Environmental Tracers
Edited by
Trevor Elliot
Printed Edition of the Special Issue Published in Water
www.mdpi.com/journal/water
Trevor Elliot (Ed.)
Environmental Tracers
This book is a reprint of the Special Issue that appeared in the online, open access journal,
Water (ISSN 2073-4441) from 2013–2014 (available at:
http://www.mdpi.com/journal/water/special_issues/environ_tracers).
Guest Editor
Trevor Elliot
Environmental Tracers Laboratory (ETL)
Environmental Engineering Research Centre (EERC)
School of Planning, Architecture & Civil Engineering (SPACE)
Queen's University Belfast
David Keir Building, Stranmillis Road
Belfast. BT9 5AG. Northern Ireland
UK
Editorial Office
MDPI AG
Klybeckstrasse 64
Basel, Switzerland
Publisher
Shu-Kun Lin
Managing Editor
Cherry Gong
1. Edition 2014
MDPI • Basel • Beijing • Wuhan
ISBN 978-3-906980-91-1 (Hbk)
ISBN 978-3-906980-92-8 (PDF)
© 2014 by the authors; licensee MDPI, Basel, Switzerland. All articles in this volume are
Open Access distributed under the Creative Commons Attribution license (CC BY), which
allows users to download, copy and build upon published articles even for commercial
purposes, as long as the author and publisher are properly credited, which ensures maximum
dissemination and a wider impact of our publications. However, the dissemination and
distribution of copies of this book as a whole is restricted to MDPI, Basel, Switzerland.
III
Dear Colleagues,
Aquifer resources continue to be overexploited, leaving the world’s most impoverished
(or vulnerable) populations and/or the aquatic environment at an ever increasing risk from
climate change. Adaptation strategies demand detailed evaluation and management of
water as a resource, requiring an understanding of the chemical, geological
(hydrogeological/geohydrological) and biological interactions that waters effect or undergo in
the hydrologic cycle. Environmental tracers are ambient natural or man-made compounds
widely distributed in the Earth’s near-surface. They may be injected into the hydrological
system from the atmosphere at recharge and/or are added/lost/exchanged inherently as waters
flow over and through materials. Variations in their chemical abundances and isotopic
compositions can be used as tracers to determine sources (provenance), pathways (of reaction
or interaction) and also timescales (dating) of environmental processes. Water dating may
invoke their characteristic decay or accumulation functions, (cf. radioactive and radiogenic
compounds and isotopes) in a system or the characteristic injection of sources. Environmental
tracers in groundwater systems can give information both on current and past flow conditions
independently of hydraulic analyses and modelling. Thus, environmental tracers generically
are important tools for developing sustainable management policies for the protection of
water resources and the aquatic environment.
Recent overviews have highlighted how most environmental tracer systematics have
become well-established through proof-of-concept studies in geochemically and hydraulically
simple aquifers. The challenge now lies in enhancing the way they are put to use by the
hydrologic community and water resource managers in more complex systems (e.g., interand intra-aquifer mixing; aquifers as distributed water systems – water coming in at one point
is going somewhere, and pumping of water represents an interception) and how they may be
used to address issues of vulnerability, sustainability, and uncertainty in water resource
systems (including resource, flooding, drought, climate justice, water and food security, water
footprints, etc.).
Therefore we would like to call for papers to disseminate and share findings especially on
the robustness or fitness-for-purpose of the application and use of environmental tracers in
water resource systems in addressing problems and opportunities scientifically. Papers are
selected by a rigorous peer review procedure with the aim of rapid and wide dissemination of
research results, development and application in the wide area of environmental tracers.
Original research papers or critical reviews are invited.
Trevor Elliot, PhD
Guest Editor
IV
Biographical Sketch - Guest Editor
Dr. Elliot is a Reader in Environmental Engineering at Queen’s University Belfast (QUB).
His training was in Geophysics & Planetary Physics (Newcastle University, 1983); his PhD in
Groundwater Geochemistry (Bath University, 1990) under the guidance of the late John N.
Andrews. After research positions at Exeter (Geology), Cambridge (Earth Sciences), and
Newcastle (Water Resources Systems Research Laboratory, Civil Engineering), in 1999 he
joined QUB and established the Environmental Tracers Laboratory (ETL) in the School of
Civil Engineering (now SPACE) as a component laboratory of the Environmental
Engineering Research Centre. The ETL aims to: advance the use of natural and applied tracers
and intelligent tracing (i.e., specifically utilising the physico-chemical properties of particular
tracers to target system characteristics) approaches in geohydrology; investigate and &
characterise environmental and engineering systems, so as to promote sustainable use and
practice and to increase fundamental understanding of gas and dissolved phase
bio/geochemical processes; mathematically model and predict fate and transport processes for
compounds of environmental concern. His research interests to date cover: intelligent tracers
for environmental & engineering systems; geohydrology; isotope hydrology; groundwater
dating; aquifer sustainability issues. He has worked on a number of major aquifer systems:
single porosity media in Algeria (sandstone), China (gravel) and Germany (gravel/sands);
dual-porosity Chalks (southern England; East Yorkshire); dual-permeability, fissured
sandstone (Northern Ireland); a swallow-hole karst system (Republic of Ireland); and the
local/regional hydrology of a sabkha system (Saudi Arabia). As well as promoting the use of
noble gases for the direct-measurement gas evasion technique to study reaeration in rivers, in
engineering systems focus has been particularly on tracing subsurface, permeable reactive
barriers (PRB’s), mixing dynamics for pumped minewater systems, and the carbon isotope
fractionation of chlorofluorocarbons (CFCs) during degradation.
V
Table of Contents
List of Contributors ........................................................................................................... VIII
Preface .................................................................................................................................XI
Trevor Elliot
Environmental Tracers
Reprinted from: Water 2014, 6(11), 3264-3269 ...................................................................... 1
http://www.mdpi.com/2073-4441/6/11/3264
Section 1: Stable Isotopes of Water (δ2H, δ18O)
Luc Lambs, Issam Moussa and Frederic Brunet
Air Masses Origin and Isotopic Tracers: A Study Case of the Oceanic and Mediterranean
Rainfall Southwest of France
Reprinted from: Water 2013, 5(2), 617-628 ........................................................................... 8
http://www.mdpi.com/2073-4441/5/2/617
Hsin-Fu Yeh, Hung-I Lin, Cheng-Haw Lee, Kuo-Chin Hsu and Chi-Shin Wu
Identifying Seasonal Groundwater Recharge Using Environmental Stable Isotopes
Reprinted from: Water 2014, 6(10), 2849-2861 ................................................................... 20
http://www.mdpi.com/2073-4441/6/10/2849
Nikolaus H. Buenning, Lowell Stott, Lisa Kanner and Kei Yoshimura
Diagnosing Atmospheric Influences on the Interannual 18O/16O Variations in Western U.S.
Precipitation
Reprinted from: Water 2013, 5(3), 1116-1140 ..................................................................... 33
http://www.mdpi.com/2073-4441/5/3/1116
Naoki Kabeya, Akira Shimizu, Jian-Jun Zhang and Tatsuhiko Nobuhiro
Effect of Hydrograph Separation on Suspended Sediment Concentration Predictions in a
Forested Headwater with Thick Soil and Weathered Gneiss Layers
Reprinted from: Water 2014, 6(6), 1671-1684 ..................................................................... 58
http://www.mdpi.com/2073-4441/6/6/1671
VI
Marco Doveri and Mario Mussi
Water Isotopes as Environmental Tracers for Conceptual Understanding of Groundwater
Flow: An Application for Fractured Aquifer Systems in the “Scansano-Magliano in Toscana”
Area (Southern Tuscany, Italy)
Reprinted from: Water 2014, 6(8), 2255-2277 ..................................................................... 72
http://www.mdpi.com/2073-4441/6/8/2255
Section 2: Multi-Isotope Studies
Christopher J. Eastoe and Ryan Rodney
Isotopes as Tracers of Water Origin in and Near a Regional Carbonate Aquifer: The Southern
Sacramento Mountains, New Mexico
Reprinted from: Water 2014, 6(2), 301-323 ......................................................................... 96
http://www.mdpi.com/2073-4441/6/2/301
Peter W. Swarzenski, Mark Baskaran, Robert J. Rosenbauer, Brian D. Edwards and
Michael Land
A Combined Radio- and Stable-Isotopic Study of a California Coastal Aquifer System
Reprinted from: Water 2013, 5(2), 480-504 ....................................................................... 119
http://www.mdpi.com/2073-4441/5/2/480
Michael Schubert, Jan Scholten, Axel Schmidt, Jean François Comanducci, Mai Khanh
Pham, Ulf Mallast and Kay Knoeller
Submarine Groundwater Discharge at a Single Spot Location: Evaluation of Different
Detection Approaches
Reprinted from: Water 2014, 6(3), 584-601 ....................................................................... 145
http://www.mdpi.com/2073-4441/6/3/584
Section 3: Investigations Using Natural Tracers in Combination with Applied
Tracers
Rory Cowie, Mark W. Williams, Mike Wireman and Robert L. Runkel
Use of Natural and Applied Tracers to Guide Targeted Remediation Efforts in an Acid Mine
Drainage System, Colorado Rockies, USA
Reprinted from: Water 2014, 6(4), 745-777 ....................................................................... 164
http://www.mdpi.com/2073-4441/6/4/745
VII
Andrew Benson, Matthew Zane, Timothy E. Becker, Ate Visser, Stephanie H.
Uriostegui, Elizabeth DeRubeis, Jean E. Moran, Bradley K. Esser and Jordan F. Clark
Quantifying Reaeration Rates in Alpine Streams Using Deliberate Gas Tracer Experiments
Reprinted from: Water 2014, 6(4), 1013-1027 ................................................................... 198
http://www.mdpi.com/2073-4441/6/4/1013
Jordan F Clark, Sheila Morrissey, Jason Dadakis, Adam Hutchinson and Roy Herndon
Investigation of Groundwater Flow Variations near a Recharge Pond with Repeat Deliberate
Tracer Experiments
Reprinted from: Water 2014, 6(6), 1826-1839 ................................................................... 213
http://www.mdpi.com/2073-4441/6/6/1826
VIII
List of Contributors
Mark Baskaran: Department of Geology, Wayne State University, Detroit, MI 48202, USA
Timothy E. Becker: Department of Earth Science, University of California, Santa Barbara,
CA 93106, USA
Andrew Benson: Program of Environmental Studies, University of California, Santa Barbara,
CA 93106, USA
Nikolaus H. Buenning: Department of Earth Sciences, University of Southern California,
Zumberge Hall of Science (ZHS), 3651 Trousdale Pkwy, Los Angeles, CA 90089, USA
Frederic Brunet: Laboratory of Functional Ecology and Environment (EcoLab), UPS, INPT,
University of Toulouse,118 route de Narbonne, Toulouse F-31062, France; Laboratory of
Functional Ecology and Environment (EcoLab), CNRS, 118 route de Narbonne, Toulouse F31062, France
Jordan F Clark: Department of Earth Science, University of California, Santa Barbara, CA
93106, USA; Program of Environmental Studies, University of California, Santa Barbara, CA
93106, USA
Jean François Comanducci: International Atomic Energy Agency (IAEA)–Environment
Laboratories, 98000 Monaco
Rory Cowie: Institute of Arctic and Alpine Research, University of Colorado Boulder,
Boulder, CO 80309, USA
Jason Dadakis: Orange County Water District, Fountain Valley, CA 92728, USA
Elizabeth DeRubeis: Department of Earth and Environmental Sciences, California State
University East Bay, Hayward, CA 94542, USA
Marco Doveri: National Research Council of Italy, Institute of Geosciences and Earth
Resources, Via G. Moruzzi 1 56124, Pisa, Italy
Christopher J. Eastoe: Geosciences Department, University of Arizona, Tucson, AZ 85721,
USA
Brian D. Edwards: U.S. Geological Survey, Menlo Park, CA 94025, USA
Trevor Elliot: Environmental Tracers Laboratory (ETL), Environmental Engineering
Research Centre (EERC), School of Planning, Architecture & Civil Engineering (SPACE),
Queen’s University Belfast, David Keir Building, Stranmillis Road, Belfast, BT9 5AG
Northern Ireland, UK
Bradley K. Esser: Chemical Science Division, Lawrence Livermore National Laboratory,
Livermore, CA 94551, USA
Roy Herndon: Orange County Water District, Fountain Valley, CA 92728, USA
Adam Hutchinson: Orange County Water District, Fountain Valley, CA 92728, USA
Naoki Kabeya: Kyushu Research Centre, Forestry and Forest Product Research Institute, 411-16 Kurokami, Kumamoto 860-0862, Japan
Lisa Kanner: Department of Earth Sciences, University of Southern California, Zumberge
Hall of Science (ZHS), 3651 Trousdale Pkwy, Los Angeles, CA 90089, USA
IX
Kay Knoeller: Helmholtz-Centre for Environmental Research–UFZ, Leipzig 04318,
Germany
Luc Lambs: Laboratory of Functional Ecology and Environment (EcoLab), UPS, INPT,
University of Toulouse,118 route de Narbonne, Toulouse F-31062, France; Laboratory of
Functional Ecology and Environment (EcoLab), CNRS, 118 route de Narbonne, Toulouse F31062, France
Cheng-Haw Lee: Department of Resources Engineering, National Cheng Kung University,
Tainan 701, Taiwan
Kuo-Chin Hsu: Department of Resources Engineering, National Cheng Kung University,
Tainan 701, Taiwan
Michael Land: U.S. Geological Survey, San Diego, CA 92101, USA
Hung-I Lin: Department of Resources Engineering, National Cheng Kung University, Tainan
701, Taiwan
Ulf Mallast: Helmholtz-Centre for Environmental Research–UFZ, Leipzig 04318, Germany
Jean E. Moran: Department of Earth and Environmental Sciences, California State
University East Bay, Hayward, CA 94542, USA
Sheila Morrissey: Department of Earth Science, University of California, Santa Barbara, CA
93106, USA
Issam Moussa: Laboratory of Functional Ecology and Environment (EcoLab), UPS, INPT,
University of Toulouse,118 route de Narbonne, Toulouse F-31062, France; Laboratory of
Functional Ecology and Environment (EcoLab), CNRS, 118 route de Narbonne, Toulouse F31062, France
Mario Mussi: National Research Council of Italy, Institute of Geosciences and Earth
Resources, Via G. Moruzzi 1 56124, Pisa, Italy
Tatsuhiko Nobuhiro: Hokkaido Research Centre, Forestry and Forest Product Research
Institute, 7 Hitsujigaoka, Toyohira, Sapporo, Hokkaido 062-8516, Japan
Mai Khanh Pham: International Atomic Energy Agency (IAEA)–Environment Laboratories,
98000 Monaco
Robert J. Rosenbauer: U.S. Geological Survey, Santa Cruz, CA 95060, USA
Ryan Rodney: Newmont Mining Corporation, 1655 Mountain City Highway, Elko, NV
89801, USA
Robert L. Runkel: U.S. Geological Survey, 3215 Marine St., Boulder, CO 80303, USA
Axel Schmidt: German Federal Institute of Hydrology, Koblenz 56068, Germany;
Helmholtz-Centre for Environmental Research–UFZ, Leipzig 04318, Germany
Jan Scholten: Institute of Geosciences, University Kiel, Kiel 24118, Germany
Michael Schubert: Helmholtz-Centre for Environmental Research–UFZ, Leipzig 04318,
Germany
Akira Shimizu: Kyushu Research Centre, Forestry and Forest Product Research Institute, 411-16 Kurokami, Kumamoto 860-0862, Japan
Lowell Stott: Department of Earth Sciences, University of Southern California, Zumberge
Hall of Science (ZHS), 3651 Trousdale Pkwy, Los Angeles, CA 90089, USA
Peter W. Swarzenski: U.S. Geological Survey, Santa Cruz, CA 95060, USA
X
Stephanie H. Uriostegui: Department of Earth Science, University of California, Santa
Barbara, CA 93106, USA; Chemical Science Division, Lawrence Livermore National
Laboratory, Livermore, CA 94551, USA
Ate Visser: Chemical Science Division, Lawrence Livermore National Laboratory,
Livermore, CA 94551, USA
Mark W. Williams: Institute of Arctic and Alpine Research, University of Colorado Boulder,
Boulder, CO 80309, USA
Mike Wireman: U.S. Environmental Protection Agency, Region 8, Denver, CO 80202, USA
Chi-Shin Wu: Department of Resources Engineering, National Cheng Kung University,
Tainan 701, Taiwan
Hsin-Fu Yeh: Department of Resources Engineering, National Cheng Kung University,
Tainan 701, Taiwan
Kei Yoshimura: Atmosphere and Ocean Research Institute, University of Tokyo, General
Research Building 211a, 5-1-5 Kashiwanoba, Kashiwa Chiba 277-8568, Japan
Matthew Zane: Department of Earth Science, University of California, Santa Barbara, CA
93106, USA
Jian-Jun Zhang: College of Soil and Water Conservation, Beijing Forestry University, No.
35 Qinghuadong Road, Haidian District, Beijing 100083, China
XI
Preface
Environmental Tracers are ambient, natural or man-made compounds. They are widely
distributed in the Earth’s near-surface and can be both (i) added/lost/exchanged inherently (as
waters flow over and through materials) and (ii) discriminated, determined analytically, and
tracked sensitively.
They can also be applied (injected) deliberately at levels higher than the ambient
background to characterize environmental (and indeed engineering) systems. Environmental
Tracers, including their stable and radioactive isotopic signatures, when considered alongside
elemental abundances, generically provide important tools for understanding the source, flow,
and mixing dynamics of water resource systems through their imprint on the system or their
sensitivity to alteration within it, and especially for subsurface waters (groundwaters).
Since the pioneering work of the 1950’s and 1960’s, isotopic signatures of waters in the
hydrological cycle have been well-characterized in terms of stable isotopes (oxygen and
hydrogen—18O; 2H) of the water molecule itself as a natural, environmental tracer. Carbon
(13C -) isotopes of dissolved compounds also started to be investigated around this period; the
range of elements covered (isotopic signatures of 15N, 34S, compounds, etc.) and the
techniques available for analyzing them in various, dissolved compounds has subsequently
increased rapidly over the ensuing decades; elements covered now include radioactive
environmental isotopes (cf. the radioactive isotope of hydrogen occurring in water itself—
tritium, 3H). Over the decades, researchers have investigated and characterized many water
resource systems with these tools in their hydrogeological arsenal. Whilst the effectiveness
(and weaknesses, as necessary) of these tracers in discriminating aquifer dynamics and
processes has been corroborated in proof of concept studies, their uptake as a standard
hydrogeological, geohydrological or water resource tool by the professions is still
disappointing. This Special Issue focuses therefore particularly on the robustness or fitnessfor-purpose of the application and use of Environmental Tracers in addressing problems and
opportunities scientifically, to promote their wider use, and to address substantive issues of
vulnerability, sustainability, and uncertainty in (ground)water resources systems and their
management.
XII
The Special Issue arose as an open call to the Water community, with rigorous peer-review
applied to selected and published papers. Eleven papers from researchers, based in seven
different countries, are included in this Special Issue. The papers reflect global research in this
area. My hope then is that the collation of these papers contributes to piquing further interest
and action in how Environmental Tracers can contribute and be used to address the
substantive issues of vulnerability, sustainability, and uncertainty in (ground)water resources
systems and their management.
Trevor Elliot
Guest Editor
School of Planning, Architecture and Civil Engineering (SPACE), Queen’s University
Belfast, UK.
5 November, 2014
1
Environmental Tracers
Trevor Elliot
Abstract: Environmental tracers continue to provide an important tool for understanding the
source, flow and mixing dynamics of water resource systems through their imprint on the system or
their sensitivity to alteration within it. However, 60 years or so after the first isotopic tracer studies
were applied to hydrology, the use of isotopes and other environmental tracers are still not routinely
necessarily applied in hydrogeological and water resources investigations where appropriate. There
is therefore a continuing need to promote their use for developing sustainable management policies
for the protection of water resources and the aquatic environment. This Special Issue focuses on the
robustness or fitness-for-purpose of the application and use of environmental tracers in addressing
problems and opportunities scientifically, to promote their wider use and to address substantive
issues of vulnerability, sustainability, and uncertainty in (ground)water resources systems and
their management.
Reprinted from Water. Cite as: Elliot, T. Environmental Tracers. Water 2014, 6, 3264-3269.
1. Introduction
Environmental tracers are ambient, natural or man-made compounds widely distributed in the
Earth’s near-surface. They may be injected naturally into the hydrological system from the
atmosphere at recharge and/or are added/lost/exchanged inherently as waters flow over and through
materials. Because of possible issues of chemical equivalence in sampled groundwater signatures, to
screen hypotheses and identify contributing processes typically detailed and/or quantitative
groundwater investigations require consideration of multiple tracers of chemical reactions: ion
ratios/correlations, minor/trace element determinands, mineralogy and reaction-path simulations,
isotopic compositions, etc. These tracers separately or in conjunction may suggest a unique solution
in detailing the hydrogeochemical processes operating. As for trace elements, environmental isotopic
signatures of dissolved compounds, and the water molecule itself, can prove particularly sensitive
tracers, as they occur generally at low levels of concentration and can be affected by chemical and
physical fractionation effects shifting their signatures. Variations then in groundwater chemical
abundances and isotopic compositions can be used as natural tracers to determine sources
(provenance), pathways (of reaction or interaction), and also timescales (dating) of environmental
processes. Dating may invoke their characteristic decay or accumulation functions (i.e., radioactive
vs. radiogenic compounds and isotopes) in a system, or the characteristic injection of sources. This
then provides a handhold on timescales of processes and water residence times to be considered
critically alongside hydraulic transit (flowthrough) and system turnover times. Moreover,
environmental tracers in groundwater systems can give information both on current and past flow
conditions independently of hydraulic analyses and groundwater modelling. Thus, generically
environmental tracers are important tools for developing sustainable management policies for the
protection of water resources (quantity and quality) and the aquatic environment. For example, where
investigations are taking place in groundwater systems in which mixing is a potentially important
2
process (cf. pumping flooded mines (see below) or even pumped public supply wells), the addition of
even modest amounts of additional sampling and analysis for isotopic environmental tracers can greatly
expand understanding of the flow and geochemical/water quality (mixing) dynamics of pumped
groundwater systems [1].
Recent overviews (e.g., [2]) have highlighted how most environmental tracers systematics have
now become well-established through proof-of-concept studies in geochemically and hydraulically
simple aquifers. The challenge now lies in enhancing the way they are put to use by the hydrologic
community and water resource managers in more complex systems, and how they may be used to
address issues of vulnerability, sustainability, and uncertainty of water resources and their systems.
Therefore the aim of this Special Issue is to disseminate and share findings especially on the
robustness or fitness-for-purpose of the application and use of environmental tracers in water
resource systems in addressing problems and opportunities scientifically. Original research papers
were selected by rigorous peer-review process with the aim of rapid, accessible and wide
dissemination of results.
2. Contributions
The selected papers presented in the Special Issue are highlighted in this section. They fall broadly
into three categories: those focused particularly on the stable oxygen- and hydrogen-isotopes and
also the radioactive hydrogen-isotope (tritium) of the water molecule to characterise its source(s),
fractionation effects, and dating young groundwater systems (five papers); those focused on
multi-isotope approaches (including the use of other radioisotopes, and significantly natural
Uranium- and Thorium-series radionuclides) (three papers); and those using natural environmental
tracers alongside or as applied tracers introduced into groundwater systems (three papers).
2.1. Stable Isotopes of Water (į2H, į18O)
Three articles [3–5] look at the start of the hydrological cycle, at the origin of rainfall and
its selection in groundwater recharge using characteristic, naturally-occurring į2H and į18O signatures
in monthly samples. The concept of “deuterium excess”, originally defined by Dansgaard [6] as
d = į2H í 8.į18O, is used in case studies by Lambs et al. [3] in France and Yeh et al. [4] in Taiwan, to
identify the predominance of different air masses contributing to rainfall patterns seasonally and the
groundwater recharge at the regional/catchment scale. Buenning et al. [5] model specifically the į18O
of annual rainfall across the western USA to identify mechanisms controlling interannual variations in
isotopic signatures. Such signatures feed into groundwater systems and might be used as climate
proxy for various surface archives and potentially [5] to monitor storm track changes. Overall,
isotope mapping and monitoring on multiple spatial and temporal scales is helping recognise and
characterise isoscapes, of use in a wide range of hydrological and other contexts [7].
A fourth contribution by Kabeya et al. [8] links a storm event over a 48-h. period to sediment
production and transport in a forested headwater catchment in Japan. Their correlation is used to
identify the source area for sediment, which might be linked potentially with mobilisation of
radioactive caesium deposited following the recent Fukushima Daiichi nuclear power plant incident.
3
Doveri and Mussi [9] use both the stable isotopes (į2H, į18O) of water and its radioactive
isotope (tritium, 3H, with a half-life of 12.3 years) monitored in springs and wells (up to six times over
a 2-year period) as natural environmental tracers to suggest a conceptual model of groundwater flow
in the Scansano-Magliano region of southern Tuscany (Italy). The local sandstone aquifer here may
provide a strategic alternative source for water supply given the overexploitation and contamination
of known, local alluvial aquifers, and for supplying isolated villages on the Toscana ridge in this
semiarid area.
2.2. Multi-Isotope Studies
In another regional aquifer study, and alongside the water isotopes (į2H, į18O, 3H), Eastoe and
Rodney [10] additionally utilise natural C-isotope systematics (stable į13C, and radioactive 14C—this
latter with a half-life of 5730 years) of dissolved inorganic carbon, and the S-isotope signature of
dissolved sulfate (į34S). The environmental tracers are used to date groundwater residence times (3H,
14
C) and to identify winter recharge at high-elevation in the Sacramento Mountains (USA) as the source
of groundwaters in the regional carbonate aquifer and flanking basins. The authors highlight therefore
that if winter rainfall decreases e.g., as a result of global climate change then aquifer recharge
also will be affected. However, the authors also show that the use of stable isotopes of water
equivocally here of themselves cannot help discriminate sources of groundwater for the more distant
Roswell basin.
Swarzenski et al. [11] also combine stable and radioactive tracers alongside hydrochemistry in a
coastal Los Angeles County (USA) study to investigate the complex groundwater scenario associated
with a historical scheme to inject freshwater at three locations as a hydraulic barrier to saline
intrusion. Here groundwater may be a “complex mixture of native groundwater, intruded seawater,
non-native injected water, and oil-field brine water.” The baseline geochemical study to discriminate
sources and characterise groundwater dynamics and mixing included looking at natural Uranium- and
Thorium-series radiogenic nuclides (223Ra, 224Ra, 226Ra, 228Ra, 222Rn) particularly to study potential
controls on the adsorption-desorption rates for dissolved cations. Schubert et al. [12] evaluate the
use of the stable isotopes of water alongside natural radium isotopic ratios (224Ra/228Ra, 228Ra/226Ra)
and radon (222Rn) contents to detect submarine groundwater discharge (SGD) in a submarine cave
near Monaco. They found 222Rn contents the most robust indicator in context and į2H, į18O suitable.
They propose particularly therefore utilising 222Rn contents alongside salinity for the investigation
of offshore fresh groundwater reserves. Such SGD and offshore groundwater reserves have recently
been highlighted as a global phenomenon potentially to be exploited [13]. Elsewhere, other
authors [14] have used dissolved uranium and 234U/238U activity ratios and natural Uranium- and
Thorium-series radiogenic nuclides, alongside dissolved noble gases (including 4He, the stable
by-product of U- and Th-series radioactive decay) to attest to the “fossil” water (palaeowater) status of
groundwaters in the regional Continental Intercalaire aquifer of Algeria and Tunsia. Their abstraction
is therefore akin to mining this groundwater resource. Where a freshwater/saline water interface is
identifiable, even the use of scavenger well technology (e.g., [15,16]) must take into account the
potential paleowater status of at least a component of the pumped water and the effect on the movement
of the interface.
4
2.3. Investigations Using Natural Tracers in Combination with Applied Tracers
In complex hydrogeological settings the use of applied tracers to analyse system flow-through
(transit) times and hydraulic connections can be problematic as the tracers can be diluted out, or if the
mixing reservoir is overestimated there can be tracer breakthrough at concentrations exceeding
environmental or analytical specifications.
Three further USA case studies [17–19] deploy applied tracers alongside ambient tracers.
Cowie et al. [17] use a two-tier investigative approach in a flooded, abandoned, underground hardrock mine system (USA). They monitored natural environmental tracers of water (į2H, į18O, 3H) to
gain a conceptual understanding of the hydrogeology, and then injected applied tracers (here ionic and
fluorescent tracers) at specific locations to focus their investigations. Remediation efforts for acid
mine drainage (AMD) then can be targetted at separating and isolating or preventing sources of poor
quality water. This approach then essentially follows the philosophy of Source-Pathway-Receptor
(SPR) risk management (for Contaminated Land investigations) rather than a traditional end-of-pipe
remedial solution for such waters. Elsewhere [1], monitoring of natural environmental tracers during
pumping of AMD waters to control the water table in the system and prevent environmental impacts
is used to characterise the sources, dynamics and evolution of mixing of waters and their water quality
for flooded, abandoned coal mines.
Benson et al. [18] use injected, applied gas tracers (SF6 and Xe) as nonpolar, partitioning tracers
for the air-water interface in streams to monitor gas loss downstream and oxygen reaeration (KDO or
K2, mass transfer coefficients). The link with environmental tracing is twofold, both because these
two tracer gases are already present in the environment sourced from the atmosphere (albeit at very
low ambient, dissolved concentrations which allows significant enhancement of dissolved
concentrations for tracing), and because the loss rate for these gases could be used also for
interpretation of 222Rn measurements in the streams to study then groundwater-surface water
interactions. Especially noble gases (like Xe) are perceived as “environmentally-friendly” tracers for
stream reaeration and gas evasion/mass transfer (K) studies, as well as for characterising
hydrodynamic properties [20].
Finally, Clark et al. [19] have applied (injected) the environmental gas tracers SF6 and Xe in
groundwater to characterise hydraulic connections between recharge facilities and production wells of a
Managed Aquifer Recharge (MAR) site. MAR involves a strategy of injecting “surface waters into
aquifers for storage and later extraction”. Each tracer was injected separately and following an
intervening period of a decade between the two tests to assess whether operation of the MAR facility
(including any changes in recharge conditions) therefore had impacted the system response.
3. Conclusions
Eleven original research articles have been selected for this Special Issue. The research findings
are novel and timely in informing the hydrological and water resources management communities
on up-to-date research and practice. Sixty years or so after the first isotopic tracer studies applied to
hydrology the use of isotopes and other environmental tracers are still not necessarily routinely
applied in hydrogeological and water resources investigations where appropriate. We trust that the
5
collation of these papers contributes to piquing further interest in how environmental tracers can
contribute and be used to address substantive issues of vulnerability, sustainability, and uncertainty
in (ground)water resources systems and their management.
Acknowledgments
The Guest Editor (TE) thanks both the research community for offering and contributing a wide
range of valuable papers, and the publisher MDPI for allocating resources and support towards this
Special Issue.
Conflicts of Interest
The author declares no conflict of interest.
References
1.
Elliot, T.; Younger, P.L. Detection of Mixing Dynamics during Pumping of a Flooded Coal
Mine. Groundwater 2014, 52, 251–263.
2. Sanford, W.E.; Aeschbach-Hertig, W.; Herczeg, A.L. Preface: Insights from Environmental
Tracers in Groundwater Systems. Hydrogeol. J. 2011, 19, 1–3.
3. Lambs, L.; Moussa, I.; Brunet, F. Air Masses Origin and Isotopic Tracers: A Study Case of the
Oceanic and Mediterranean Rainfall Southwest of France. Water 2013, 5, 617–628.
4. Yeh, H.F.; Lin, H.I.; Lee, C.H. Identifying Seasonal Groundwater Recharge Using
Environmental stable isotopes. Water 2014, 6, 2849–2861.
5. Buenning, N.H.; Stott, L.; Kanner, L.; Yoshimura, K. Diagnosing Atmospheric Influences on
the Interannual 18O/16O Variations in Western U.S. Precipitation. Water 2013, 5, 1116–1140.
6. Dansgaard, W. Stable Isotopes in Precipitation. Tellus 1964, 16, 436–468.
7. West, J.B., Bowen, G.J., Dawson, T.E., Tu, K.P., Eds. Isoscapes: Understanding Movement,
Pattern and Process on Earth through Isotope Mapping; Springer: New York, NY, USA, 2010;
ISBN 978-90-481-3353-6; pp. 1–487.
8. Kabeya, N.; Shimizu, A.; Zhang, J.J.; Nobuhiro, T. Effect of Hydrograph Separation on
Suspended Sediment Concentration Predictions in a Forested Headwater with Thick Soil and
Weathered Gneiss Layers. Water 2014, 6, 1671–1684.
9. Doveri, M.; Mussi, M. Water Isotopes as Environmental Tracers for Groundwater Flow
Understanding: An Application on Fractured Aquifer Systems in the Scansano-Magliano in
Toscana area (southern Tuscany-Italy). Water 2014, 6, 2255–2277.
10. Eastoe, C.J.; Rodney, R. Isotopes as Tracers of Water Origin in and Near a Regional Carbonate
Aquifer: The Southern Sacramento Mountains, New Mexico. Water 2014, 6, 301–323.
11. Swarzenski, P.W.; Baskaran, M.; Rosenbauer, R.J.; Edwards, B.D.; Land, M. A Combined
Radio- and Stable-Isotopic Study of a California Coastal Aquifer System. Water 2013, 5,
480–504.
6
12. Schubert, M.; Scholten, J.; Schmidt, A.; Comanducci, J.F.; Pham, M.K.; Mallast, U.; Knoeller, K.
Submarine Groundwater Discharge at a Single Spot Location: Evaluation of Different Detection
Approaches. Water 2014, 6, 584–601.
13. Post, V.E.A.; Groen, J.; Kooi, H.; Person, M.; Ge, S.; Edmunds, W.M. Offshore Fresh
Groundwater Reserves as a Global Phenomenon. Nature 2013, 504, 71–78.
14. Elliot, T.; Bonotto, D.M.; Andrews, J.N. Uranium, Radium and Radon evolution in the
Continental Intercalaire aquifer, Algeria and Tunisia. J. Environ. Radioac. 2014, 137,150–162.
15. Stoner, R.F.; Bakiewicz, W. Scavenger Wells-1-Historic Development. In Study and Modelling
of Saltwater Intrusion into Aquifers, Proceedings of the 12th Saltwater Intrusion Meeting,
Barcelona, Spain, 1–6 November 1992; Custodio, E., Galofr, A., Eds.; CIMNE: Barcelona, Spain;
pp. 545–556.
16. Aliewi, A.S.; Mackay, R.; Jayyousi, A.; Nasereddin, K.; Mushtaha, A.; Yaqubi, A. Numerical
Simulation of the Movement of Saltwater Under Skimming and Scavenger Pumping in the
Pleistocene Aquifer of Gaza and Jericho Areas, Palestine. Transp. Porous Media 2001, 43,
195–212.
17. Cowie, R.; Williams, M.W.; Wireman, M.; Runkel, R.L. Use of Natural and Applied Tracers to
Guide Targeted Remediation Efforts in an Acid Mine Drainage System, Colorado Rockies,
USA. Water 2014, 6, 745–777.
18. Benson, A.; Zane, M.; Becker, T.E.; Visser, A.; Uriostegui, S.H.; DeRubeis, E.; Moran, J.E.;
Esser, B.K.; Clark, J.F. Quantifying Reaeration Rates in Alpine Streams Using Deliberate Gas
Tracer Experiments. Water 2014, 6, 1013–1027.
19. Clark, J.F.; Morrissey, S.; Dadakis, J.; Hutchinson, A.; Herndon, R. Investigation of
Groundwater Flow Variations near a Recharge Pond with Repeat Deliberate Tracer
Experiments. Water 2014, 6, 1826–1839.
20. Semuwemba, J.; Elliot, T.; Mackinnon, P.A. Determining the Reaeration Coefficient and
Hydrodynamic Properties of Rivers Using Inert Gas Tracers. In Proceedings of the Second
International Conference on Advances in Engineering and Technology (AET 2011), Entebbe,
Uganda, 30 January–2 Febuary 2011; Mwakali, J.A., Alinaitwe, H.M., Eds.; Macmillan
Publishers: Kampala, Uganda, 2011; ISBN 978-214-00-7; pp. 670–676.
Section 1: Stable Isotopes of Water (į2H, į18O)
8
Air Masses Origin and Isotopic Tracers: A Study Case of the
Oceanic and Mediterranean Rainfall Southwest of France
Luc Lambs, Issam Moussa and Frederic Brunet
Abstract: Aquifers recharge mainly by local rainfall, which depend on the air mass humidity and
orographic lifting, causing rain. The stable isotopes of the water molecule, i.e., oxygen-18 and
deuterium, are useful tracers to determine the water source origin. Moreover, the calculation of the
deuterium excess enables one to differentiate between the air masses from the Atlantic Ocean or the
Mediterranean Sea. A transect from one coast to the other one and going through the city of Toulouse
have been made to sample the groundwater and determine their isotopic characteristic. A monthly
rainfall sampling has also been done over one year, close to the city Toulouse, to see how the d-excess
values range over the season. The discussion replaces these results in available isotopic data.
Reprinted from Water. Cite as: Lambs, L.; Moussa, I.; Brunet, F. Air Masses Origin and Isotopic
Tracers: A Study Case of the Oceanic and Mediterranean Rainfall Southwest of France. Water 2013,
5, 617-628.
1. Introduction
France presents a wide range of coasts and is influenced by both the Atlantic Ocean and the
Mediterranean Sea. The morphology of this country, which impacts the rainfall pattern, is complex,
with the Massif Central in the center, the Alps to the east and the Pyrenees to the south. The main
processes controlling the į18O and įD isotopic signatures in precipitations were summarized by
Rozanski et al. [1]; i.e., the rainfall amount, continental and altitude effects and the origin of air
masses. For France, some pioneer works have studied the isotopic pattern of rainfall, such as Celle et
al., Celle-Jeanton et al., Ladouche et al. and Millot et al. [2–5], but still, some areas lack for valuable
data, like the Southwest area. It is, therefore, important to constrain the signature of the atmospheric
signal in these geographical and geomorphological contexts by means of a rainfall-monitoring
network and groundwater measurements.
The Southwest of France presents the characteristics of having two major winds blowing in
an opposite direction. For instance, in Toulouse, the northwest wind (280° to 340°), called “Cers”,
blows during 43% of the year, bringing the moisture from the Atlantic Ocean, and the southeast wind
(120° to 180°), called “Autan”, blows during 29% of the year, bringing the Mediterranean influences.
Therefore the airport runways are oriented N-W/S-E, the planes taking them from one end or the
other depending on the wind origin.
The Autan wind is a föhn-type wind deriving from the Marin wind from the Mediterranean and
blowing over the Languedoc into the Tarn and the Garonne valleys, affecting the Lauragais (the hilly
area between Castelnaudary and Toulouse) and the Toulouse area. Channeled and intensified by the
narrowing lowlands between the Pyrenees and the south of the Massif Central, it undergoes sudden
9
acceleration between the Corbières and the Montagne Noire (Venturi effect). Gusts can exceed twice
the mean speed, possibly enhanced by lee-wave activity.
To understand the Atlantic Ocean and the Mediterranean Sea on the southwest rainfall of France,
the present work summarizes over 75 isotopic measurements of shallow groundwater, used as rainfall
natural gauges, located between these two coasts. This study is completed by an 18 months’ rainfall
survey near Toulouse in view to follow the air masses origin for the rainfall over the season and
the influence of the Mediterranean income mainly through the Autan wind.
2. Experimental Section
2.1. Study Area
The Southwest of France stretches between the Atlantic and the Mediterranean coast over
350 km; see Figures 1 and 5. For the groundwater, the sampling places were taken in alluvial plains,
i.e., along the Garonne and Adour valley at the west and along the Aude valley at the east. The
connection between the two basins is Seuil de Naurouze (189 m) located 40 km southeast of
Toulouse. Farm wells or piezometers, if available, were used for the shallow groundwater sampling.
Typical water depth ranges between 2 and 12 m. The longer groundwater survey has been made near
Verdun sur Garonne, 35 km downstream (northwest) of Toulouse, in a farm well (depth 2–4 m), one
km away from the Garonne River. It is located in the middle of the Garonne Channel and would be
classified as mixed bedrock-alluvial stream [6], and the valley contains a classic flight of clay terraces
that represent an episodic bedrock valley deepening and punctuated by lateral migration of the
deposition of coarse gravels and sands.
Figure 1. The Southwest of France with the localization of the sample sites with
the typical į18O value of the groundwater (in white) and the upstream basin (in red).
The resulting isovalue contour for į18O = í5‰, í6‰ and í7‰ are given with a white
line, and the two dominant wind directions are given by both arrows.
10
Along the west to east transect, the rainfall cumulates at the feet of the Western Pyrenees, with
1156 mm per year in Dax and a little less for Cestas-Pierroton, located north of Bordeaux with
920 mm. After, the rainfall amounts lower along the Garonne Valley with 763 mm in Agen and
660 mm in Toulouse. It increases slowly in the area of Carcasonne 736 mm, due to the proximity of
the mountains, before slowing down again in the direction of the Mediterranean Sea: 609 mm in
Narbonne, 549 mm in Perpignan and 605 mm in Sète. On the contrary, the evaporation enhance from
west to east due to the increasing insulation.
2.2. Rainfall Sampling
Close to our isotopic laboratory in Auzeville, 8 km southeast of Toulouse, a rainfall sampling was
installed during June 2011. Its consist of a polypropylene funnel (22 cm diameter) covering a 5 L
polypropylene jerrycan, containing 300 mL of paraffin oil to avoid any evaporation process. The
rainfall aliquots were sampled monthly, between July 2011 and December 2012. The meteorological
data used were obtained from the close INRA station (less than 1 km) and compared with the
long-term Toulouse airport station, Blagnac, where the record started in 1809, and located
northwest, at 18 km. For each monthly sample, 10 mL were used for į18O and įD determination.
2.3. Groundwater Sampling
The shallow groundwater survey started in 1997 at Monbequi near Verdun sur Garonne in farm
wells and piezometers to follow the influence of the Garonne River in this meander area [7,8]. The
upper well was chosen as the local ground water reference and samplings were made over 6 years.
To get a wider understanding of the alluvial groundwater characteristic, complementary samplings
were done along the Garonne Valley between 2004 and 2007, for the Adour Valley between 2005
and 2006 and at the Seuil de Naurouze and Aude Valley in 2010.
2.4. Isotopic Measurements
Glass vial of 10 mL with tight caps were used for water sampling. The vials were kept at stable
temperature before the measurement of the į18O and įD determination on a continuous flow isotope
ratio mass spectrometer. Before 2007, the water samples were send to an iso-analytical laboratory in
England (ANCA-GSL and GEO 20–20 IRMS, Europa Scientific, Crewe UK). Each water sample
was measured in triplicate for each isotope ratio. The overall mean standard deviation (SD) for the
į18O values was around ±0.15‰, and the mean SD for the į2H values was ±1.7‰. From 2010, the
measurements were done on our own isotopic platform, Shiva (Isoprim 100 and Geo-multi-flow,
Elementar, Hamburg, Germany). Each water sample was measured in duplicate. The mean standard
deviation (SD) for the rainfall measurements was around ±0.29‰ for į18O, and the overall mean SD
was ±2.5‰ for į2H.
The results are given relative to V-SMOW standards from IAEA. The deuterium excess was
calculated using the Global Meteorological Water Line (GMWL), as defined by Craig [9] and
completed by Dansgaard [10]:
d-excess = įD – 8 × į18O
(1)
11
A value of d-excess around 10 or below, i.e., similar to the GMWL, was taken as originating from
Atlantic moisture and a value around 14 [2] as moisture with water recycling as one coming from
the close West Mediterranean basin.
3. Results and Discussion
3.1. Rainfall Results
The data were obtained between July 2011 and December 2012. The annual amount of rainfall
for 2011 was 538 mm, and for 2012 it was 548 mm in the INRA station and around 20% over
the amount values from Blagnac Airport station. These numbers are under the 30 years mean
(CLINO period 1961–1990 [11]), 671 ± 121 mm per year monthly, as since 2003, the Toulouse area
is in a dry period. The obtained isotopic value for į18O ranges from í10.7‰ to í4.0‰, showing
the continental influence, with a mean monthly temperature ranging from í2.5 °C (February 2012)
for the minimum to +30.4 °C (August 2012) for the maximum. Table 1 reports the obtained isotopes
values for the monthly rainfall, with the meteorological data from the nearby INRA station.
In Figure 2, both years are plotted along a whole year and present a similar pattern. The į18O
signal presents the more depleted values in winter and the less depleted in summer, following
the temperature shift. The negative peak from February 2012 corresponds to a cold period, until
í12.5 °C on February 9, with one week of snow. The d-excess displays more erratic variations,
perhaps with some evaporative process in summer (July) and a possible Mediterranean influence
in October.
Figure 3 gives the representation of į18O versus į2H for both years. For 2011, the regression gives
in 2011 a slope of 7.70 and a d-excess of 11.28 with R2 = 0.979, and in 2012, the slope is equal to
7.70 and a d-excess of 9.85 with a R2 = 0.909. Both sets are very close with the Global Meteorological
Water Line (GMWL) defined by Craig 1961 [9]. The point above the GMWL (į18O = í7.83‰,
d-excess = +22.12) corresponds to a possible Autan wind influence with the incoming of
Mediterranean rain. The weighted mean (w-mean) of į18O or the d-excess can be calculated by
summing the monthly amount of rainfall (rainfall amountmonth) relative to the total rainfall for one
year and multiplying both monthly isotopic values, according to the following equation:
į18Ow-mean = Ȉ i = 1 to 12 į18Omonth × Rainfall amountmonth ÷ Rainfall amountyear
(2)
The obtained value for the 2012 values is į18Ow-mean = í6.32‰ ± 0.25‰ and a d-excessw-mean =
11.76 ± 0.59.
12
Table 1. Isotopic results, calculated d-excess and meteorological conditions for the
rainfall samples near Toulouse.
Month
July 2011
August 2011
September 2011
October 2011
November 2011
December 2011
January 2012
February 2012
March 2012
April 2012
May 2012
June 2012
July 2012
August 2012
September 2012
October 2012
November 2012
December 2012
į18O
( ‰)
į 2H
( ‰)
d-excess
Rainfall
Min. T
Max. T
í4.48
í4.00
í5.18
í6.59
í10.71
í5.62
í6.55
í10.22
í5.80
í6.71
í4.80
í4.67
í4.53
í6.28
í5.25
í7.83
í8.89
í8.09
í28.15
í19.02
í27.18
í36.73
í72.56
í30.45
í40.87
í71.81
í32.00
í44.17
í24.84
í27.23
í29.61
í39.35
í28.96
í40.53
í55.92
í59.80
7.69
12.98
14.24
15.98
13.12
14.51
11.53
9.95
14.40
9.51
13.56
10.10
6.63
10.89
13.05
22.12
15.20
4.93
(mm)
86.5
21.5
73.5
24.5
31.0
53.0
39.0
4.0
22.0
69.0
75.5
53.5
58.0
48.5
26.0
53.0
49.5
50.5
(°C)
14.9
16.3
14.8
9.6
9.6
5.0
3.8
í2.5
4.1
7.6
11.6
15.0
14.7
17.5
13.9
11
6.8
4.6
(°C)
25.7
29.0
27.2
21.7
15.9
12.1
10.3
6.0
17.5
16.2
22.2
26.5
27.5
30.4
24.7
20.5
14
11.6
Figure 2. Monthly variation of (a) į18O and (b) d-excess with error bars for the rainfall
in 2011 (dotted line, no symbol) and 2012 (plain line, cross symbol) near Toulouse.
(a)
13
Figure 2. Cont.
(b)
Figure 3. Relationship between the į18O and į2H rainfall values for 2011 and 2012
compared to the GMWL line (dotted line).
3.2. Groundwater Results
3.2.1. Reference Well
The survey of the reference well for the groundwater over six years (see Figure 4) reveals that
the į18O values range from í6.18‰ to í7.10‰.The overall mean value of į18O= í6.60‰ ± 0.24‰
(n = 31), whereas the annual mean values range from í6.43‰ to í6.86‰. The seasonal pattern
appears very reduced (ǻį18O = 0.2 units) for these 31 samples, with a slight depletion in spring. This
dampened isotopic amplitude (AGW) response compared to the rainfall isotopic amplitude (Arainfall)
14
can be used to calculate the mean residence time (tmr) according to the exponential model equation
of Stichler and Herrmann [12]:
(3)
tmr= 0.5ʌ × (Arainfall2/AGW2 í 1)0.5
This calculation gives a residence time of about five years.
Figure 4. Variation of the į18O values of the reference groundwater well near Verdun
sur Garonne (between 1997 and 2003) over the season. The solid line represents the
polynomial regression line.
3.2.2. Groundwater between Atlantic and Mediterranean Coasts
The numerous alluvial ground water sample along the Garonne, Adour and Aude Valleys shows
that there is a gradient from around į18O = í5.0‰ close to the Atlantic coast (í4.6‰ in Dax and í5.4‰
in Bordeaux) and slowly becoming more depleted when coming more inland. At Verdun/Garonne,
50 km downstream of Toulouse, and until Seuil de Naurouze, the highest point (189 m) where the
water flux divides itself between the Atlantic and the Mediterranean slope, the į18O reaches í6.6‰.
The continental effect continues on the east slope until Raissac, 25 km before the Mediterranean Sea,
with a value of į18O around í6.8‰. Along this sea coast, the isotopic values become less depleted,
with į18O value close to í5.5‰.
On the contrary, the d-excess starts close to 10 (i.e., the value of GMWL) along the Atlantic coast,
before dropping inland, until Toulouse, around 6. From Seuil de Naurouze, the d-excess increases
from 8.2, to 9.8 in Raissac, before reaching 12.6 at La Franqui. The Mediterranean Sea is a closed
system and displays a higher d-excess value, around 14 [2]. The effect of the Mediterranean Sea
through the Autan wind affects the d-excess, meaning that the water vapor recycles until Seuil de
Naurouze, but the isotopic signal of the of the Continental effect coming from the Atlantic coast
stops only a few ten km before reaching the Mediterranean coast, see Figure 5.
15
Figure 5. Continental gradient from the west Atlantic coast to east Mediterranean coast:
mean isotopic values (į18O, d-excess) for the different alluvial ground water stations along
the Garonne, Adour and Aude Valleys. The altitude of the station is given between brackets.
St Maurice (76 m)
-6.77 ± 0.03 n = 3
6.5 +/- 1.9
Atlantic coast
Mas d’Agenais (25 m)
-6.25 ± 0.22 n = 2
Bordeaux wineyard (5–20 m) 6.63 ± 3.5
Seuil de Naurouze (189 m)
Verdun / G (88m)
-6.60 ± 0.24 n=31
6.0 ± 1.7 n=3
-6.60 ± 0.28 n = 2
8.2 ± 1.9
Mediterranean coast
Raissac (20 m)
-6.83 ± 0.41 n=7
9.8 ± 3.1
-5.59 ± 0.39 n = 22
La Franqui (5 m)
-5.53 ± 0.10 n=2
12.6 ± 4.4
0
250
350 km
Messange (5 m)
Gave Pau (151 m)
-4.86 ± 0.47 n = 2
-6.07
± 0.03 n = 4
9.2 ± 3.1 Adour (5–25 m)
-5.59 ± 0.19 n = 2
3.3. Discussion
Most shallow groundwater is of meteoric, i.e., atmospheric origin. Rainwater directly infiltrates
the ground or indirectly via the inflow of surface water or from bank storage in streams [13]. Shallow
and locally-derived ground waters are often used to characterize the isotopic content of meteoric
waters, due to the conservative nature of the stable isotope composition of water in an aquifer [14].
Unconsolidated sediments, like in a river valley, are excellent and most efficient aquifers. Their
porosity and permeability are usually high [13]. We have measured the groundwater velocity in Verdun
near our reference well, and the velocity ranges from 2.1 to 3.2 m/day [15]. Such alluvial plains
present high precipitation response to rainfall, and the residence time of ground water is short.
In general, alluvial shallow groundwaters are known to be two to 50 years old [16]. In our case,
with the equation of Stichler [12], we found a resident time of five years for our groundwater
reference well.
On the contrary, deeper groundwater in the Adour-Garonne basin is characterized by more
depleted values than the actual mean isotopic rainfall, with values ranging from į18O = í5.6‰
to í10.6‰ [17]. For instance, in the Southwest of France, we found isotopic values ranging from
į18O= í7.2‰ to í7.8‰, (Lambs, unpublished results) for many tens of meters-depth bore wells.
Such heterogeneity in the į18O signature for a Paleocene aquifer reflects a variable recharge, either
in space and time. The most depleted values correspond to a water recharged with a colder climate
than the present one [17].
From the isotopic results around Toulouse, there is very good agreement between the Verdun
groundwater annual average value (į18Omean = í6.60‰ ± 0.25‰) and the rainfall (į18Omean = í6.63‰
± 1.81‰). For the calculated d-excess, there is a small shift with a mean value of d-excessmean = 6.0 ± 1.7
for the groundwater and d-excessmean = 11.8 ± 4.43 for the rainfall. Notice also that the groundwater
16
taken at the Naurouze pass in 2010 gives similar values (į18O = í6.80‰ ± 0.28‰, d-excess = 8.2 ± 1.9)
compared to Verdun groundwater sampled between 1997 and 2003.
On a wider scale, the isotopic fractionation with the altitude is around í0.28 į18O per mil for
100 m of elevation [2]. Along our transect, as the mean values of the groundwater around Bordeaux
are į18O = í5.6‰, at the highest point, Seuil de Naurouze (189 m), the altitude isotopic should at
least be í6.2‰, but as the mean Lauragais hill is around 290 m, a correct value would be í6.5‰.
Another way is to calculate the continental effect with the ratio of í3.2 į18O per mil for 1000 km
inland [5]. From the Atlantic coast at the level of Bordeaux until Seuil de Naurouze, there is 330 km,
which should give a fractionation of í1.06, i.e., around í6.66‰, which is in better accordance with
the measured values. If one calculates the continental fractionation until Raissac, 420 km from
the Atlantic coast, the new į18O would become í6.94‰, a value more depleted than the observed
mean value of í6.83‰.
Rainfall coming from the western Mediterranean basin is known to present less depleted į18O
values, but with a higher d-excess value, around +14 [3,18]. Such an influence is seen on groundwater
į18O values of la Franqui and perhaps also in the higher d-excess of the Raissac area. Even in
Toulouse, in October 2012, the abnormal d-excess of the rainfall could come from this Mediterranean
inflow. In general, the Mediterranean rainfall seems to be located much closer to the coast, a few tens
of kilometers wide. Only the black Autan wind, a particular case of the Autan wind blowing northwest,
can bring Mediterranean moisture more inland. However, the amount of this East Mediterranean
rainfall remains low compared to the West Atlantic rainfall. The Autan wind is a continental high
pressure wind bringing heat and dryness and is influenced by the local orography. Even if different
meteorological models try to understand this particular wind [19–21], it is often not announced in
the weather forecast.
Table 2 reports the rainfall isotopic data available in the Global Network of Isotopes in
Precipitation (GNIP) database from IAEA [22]. For Toulouse, the data is from the present work
(year 2012), as there is no other data on the GNIP database. However, it is possible to get the
calculated interpolated data from the area of Toulouse [23], which is į18O = í6.2‰ ± 0.3‰. This
value is a little less than our mean value (į18Omean = í6.6‰ ± 1.8‰), but equal to our weighted mean:
į18Ow-mean = í6.32‰ ± 0.25‰. For the slope between the į18O and įD values, only Toulouse presents
a value close to the GMWL (respectively, 7.70 and 8.00). All the other places on both coasts display
a lower slope, due to a smaller į18O. It is not necessarily a problem of coastal influence or the
evaporation process. In fact, the GMWL as a global line is made of many small local lines with
slopes that are often smaller than eight. There is a good correspondence between the calculated į18O
weighted mean of Bordeaux (in fact, Cestas-Pierroton) and our mean groundwater of this area
(respectively, í5.69‰ and í5.59‰) and also the calculated į18O weighted mean of Dax and our
mean groundwater of this area (respectively, í4.59‰ and í4.86‰). With these į18O weighted
means, the continental effect from west to east between Bordeaux and Toulouse is around one unit
and the latitude effect between Bordeaux and Dax, also one unit of į18O. One can also notice the
equivalent of į18O weight mean values of Dax and Montpellier (respectively, í4.59‰ and í4.17‰),
nearly on the same latitude, but on a different sea coast. The calculated weighted d-excess mean is
always high, i.e., between nine and 12, and only in Avignon, it drops at around eight.
17
The study of the West Mediterranean influence is complex, as it is a transition zone between
the cool North Atlantic air masses and the warm and wet air flowing from the Mediterranean basin.
The isotopic signal of the air masses from these different origins crossing the Mediterranean may be
modified by water vapor produced in the Mediterranean [3]. The d-excess value of the water vapor
of the Mediterranean region results from the intensive evaporation near the coast, under conditions
of a large humidity deficit [24]. All these particular features from the Mediterranean water vapor
reveal the importance of the study to understand the local climate by monitoring numerous and
long-term rainfall and groundwater sampling.
Table 2. Comparison of the isotopic features of the GNIP stations and our Toulouse station.
City
Dax
Bordeaux
Toulouse
Montpellier
Avignon
Slope d-excess
6.09
6.90
7.70
6.49
7.19
0.32
3.90
9.86
3.07
2.31
R2
į18Ow mean
0.948
0.925
0.909
0.825
0.984
(‰)
í4.59 ± 0.95
í5.69 ± 0.65
í6.32 ± 0.25
í4.17 ± 0.14
í5.49 ± 1.20
d-excessw
9.71 ± 1.24
10.18 ± 0.32
11.76 ± 0.59
9.07 ± 2.07
7.39 ± 2.95
į18O range
Rainfall
Year
(‰)
(mm)
í6.84–í2.25 1156 1999–2005
í7.16–í4.33
920 2007–2009
í10.71–í4.00 548
2012
í6.65–í1.79
595 1997–1998
í7.83–í2.23
619 1997–2009
4. Conclusions
The present study is the first one to provided isotopic rainfall data for the Toulouse area in
connection with numerous alluvial groundwater sampling over the Southwest of France. These
results reveal that this area is influenced by both the Atlantic and the Mediterranean climate, but also,
a continental effect is perceptible through the wide isotopic range. The Mediterranean influence is
mainly brought by the Autan wind, but the isotopic composition of the groundwater is only
influenced closer to a few tens of kilometers from the Mediterranean shore. Only some d-excess
higher value can reveal it. More sampling of rainfall at the month scale still needs to be done for this
area to better understand this specific climate. Furthermore, sampling of individual special rain
events, like after the Autan wind, is now also performed for trying to refine these possible
Mediterranean moisture inflows.
Acknowledgments
We would like to thank Florent Barbecot (University of Paris Sud) and Stephan Terzer (IAEA,
Vienna) for their advice for the rainfall isotope sampling, Romain Walker (Ecolab) for Figure 1 and
Daniel Dalger (Ecolab) for help with the isotopic analysis. For the meteorological data, we are
grateful to Pierre Perrin (INRA Auzeville) and Meteofrance (Blagnac airport), and for the two
anonymous reviewers for their valuables comments.
18
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Rozanski, K.; Araguas-Araguas, L.; Gonfiantini, R. Isotopic patterns in modern global
precipitation. In Climate Change in Continental Isotopic Records; Geophysical Monograph
Series Volume 78; Swart, P.K., Lohmann, K.C., Mckenzie, J., Savin, S., Eds.; American
Geophysical Union: Washington, DC, USA, 1993; pp. 1–36.
Celle, H.; Daniel, M.; Mudry, J.; Blavoux, B. Signal pluie et traçage par les isotopes stables en
Méditerranée occidentale: Exemple de la région avignonnaise (Sud-Est de la France) [in
French]. C. R. d’Acad. Sci. Ser. IIA Earth Planet. Sci. 2000, 331, 647–650.
Celle-Jeanton, H.; Travi, Y.; Blavoux, B. Isotopic typology of the precipitation in the Western
Mediterranean region at three different time scales. Geophys. Res. Lett. 2001, 28, 1215–1218.
Ladouche, B.; Aquilina, L.; Dörfliger, N. Chemical and isotopic investigation of rainwater in
Southern France (1996–2002): Potential use as input signal for karst functioning investigation.
J. Hydrol. 2009, 367, 150–164.
Millot, R.; Petelet-Giraud, E.; Guerrot, C.; Négrel, Ph. Multi-isotopic composition (Li, B, D, O)
of rainwaters in France: Origin and spatio-temporal characteristization. Appl. Geochem. 2010,
25, 1510–1524.
Howard, A.D.; Dietrich, W.E.; Seidl, M.A. Modeling fluvial erosion on regional to continental
scales. J. Geophys. Res. Solid Earth 1994, 99, 13971–13986.
Lambs, L. Correlation of conductivity and stable isotope 18O for the assessment of water origin
in river system. Chem. Geol. 2000, 164, 161–170.
Lambs, L. Interactions between groundwater and surface water at river banks and the confluence
of rivers. J. Hydrol. 2004, 288, 312–326.
Craig, H. Isotopic variation in meteoric waters. Science 1961, 133, 1702–1703.
Dansgaard, W. Stable isotopes in precipitation. Tellus 1964, 16, 436–468.
Word Meteorological Organization (WMO). Climatological Normals (CLINO) for the Period
1961–1990; Secretariat of the Word Meteorological Organization: Geneva, Switzerland, 1996.
Stichler, W.; Herrmann, A. Application of Environmental Isotope Techniques in Water Balance
Studies of Small Basins. In New Approaches in Water Balance Computations, Proceedings of
the 18th General Assembly of the International Union of Geodesy and Geophysics, Hamburg,
Germany, August 1983; IAHS Publication 148: Oxfordshire, UK; pp. 93–112.
Geyh, M. Groundwater. In Environmental Isotopes in the Hydrological Cycle; Mook, W.G.,
Ed.; International Hydrological Programme: Paris, France; International Atomic Energy
Agency: Vienna, Austria, 2000; Volume 4, pp. 1–193.
Longinelli, A.; Selmo, E. Isotope geochemistry and the water cycle: A short review with special
emphasis on Italy. Mem. Descr. Carta Geol. d’It. 2010, 90, 153–164.
Bats-Landalle, G. Analyse Géomorphologique de la Plaine d’Inondation de la Garonne [in French].
Ph.D. Thesis, University of Toulouse, Toulouse, France, September 1998.
Weisman, G.S.; Zhang, Y.; LaBolle, E.M.; Fogg, G.E. Dispersion of groundwater age in an
alluvial aquifer system. Water Resour. Res. 2002, 38, 1–13.
19
17. Negrel, Ph.; Petelet-Giraud, E. Isotopes in groundwater: Indicators of climate changes. Trends
Anal. Chem. 2011, 30, 1279–1290.
18. Gat, J.R.; Carmi, I. Effect of Climate Changes on the Precipitation Patterns and Isotopic
Composition of Water in a Climatic Transition Zone: Case of the Eastern Mediterranean Sea
Area. In The Influence of Climate Change and Climatic Variability on the Hydrologic Regime
and Water Resources, Proceedings of the Vancouver Symposium, Vancouver, Canada,
9–22 August 1987; IAHS Publication 168: Oxfordshire, UK; pp. 513–523.
19. Von der Emde, K.; Bougeault, Ph. High resolution simulation of a flow past the Pyrenees.
Meteorol. Atmos. Phys. 1997, 62, 1–8.
20. Benech, B.; Koffi, E.; Druilhet, A.; Durand, P. Dynamic characteristics of regional flows around
the Pyrenees in view of the Pyrex experiment. J. Appl. Meteorol. 1998, 37, 32–51.
21. Aouzerats, B.; Tulet, P.; Pigean, G.; Masson, V.; Gomes, L. High resolution modeling of aerosol
dispersion regimes during the Capitoul field experiment: From regional to local scale
interactions. Atmos. Chem. Phys. Discuss. 2010, 10, 29569–29598.
22. International Atomic Energy Agency Web Page. Global Network of Isotopes in Precipitation.
Available online: http://www-naweb.iaea.org/napc/ih/IHS_resources_gnip.html (accessed on
9 May 2013).
23. Bowen, G.J. The Online Isotopes in Precipitation Calculator, Version 2.2. Available online:
http://www.waterisotopes.org (accessed on 13 May 2013)
24. Gat, J.R.; Klein, B.; Kushnir, Y.; Roether, W.; Wernli, H.; Yam, R.; Shemesh, A. Isotope
composition of air moisture over the Mediterranean Sea: An index of the air-sea interaction
pattern. Tellus 2003, 55, 953–965.
20
Identifying Seasonal Groundwater Recharge Using
Environmental Stable Isotopes
Hsin-Fu Yeh, Hung-I Lin, Cheng-Haw Lee, Kuo-Chin Hsu and Chi-Shin Wu
Abstract: In this study, the stable isotope values of oxygen and hydrogen were used to identify
the seasonal contribution ratios of precipitation to groundwater recharge in the Hualien River basin
of eastern Taiwan. The differences and correlations of isotopes in various water bodies were
examined to evaluate the groundwater recharge sources for the Hualian River basin and the
interrelations between groundwater and surface water. Proportions of recharge sources were
calculated based on the results of the mass balance analysis of the isotope composition of hydrogen
and oxygen in the basin. Mountain river water accounted for 83% and plain rainfall accounted
for 17% of the groundwater recharge in the Huanlian River basin. Using the mean d-values,
a comparison of d-values of precipitation and groundwater indicates the groundwater consists of
75.5% wet seasonal sources and 24.5% dry seasonal sources, representing a distinct seasonal variation
of groundwater recharge in the study area. Comparisons between hydrogen and oxygen isotopes in
rainwater showed that differences in the amount of rainfall resulted in depleted oxygen and hydrogen
isotopes for precipitation in wet seasons as compared to dry seasons. The river water contained more
depleted hydrogen and oxygen isotopes than was the case for precipitation, implying that the river
water mainly came from the upstream catchment. In addition, the hydrogen and oxygen isotopes
in the groundwater slightly deviated from the hydrogen and oxygen isotopic meteoric water line in
Huanlian. Therefore, the groundwater in this basin might be a mixture of river water and precipitation,
resulting in the effect of the river water recharge being greater than that of rainfall infiltration.
Reprinted from Water. Cite as: Yeh, H.-F; Lin, H.-I; Lee, C.-H; Hsu, K.-C; Wu, C.-S. Identifying
Seasonal Groundwater Recharge Using Environmental Stable Isotopes. Water 2014, 6, 2849-2861.
1. Introduction
Oxygen and hydrogen isotopes of water are widely used as tracers to understand hydrogeological
processes such as precipitation, groundwater recharge, groundwater-surface water interactions, and basin
hydrology [1–6]. A comparison of the oxygen and hydrogen isotopic compositions of precipitation
and groundwater provides an excellent tool for evaluating the recharge mechanism [7–11].
Determining the sources of groundwater recharge is important for the effective management of
groundwater resources.
In hydrology, fractionation of į18O and įD is driven by kinetic processes during evaporation and
condensation [12]. In the process of oceanic water evaporation becoming inland rainfall, a sequence of
isotope fractionations causes variations in the composition of the isotope values of oxygen and
hydrogen in continental meteoric water. Since this fractionation process is based on the equilibrium
processes of the isotopes of evaporation and condensation, there is a specific relationship that governs
the distributions of isotope values of oxygen and hydrogen in rainfall.
21
The empirical equation was found by Craig [13] when he used a linear regression method to
analyze the composition of the isotopes of oxygen and hydrogen in samples of precipitation,
snow water, and river water from all over the world. His finding is known as the Global Meteoric
Water Line (GMWL):
įD = 8 į18O + 10
(1)
A later study by the IAEA (International Atomic Energy Agency), water samples from rainfall
stations were collected globally showing a similar result [12,14]:
įD = (8.17 ± 0.08) į18O + (10.56 ± 0.64)
(2)
Most of the precipitation in the world follows this relationship. However, some specific areas that
have different evaporation and condensation conditions (e.g., temperature and humidity), or that have
a unique terrain environment, create their own special local meteoric water line with a different slope
and intercept [12,15]. In the meteoric water line of oxygen and hydrogen isotopes, the slope
represents the ratio of the temperature relationship between įD and į18O when condensation occurs;
the value of the intercept is based on the evaporative conditions in the water source region.
The intercept is also called deuterium excess or d-excess (d = įD í 8į18O) [14]. The intercepts in
most places around the world are about 10‰. However, areas may have different slopes and intercepts
due to different rainfall evaporation conditions or source evaporation conditions in various air
mass sources. For example, North America: įD = 7.95 į18O + 6.03 [12]; Tropical Island area:
įD = 6.17 į18O + 3.97 [12]; Japan: įD = 8 į18O + 17.5 [16]. Generally speaking, if evaporation is faster,
or if rainfall evaporation occurs, intercepts are higher. Some studies have used d-excess to identify
the air mass source of meteoric water and to define the seasonal recharge of groundwater [4,5,7,17].
Taiwan is located in the West Pacific Ocean monsoon climate area. The air mass of the Northeast
monsoon prevalent in winter originates from the Antarctic Continent. The air mass of the Southwest
monsoon prevalent in summer originates from the equator in the Pacific Ocean and from the North
Pacific Ocean tropical marine air mass. Rainfall in Taiwan is related to these three air masses.
The summer rainfall in Taiwan is mostly typhoon cloudbursts and afternoon rainfall caused by
thermal convection. Due to the effects of the Central Mountain Range, Northern Taiwan faces the
Northeast monsoon [18]. During the winter, southward cold fronts bring plentiful water vapor from
the East China Sea which results in rainfall. Although Eastern Taiwan is to the east of the Central
Mountain Range, rainfall is low because the Northeast monsoon moves parallel to the coast. In
addition, Western Taiwan, protected by the Central Mountain Range, is not obviously affected by the
Northeast monsoon, causing winter drought [18].
The purpose of this study is to use oxygen and hydrogen isotopes as natural tracers to identify the
possible sources of groundwater and the seasonal variations in groundwater recharge in the eastern
Taiwan Huanlian River basin. The results provide useful information about hydrological processes,
such as the interaction of precipitation, river water, and groundwater.
22
2. Study Area
The Huanlian River is located in the Huatung Valley in Hualian County, Eastern Taiwan. To
the north, this river connects with the Taroko River system basin; to the west, it is adjacent to
the Chuoshui River; to the south, it is next to the Xiuguluan River basin; and to the east, it is next to
the Fengping River system basin (Figure 1). The Hualian River originates from Bazi Mountain,
which is a sub-range of the Dan Mountain. The main stream is approximately 57.28 km long, with a
basin area of 1507 km2 and an annual runoff of 3813 million m3. The Hualian River basin water
resources are primarily used for agriculture irrigation, followed by domestic and industrial water use.
Despite the rich amount of surface water offered by the Huanlian River, the surface runoff during
wet and dry seasons fluctuates significantly. Each year’s high-flow period starts from May to
October, which accounts for 70% of the runoff of the entire year. After October, the flow declines
significantly, and the driest period occurs in February and March. From November to April of the
following year, the runoff in the drought period accounts for approximately 30% of the annual runoff.
The main stream of the Huanlian River exits the valley from Dafong Mountain and enters the
plains areas. It flows along Huatung Valley from southwest to northeast. The primary sub-ranges
include the Guangfu River, the Maan River, the Wanli River, the Shoufeng River, and the Mugua
River. These rivers run into the ocean near Hualian Mountain, at the north of the Coastal Range. The
hydrogeology of the Hualian River basin can be divided into three areas based on location: the Central
Mountain Range, Huatung Valley, and the Coastal Range. Depending on the terrain, Hualian can be
divided into two sections, Hualian Plain and Huatung Valley Plain. Hualian Plain is located to the
north of Huatung Valley. Good gravel aquifers can be found at depths of 80–90 m underground. The
width of the Huatung Valley’s shallow gravel layer tends to become thinner from the top to the bottom
of the alluvial fan, whereas the deeper layer is characterized by coarse sands and occasional mud
layers. With respect to geological characteristics, Huatung Valley is located at the line of collision
between the Eurasian Plate and the Philippine Sea Plate. The two sides of the valley are delimited by
upthrust with high angles.
Hualian River is abundant in runoff and high in sediment transport capacity. Most surface water
is utilized for agricultural purposes, whereas domestic and industrial water rely on the copious
amount of available groundwater. To understand the use of water resources in this basin, the
fundamental characteristics and recharge sources of the groundwater need to be analyzed. Based on
the effective fractional porosity volume in the alluvium, it can be determined that the groundwater
reserve in the Hualian River basin is approximately 5 billion m3, 370 million m3 of which can be
exploited per year [19]. Based on water balance, the river infiltration recharge for the whole region is
1.45 billion m3 [20]. A report published by the Taiwan Water Resource Agency [21] shows a simulation
of the hydraulic characteristics of regional groundwater using the MODFLOW model in the
Groundwater Modeling System (GMS). According to the results of the groundwater budget in the
Hualian River basin, 27% of the infiltration is direct rainfall, 18% of the recharge is the lateral flow
of boundary, and 55% of the groundwater recharge comes from river water.
23
Figure 1. The location of the study region. Precipitation sampling sites (circles), river
water (squares), and groundwater (triangles) samples are shown.
3. Sampling and Analytical Method
Precipitation, river water, and groundwater samples were collected for oxygen and hydrogen
isotopic analyses from 2003 to 2012. Sampling was carried out during both wet and dry periods.
Sampling procedures for precipitation were in accordance with IAEA guidelines [22]; in short, the
procedures are designed to avoid evaporation of precipitation samples. Sampling locations are shown
in Table 1 and Figure 1. Stable oxygen isotopic compositions were analyzed using the CO2-H2O
equilibration method [23]. The equilibrated CO2 gas was measured using a VG SIRA 10 isotope ratio
mass spectrometer. The hydrogen isotopic compositions were determined on a VG MM602D isotope
ratio mass spectrometer after water was reduced to H2 using zinc shots made by the Biogeochemical
Laboratory of Indiana University [24]. All isotopic ratio results were reported as the į-notation (‰)
relative to the international VSMOW (Vienna Standard Mean Ocean Water) standard. The precisions
(2ı) for į18O and įD were 0.1‰ and 1.5‰, respectively.
24
Table 1. Sampling locations of the study region.
Site
S1
S2
S3
S4
S5
P1
P2
P3
G1
G2
G3
G4
Longitude
121.59
121.52
121.45
121.41
121.39
121.55
121.48
121.48
121.55
121.56
121.45
121.42
Latitude
23.92
23.95
23.81
23.73
23.62
23.87
23.81
23.70
23.94
23.89
23.73
23.66
Elevation (m)
7
96
140
143
160
76
91
106
52
18
110
116
4. Results and Discussion
4.1. Isotopic Compositions of Precipitation
For the purposes of this study, a total of 385 samples of precipitation in the Hualian River basin
were analyzed to discuss the characteristic isotopic signatures of precipitation. The įD of the
precipitation was between í149.5‰ and 22.7‰, with a mean of í22.9‰ ± 31.9‰. The į18O ranged
between í20.2‰ and 1.5‰, with a mean í4.4‰ ± 3.7‰. The mean d was 12.1‰. Linear
regression analysis showed the Local Meteoric Water Line (LMWL) of the Hualian River basin to be
įD = 8.40 į18O + 13.89.
In this study, two local meteoric water regression lines (LMWL) were plotted to describe the
isotopic data for different seasons: įD = 8.03į18O + 9.73 for the wet season precipitation (May to
October) and įD = 8.04į18O + 15.03 for the dry season precipitation (November–April). The slope
and intercept of the regression line for the wet season precipitation are virtually identical to those of
the global meteoric water line (GMWL) of Craig (1961) [15]. The dry season precipitation was found
to have an intercept of 15.03, which is much higher than that of the GMWL of 10 due to the different
air masses affecting the study (see Figure 2).
In this study, the isotopic composition of precipitation during the wet and dry seasons in Hualian
was also examined. The įD in wet seasons ranged between í149.5‰ and 17.5‰, with a mean of
í40.6‰ ± 32.8‰. The į18O was between í20.2‰ and 0.8‰, with a mean of í6.3‰ ± 4.1‰.
The mean of d was 9.5‰. The įD in dry seasons ranged between í69.8‰ and 22.7‰, with a mean of
í3.7‰ ± 16.1‰. The į18O was between í9.6‰ and 1.5‰, with a mean of í2.3‰ ± 1.9‰. The mean
of d was 14.7‰. More depleted composition of hydrogen and oxygen isotopes was found in the
summer wet seasons than the dry seasons. This feature has been commonly observed in other regions
of Taiwan [18,25–27]. In Taiwan, the composition of hydrogen isotopes, compared to the rainfall of
the Northeast monsoon in winters, is depleted in the rainfall of the Southwest monsoon in summers.
As for the effects of temperature, higher temperature may have enriched the signatures of hydrogen
and oxygen isotopes in precipitation. In seasons with great rainfall, the rainfall amount effect [12]
caused by the rain out of heavy precipitation amounts over a relatively short time duration,
25
the hydrogen and oxygen isotopes in precipitation are significantly depleted. Thus, rainfall also has
an effect on isotopes. Temperature and rainfall amount exert opposite effects on the fractionation of
hydrogen and oxygen isotopes. Moreover, the isotopes in the Hualian River basin are depleted more in
wet seasons than is the case in dry seasons. The signatures of hydrogen and oxygen isotopes can
be explained by the mutual influence of rainfall amount and temperature, with the former have
a stronger effect than the latter. During rainy seasons in the summer, rainfall is often heavier with
a greater amount of rainfall during a specific period of time and so a higher precipitation rate, despite
higher temperatures. Consequently, the composition of the hydrogen and oxygen isotopes at this time
is depleted relative to į18O and įD. Yurtsever and Gat (1981) [28] have pointed out generally that
the temperature effect is normally pronounced in high-latitude continental regions, whereas the amount
effect is pronounced in tropical regions. It is well known that the hydrogen and oxygen heavy isotope
contents of precipitation decrease with increasing altitude. In this study, precipitation sites are located
in the lowest area of the basin. Therefore, in this study it is difficult to discuss the range of altitude effect
in the basin. Precipitation sites may not adequately represent the average precipitation in the basin.
If the altitude effect is large, groundwater can also derive from precipitation in the mountainous area
with depleted isotopic compositions similar to streams.
Figure 2. Plot of įD vs. į18O for precipitation samples. LMWL represents the local
meteoric water line.
4.2. Isotopic Compositions of River Water
The įD of the river water in the Hualian River basin was between í70.0‰ and í46.2‰, with a mean
of í61.3‰ ± 5.9‰. The į18O ranged between í10.3‰ and í7.8‰, with a mean of í9.1‰ ± 0.6‰
(Figure 3). The comparison of hydrogen and oxygen isotope compositions between rainwater and river
water demonstrated that the composition of the hydrogen and oxygen isotopes from river water
matched that of the local meteoric water across the Hualian River basin. This indicated that rainfall is
26
the primary source of the river water. Furthermore, the composition of the hydrogen and oxygen
isotopes from the river water was more depleted as compared to that of the precipitation in the valley,
indicating that the river water is composed of rainfall in the upstream catchment. Therefore, the
precipitation in the valley had smaller effects on the river water. In wet seasons, the hydrogen and
oxygen isotopes exhibited a depleted composition compared to that of dry seasons, a similar
phenomenon to the rainwater. Thus, the composition of the rainwater was influenced by the season. In
short, it can be concluded that the water recharge of the Hualian River basin is significantly affected
by seasonal rainfall. Also, as mentioned, the hydrogen and oxygen isotope composition of the rainfall
brought by the Southwest monsoon in the summer is more depleted than that brought by the
Northeast monsoon in the winter.
Figure 3. Plot of įD vs. į18O for river water samples. The LMWL is established as
įD = 8.40 į18O + 13.89 for local precipitation.
-30
Stable isotopic composition of river water
LMWL
Dry season
-40
Wet season
δD (0/00)
-50
-60
-70
-80
-90
-12
-11
-10
-9
-8
-7
-6
δ18O (0/00)
In this study, the signatures of hydrogen and oxygen isotopes for river water in the main stream and
tributaries of the Hualian River (the Mugua River, the Shoufeng River, the Maan River, and the Wanli
River) were also compared during dry and wet seasons. The results are illustrated in Figure 4. As shown
in the figure, significant differences were observed in the hydrogen and oxygen isotope signatures in
the mainstream and tributaries of the Hualian River. In wet seasons, streams leak water into the
groundwater system. Alternatively, water can be discharge from the groundwater to surface waters
in dry seasons. Thus, the groundwater compositions were similar to those of the river water,
indicating that the source of the groundwater in this area may be related to river water. In both dry
and wet seasons, the isotope signatures in the tributaries of the Hualian River were more depleted
than those in the mainstream because the tributaries in mountainous catchment areas consisted of
27
primarily interflow before they entered the mainstream. Interflow is the lateral movement of water in
the vadose zone, that first returns to the surface or enters a stream prior to becoming groundwater. The
interflow blended with groundwater and then joined the Hualian River. Consequently, the
compositions in the two were significantly different. These results may need to be verified by
collecting related data.
Figure 4. Relationship between įD and į18O for river water in comparison to the main
stream and tributaries of the Hualian River during dry and wet seasons. (A) Dry seasons;
(B) Wet seasons.
(A)
(B)
4.3. Isotopic Compositions of Groundwater
This research consisted of an analysis of the groundwater in the Hualian River basin. As illustrated
in Figure 5, the hydrogen and oxygen isotope compositions in the groundwater samples obtained
from observation wells corresponded to those of the rainfall along the Hualian River basin. The įD
of the groundwater in the Hualian River basin was between í73.7‰ and í42.0‰, with a mean of
í56.2‰ ± 9.1‰. The į18O ranged between í10.4‰ and í7.2‰, with a mean of í8.8‰ ± 1.0‰
(Figure 5). The isotopic composition of river water is controlled by the mixing rates of three major
components: surface runoff, interflow, and groundwater (base flow). The difference between arrival
times for interflow and surface runoff is of the order of hours, so they are both from recent storms
and have similar isotopic compositions. Therefore, from the point of view of isotopic composition,
river water can be considered as being composed of groundwater and runoff [29]. The basin
groundwater is recharged from rainfall and river water. In this study, the groundwater compositions
were similar to those of the river water, indicating that the source of the groundwater in this area may
be related to river water. The primary groundwater source may be river water rather than simply
rainfall recharge. In addition, the hydrogen and oxygen isotopes in the groundwater near the Hualian
River slightly deviated from the hydrogen and oxygen isotopic local meteoric water line in Hualian.
28
Therefore, the groundwater might be a mixture of river water and rainwater, which explains why
the effect of the river water recharge was greater than the rainfall infiltration.
Figure 5. Plot of įD vs. į18O of groundwater samples.
-30
Stable isotopic composition of groundwater
LMWL
Dry season
-40
Wet season
δD (0/00)
-50
-60
-70
-80
-90
-14
-12
-10
-8
-6
δ18O (0/00)
4.4. Mass Balance Analysis
The basin groundwater is recharged from rain that falls on the basin and from the tributaries of
the Hualian River (the Mugua River, the Shoufeng River, the Maan River, and the Wanli River)
drained from mountain watersheds (Figure 1). According to the depth of the well screens (80–90 m
below ground) and hydro-geological profiles, groundwater from the four wells (G1, G2, G3, and G4)
can be classified as shallow groundwater. The meteoric į18O-įD signature is important for
understanding the groundwater recharge. The isotopic composition of groundwater equals the average
weighted values of recharge sources, such as the annual composition of precipitation and river water.
Therefore, deviations of the groundwater isotopic ratio from that of precipitation are expected. The
transfer function from precipitation to groundwater must be understood for groundwater provenance
studies. The transfer function also provides basic information about the mechanisms of recharge [2].
In this study, the mean value of oxygen isotopic compositions of groundwater for the Huanlian River
basin was í9.65‰ (ranging from í9.32‰ to í10.65‰). The mean values of oxygen isotopic
compositions of dry and wet seasons for the Huanlian River basin were í4.18‰ and í6.46‰,
respectively. The ratio of precipitation for dry and wet seasons was 0.19:0.81 from 2000 to 2012
(according to the Central Weather Bureau). The weighted average į18O of precipitation was í6.02‰
in the Huanlian River basin. The values of oxygen isotopic compositions of river water for the dry
and wet seasons were í10.20‰ and í10.42‰, respectively. The ratio of stream flow for the dry and
29
wet seasons was 0.14:0.86 from 1980 to 2011 (according to the Water Resources Agency). The river
water weighted average value for į18O was í10.39‰ in the Huanlian River (see Table 2).
Table 2. Precipitation and river water weighted average įD, į18O and d-excess and their
standard error.
Weighted Average:
Ratio
n (used for average)
įD (‰)
į18O (‰)
d-excess (‰)
Precipitation
Dry Season
Wet Season
0.19
0.81
185
200
í3.7 ± 16.1
í40.6 ± 32.8
í2.3 ± 1.9
í6.3 ± 4.1
14.7 ± 5.6
í9.5 ± 4.6
River Water
Dry Season
Wet Season
0.14
0.86
41
64
í62.9 ± 6.5
í66.2 ± 10.6
í9.6 ± 0.7
í9.8 ± 1.2
13.5 ± 3.2
12.1 ± 3.5
In basin water budget studies, it is important to assess the proportion of the precipitation and river
water from the mountain that actually recharges the groundwater. The stable isotopic composition of
groundwater is determined by oxygen and hydrogen isotopic compositions and recharge percentages
of concerned sources. Using mass balance analysis for the oxygen and hydrogen isotopic
compositions, the groundwater recharge percentages of every recharge source can be evaluated.
In this study, mixing between two distinct recharge sources can be quantified by a simple linear
algebraic equation:
C(VA + VB ) = AVA + BVB
C=A
VA
VB
+B
= A(1 − X) + BX
VA + VB
VA + VB
(3)
where A is the precipitation stable isotope value of the basin; B is the river water stable isotope value
of the mountain watershed; C is the groundwater stable isotope value of the basin; VA is the amount
of precipitation; VB is the amount of river water; X is the recharge proportion of river water; and
(1–X) is the recharge proportion of precipitation.
Based on stable isotopic characteristics, the results show that 83% of the groundwater in the
Huanlian River basin is derived from river water from the mountain watershed, and 17% is from the
rain that falls on the basin. This indicates that the basin groundwater is mainly recharged from the river
water from the mountain watershed, primarily due to the abundant precipitation in the mountain area.
Using the mean d-value, the relative contributions of the wet and dry seasonal sources to the
groundwater recharge can be calculated using a mass-balance equation:
dgroundwater = X dwet season + (1 í X) ddry season
(4)
where X and (1 í X) are the fractions of wet and dry seasonal sources, respectively. Based on their
d-values, the groundwater sources are composed of an average of approximately 75.8% wet seasonal
sources and 24.2% dry seasonal sources.
30
5. Conclusions
The present study examined the stable isotopic composition of precipitation, river water, and
groundwater in the Hualian River basin. Mountain river water accounted for 83%, and plain rainfall
accounted for 17% of the groundwater recharge in the Huanlian River basin. Comparisons between
hydrogen and oxygen isotopes in precipitation showed that differences in the amount of rainfall
resulted in depleted oxygen and hydrogen isotopes for precipitation in wet seasons as compared to dry
seasons. River water contained more depleted hydrogen and oxygen isotopes than precipitation did,
implying that the river water mainly came from the upstream catchment. Using a mass balance
equation, a comparison of d-values of precipitation and groundwater indicated the groundwater
consists of 75.5% wet seasonal sources and 24.5% dry seasonal sources, representing a distinct
seasonal variation of groundwater recharge in the study area.
Acknowledgments
This study was financially supported by funds from the National Science Council (NSC), Taiwan,
under grant NSC 101-2221-E-006-196-MY2 and the Central Geological Survey of Taiwan under grant
102-5226904000-01-02. We would also like to thank Chung-Ho Wang of the Institute of Earth
Sciences at the Academia Sinica for help with analyzing the water samples. Special thanks go to the
two anonymous reviewers and the editors for their critical reviews and helpful comments.
Author Contributions
Hsin-Fu Yeh conceived the subject of the article, literature review and contributed to the writing
of the paper; Hung-I Lin participated in the composition of the manuscript in the method, results and
conclusion sections; Cheng-Haw Lee and Kuo-Chin Hsu provided expertise on groundwater-surface
water interactions and hydrogeology in study area; Chi-Shin Wu participated in data processing,
elaborated the statistical analysis, and figures.
Conflicts of Interest
The authors declare no conflict of interest.
References
1.
2.
3.
Gat, J.R. Oxygen and hydrogen isotopes in the hydrologic cycle. Ann. Rev. Earth Planet Sci.
1996, 24, 225–262.
Clark, I.D.; Fritz, P. Environmental Isotopes in Hydrology; Lewis Publishers: New York, NY,
USA, 1997.
Vandenschrick, G.; van Wesemael, B.; Frot, E.; Pulido-Bosch, A.; Molina, L.; Stiévenard, M.;
Souchez, R. Using stable isotope analysis (įD and į18O) to characterise the regional hydrology
of the Sierra de Gador, south east Spain. J. Hydrol. 2002, 265, 43–55.
31
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Deshpande, R.D.; Bhattacharya, S.K.; Jani, R.A.; Gupta, S.K. Distribution of oxygen and
hydrogen isotopes in shallow groundwaters from Southern India: Influence of a dual monsoon
system. J. Hydrol. 2003, 271, 226–239.
Gibson, J.J.; Edwards, T.W.D.; Birks, S.J.; St Amour, N.A.; Buhay, W.M.; McEachern, P.;
Wolfe, B.B.; Peters, D.L. Progress in isotope tracer hydrology in Canada. Hydrol. Process.
2005, 19, 303–327.
Gammons, C.H.; Poulson, S.R.; Pellicori, D.A.; Reed, P.J.; Roesler, A.J.; Petrescu, E.M.
The hydrogen and oxygen isotopic composition of precipitation, evaporated mine water, and
river water in Montana, USA. J. Hydrol. 2006, 328, 319–330.
Blasch, K.W.; Bryson, J.R. Distinguishing sources of ground water recharge by using į2H and
į18O. Ground Water 2007, 45, 294–308.
Li, F.; Pan, G.; Tang, C.; Zhang, Q.; Yu, J. Recharge source and hydrogeochemical evolution
of shallow groundwater in a complex alluvial fan system, southwest of North China Plain.
Environ. Geol. 2008, 55, 1109–1122.
Heilweil, V.M.; Solomon, D.K.; Gingerich, S.B.; Verstraeten, I.M. Oxygen, hydrogen, and
helium isotopes for investigating groundwater systems of the Cape Verde Islands, West Africa.
Hydrogeol. J. 2009, 17, 1157–1174.
Yin, L.; Hou, G.; Su, X.; Wang, D.; Dong, J.; Hao, Y.; Wang, X. Isotopes (įD and į18O) in
precipitation, groundwater and surface water in the Ordos Plateau, China: Implications with
respect to groundwater recharge and circulation. Hydrogeol. J. 2011, 19, 429–443.
Singh, M.; Kumar, S.; Kumar, B.; Singh, S.; Singh, I.B. Investigation on the hydrodynamics of
Ganga Alluvial Plain using environmental isotopes: A case study of the Gomati River Basin,
northern India. Hydrogeol. J. 2013, 21, 687–700.
Dansgaard, W. Stable isotopes in precipitation. Tellus 1964, 16, 436–468.
Craig, H. Isotopic variations in meteoric waters. Science 1961, 133, 1702–1703.
Gat, J.R. The isotopes of hydrogen and oxygen in precipitation. In Handbook of Environmental
Isotope Geochemistry; Fritz, P., Fontes, J.C., Eds.; Springer: Berlin, Germany, 1980; pp. 21–47.
Darling, W.G.; Armannsson, H. Stable isotopic aspects of fluid flow in the Krafla, Namafijall
and Theistareykir geothermal systems of northeast Iceland. Chem. Geol. 1989, 76, 197–213.
Sakai, H.; Matsubaya, O. Stable isotopic studies of Japanese geothermal system. Geothermics
1977, 5, 97–124.
Lee, K.S.; Wenner, D.B.; Lee, I. Using H- and O-isotopic data for estimating the relative
contributions of rainy and dry season precipitation to groundwater: Example from Cheju Island,
Korea. J. Hydrol. 1999, 222, 65–74.
Peng, T.R.; Wang, C.H.; Huang, C.C.; Fei, L.Y.; Chen, C.T.A.; Hwong, J.L. Stable isotopic
characteristic of Taiwan’s precipitation: A case study of western Pacific monsoon region.
Earth Planet Sci. Lett. 2010, 289, 357–366.
Water Resource Agency (WRA). Investigation of Water Resource in Hualien; Ministry of
Economic Affairs: Taipei, Taiwan, 1997; p. 265.
Water Resource Agency (WRA). Investigation of Groundwater Resource in Hualien Alluvial
Valley; Ministry of Economic Affairs: Taipei, Taiwan, 2005; p. 395.
32
21. Water Resource Agency (WRA). Conjunctive Utilization of Surface Water with Groundwater
for Hualien Area; Ministry of Economic Affairs: Taipei, Taiwan, 2010; p. 525.
22. International Atomic Energy Agency (IAEA). Guidebook on Nuclear Techniques in Hydrology;
IAEA Technical Reports Series No. 91; IAEA: Vienna, Austria, 1983; p. 439.
23. Epstein, S.; Mayeda, T. Variation of 18O content of waters from natural sources. Geochim.
Cosmochim. Acta 1953, 4, 213–224.
24. Coleman, M.L.; Shepherd, T.J.; Durham, J.J.; Rouse, J.E.; Moore, G.R. Reduction of water with
zinc for hydrogen isotope analysis. Anal. Chem. 1982, 54, 993–995.
25. Wang, C.H.; Peng, T.R. Hydrogen and oxygen isotopic compositions of Taipei precipitation:
1990–1998. West. Pac. Earth Sci. 2001, 1, 429–442.
26. Peng, T.R.; Wang, C.H.; Lai, T.C.; Ho, S.K. Using hydrogen, oxygen, and tritium isotopes to
identify hydrological factors contributing to landslides in a mountainous area, central Taiwan.
Environ. Geol. 2007, 52, 1617–1629.
27. Yeh, H.F.; Lee, C.H.; Hsu, K.C. Oxygen and hydrogen isotopes for the characteristics of
groundwater recharge: A case study from the Chih-Pen Creek basin, Taiwan. Environ. Earth
Sci. 2011, 62, 393–402.
28. Yurtesever, Y.; Gat, J.R. Atmospheric waters. In Stable Isotope Hydrology: Deuterium and
Oxygen-18 in the Water Cycle; IAEA Technical Report Series No. 210; Gat, J.R., Gonfiantini,
R., Eds.; International Atomic Energy Agency: Vienna, Austria, 1981; pp. 103–142.
29. Lu, H.Y.; Peng, T.R.; Liu, T.K.; Wang, C.H.; Huang, C.C. Study of stable isotopes for highly
deformed aquifers in the Hsinchu-Miaoli area, Taiwan. Environ. Geol. 2006, 50, 885–898.
33
Diagnosing Atmospheric Influences on the Interannual
18
O/16O Variations in Western U.S. Precipitation
Nikolaus H. Buenning, Lowell Stott, Lisa Kanner and Kei Yoshimura
Abstract: Many climate proxies in geological archives are dependent on the isotopic content of
precipitation (į18Op), which over sub-annual timescales has been linked to temperature, condensation
height, atmospheric circulation, and post-condensation exchanges in the western U.S. However,
many proxies do not resolve temporal changes finer than interannual-scales. This study explores
causes of the interannual variations in į18Op within the western U.S. Simulations with the
Isotope-incorporated Global Spectral Model (IsoGSM) revealed an amplifying influence of
post-condensation exchanges (i.e., raindrop evaporation and vapor equilibration) on interannual
į18Op variations throughout the western U.S. Mid-latitude and subtropical vapor tagging simulations
showed that the influence of moisture advection on į18Op was relatively strong in the Pacific
Northwest, but weak over the rest of the western U.S. The vapor tags correlated well with interannual
variations in the 18O/16O composition of vapor, an indication that isotopes in vapor trace atmospheric
circulation. However, vertical-tagging simulations revealed a strong influence of condensation height
on į18Op in California. In the interior of the western U.S., a strong temperature effect was found only
after annual mean temperatures were weighted by monthly precipitation totals. These multiple
influences on į18Op complicate interpretations of western U.S. climate proxies that are derived from
isotopes in precipitation.
Reprinted from Water. Cite as: Buenning, N.H.; Stott, L.; Kanner, L.; Yoshimura, K. Diagnosing
Atmospheric Influences on the Interannual 18O/16O Variations in Western U.S. Precipitation. Water
2013, 5, 1116-1140.
1. Introduction
Stable oxygen and hydrogen isotopes (primarily 18O and deuterium) preserved in geological
archives are commonly used as proxies of past climatic changes (e.g., see reviews: [1–4]). Many of
these proxies are derived from meteoric waters and thus record temporal variations in the isotopic
composition of precipitation (hereafter į18Op, where the isotopic composition is defined as:
į = R/RSTANDARD í 1, where R is the heavy to light isotope ratio). It is important for the interpretation
of climate proxies to better understand the causes of temporal variations in į18Op within a given
region. Within the tropics and subtropics, į18Op is typically linked to changes in precipitation rates
(the amount effect) [5–12], whereas measurements from outside of the tropics and subtropics usually
show correlations between į18Op and air temperature (the temperature effect) [5,13–16]. However,
within many middle latitude regions the relationship between į18Op and climatic variables is not
always as clear, which creates difficulties when interpreting climate proxies within these latitudes.
It is likely that the isotopic composition of precipitation that falls in the middle latitudes is subject to
a combination of influences. This study focuses on the western U.S., with the intent to understand
the primary influences on the year-to-year changes in į18Op.
34
Previous studies of the western U.S. have examined į values from ice cores [17,18], tree
cellulose [16,19,20], speleothems [21–23], leaf wax n-alkanes [24], and lacustrine sediment
archives [25–28]. These records provide estimates of į18Op variations on several timescales,
including seasonal, interannual and interdecadal. The variations within each record have been
attributed to changes in near surface air-temperature, precipitation amount, surface latent heat
release, and/or storm trajectories.
On sub-hourly timescales, precipitation į variations have been linked to changes in condensation
height [29] and raindrop re-evaporation [30]. At weekly to seasonal timescales, studies have
suggested that seasonal temperature variations strongly impact į18Op in the western U.S. [31,32],
while several other studies [33–36] have suggested that storm trajectories were the primary influence
on precipitation į values. Recently, Buenning et al. [37] found through model sensitivity experiments
and tagging simulations that the seasonal cycle of į18Op was primarily a result of seasonal changes
in vapor condensation height along the west coast of the U.S. However, one of the remaining issues
of Buenning et al. [37] was whether or not the same mechanism drives interannual variations in
precipitation į values within the region, which would likely be more valuable for the interpretation
of paleoclimate archives.
This study aims to answer the question of which mechanism(s) primarily drive interannual į18Op
variations in the western U.S. by using model simulations. The following sections discuss the model,
the model experiments (Section 2), and validation (Section 3). Sensitivity experiments are assessed
to find which fractionation process has the largest influence on interannual į18Op variations within
the model, and the influences of post-condensation exchanges, condensation height, moisture
advection, and temperature on interannual į18Op variations are discussed and quantified (Section 4).
Finally, the paper concludes with a summary of the results and a discussion of the implications of
the findings (Section 5).
2. Methods
A methodology similar to Buenning et al. [37] is used here to diagnose the cause of the
interannual variations in į18Op. The model used for this study is the Experimental Climate Prediction
Center’s (ECPC) Isotope-incorporated Global Spectral Model (IsoGSM) [38]. The atmosphere in
IsoGSM has 28 vertical layers, and the chosen horizontal resolution of the model is given by
triangular truncation of the spherical harmonic spectrum at wave number 62 (approximately 1.85°
longitude by 1.85° latitude). The model is forced with prescribed sea surface temperatures and
sea-ice conditions from the optimal interpolation weekly data set, downloaded from the ECPC
database [39]. Each IsoGSM simulation runs from 1953 through 2010 with a 10-minute time-step,
and each simulation is spectrally nudged to the same wind and temperature fields (The National
Centers for Environmental Predictions and National Center for Atmospheric Research Reanalysis
version 1 [40]) every six hours. Yoshimura and Kanamitsu [41] describe the specific details of the
spectral nudging technique.
IsoGSM accounts for water isotopologues (H2O, H218O, and HDO) in all three phases within the
simulated atmosphere, and fractionation factors (defined as Į = Rcd = Rg, where subscripts cd and g
refer to condensed phase and gas, respectively) are computed and applied when phase changes occur.
35
Using the temperature-dependent formulations of Majoube et al. [42,43], equilibrium oxygen
isotopic fractionation is applied during vapor condensation in the atmosphere (Įeq–con) and during
both ocean (Įeq–ev) and raindrop (Įeq–rev) evaporation. Kinetic fractionation is also accounted for in
IsoGSM during ocean evaporation (Įk–ev) [44] and vapor deposition onto ice crystals (Įk–con) [45].
Isotopic exchange and both equilibrium and kinetic fractionation during raindrop evaporation is
estimated following Stewart [46]. It is assumed that 45% of falling raindrops equilibrate with
surrounding vapor for convective precipitation and 95% equilibrates for stratiform precipitation.
Interannual į18Op variability is first evaluated in an unperturbed control simulation (CTRL).
The model is then used to determine which fractionation processes contribute to the interannual
į18Op variations. Sensitivity experiments were conducted with IsoGSM by “turning off” isotope
fractionation. Note, these model experiments were conducted previously by Buenning et al. [37].
Each sensitivity experiment (name in all capitals) turned off individual fractionation processes by
setting the isotopic fractionation factors (Į variables defined above) equal to one (a list and
description of each experiment is given in Table 1). Two of the model experiments assess the
influence of equilibrium oxygen isotope fractionation processes [42,43] on interannual į18Op
variance. These processes are equilibrium oxygen isotope fractionation during ocean evaporation
(Įeq–ev, NOFEQ1) and condensation in the atmosphere (Įeq–con, NOFEQ2). These two experiments
removed both the temperature dependence of Įeq and the isotope effect (i.e., preferential evaporation
or rainout). As such, two additional simulations were conducted where only the temperature
dependence is removed by setting the equilibration temperature to a constant value, globally
(CONFEQ1 and CONFEQ2, with Įeq–ev = 1.00980653 and Įeq–con = 1.01162795, respectively).
Simulations that removed kinetic oxygen isotope fractionation were performed, which included
isotope fractionation during ocean evaporation (Įk–ev, NOFKI1) [44] and vapor deposition onto ice
crystals (Įk–con, NOFKI2) [45]. The final fractionation experiment was performed where raindrop
equilibration rates were set equal to zero. This experiment removes all isotope effects associated with
post-condensation exchanges, including vapor-rain equilibration and both equilibrium and kinetic
fractionation during raindrop evaporation (NORNEV) [11,30,46–48] (see Yoshimura et al. [30] for
equations used in IsoGSM’s rain exchange model).
Vertical tagging simulations were performed with IsoGSM, such that tagged vapor mixing ratios
were set equal to “normal” vapor mixing ratios at specific levels (TAGZ). Unlike the TAGLEV
simulation of Buenning et al. [37] (which only tagged two layers), 14 individual tags were assigned
to the 28 individual layers in IsoGSM to account for condensation height (one tag per 2 atmospheric
levels). Tags were removed from the atmosphere if the tagged vapor advects vertical tagging level.
Thus, resulting precipitation of each vertical water tag will quantify the amount of total precipitation
from 14 different vertical levels of the atmosphere.
36
Table 1. Name and description of Isotope-incorporated Global Spectral Model
(IsoGSM) simulations.
Simulation
name
CTRL
NOFEQ1
NOFEQ2
NORNEV
CONFEQ1
CONFEQ2
NOFKI1
NOFKI2
TAGY
TAGZ
Description
Unperturbed control simulation
Equilibrium oxygen isotopic fractionation during ocean water evaporation is turned off
(Įeq–ev = 1)
Equilibrium oxygen isotopic fractionation during condensation is turned off (Įeq–con = 1)
All oxygen isotopic fractionation associated with raindrop evaporation is turned off
Equilibrium oxygen isotopic fractionation during ocean water evaporation is set to constant,
removing the temperature dependence (Įeq–ev = 1.00980653, T = 293 K)
Equilibrium oxygen isotopic fractionation during condensation is set to a constant, removing
the temperature dependence (Įeq–con = 1.01162795, T = 274 K)
Kinetic oxygen isotopic fractionation during ocean water evaporation is turned off (Įk–ev = 1)
Kinetic oxygen isotopic fractionation during vapor deposition onto ice is turned off (Įk–con = 1)
Tagging simulation where tag1 is applied within 20°–40° N and 140°–170° W; Tag 2 is applied
within 40°–60° N and 140°–170° W.
Tagging simulation where 14 separate tags are applied to the 28 vertical ı layers (2 layers per
tag). Tag 1 is applied to the surface layer and layer 2; tag 2 is applied to layers 3 and 4; tag 3 is
applied to layers 5 and 6; and so on to up to layers 27 and 28. This simulation is actually a
combination of 7 simulations, with each model run simulating two of the tags.
To account for interannual variations in moisture advection, a horizontal tagging simulation was
performed with IsoGSM (similar to the TAGLAT simulation of Buenning et al. [37]), where
subtropical moisture was tagged within 20°–40° N and 140°–170° W and middle latitude vapor was
tagged within 40°–60° N and 140°–170° W (TAGY). For the TAGY simulation, tagged vapor mixing
ratios were set to “normal” vapor mixing ratios within each of the two tagging regions at every model
time-step. When vapor from one tagging region is transported by the winds into the other region, it
is immediately removed to avoid vapor having both tags. Outside of the tagging regions, the tagged
vapor is treated the same as normal vapor, except they are not used in the energy calculations
(i.e., the vapor tags are allowed to advect, mix, condense, and rainout). It is important to note that
the tracer method used here differs with the typical method of tagging vapor that evaporates off the
ocean surface. The method used here allows a direct quantification of the amount of precipitation
and vapor that was advected from each tagging region. Because both the TAGY and the CTRL
simulations are nudged to the same reanalysis wind fields, į18O values can be compared to the
fraction of vapor and precipitation with each tag. Correlation coefficients were calculated at each
grid cell in the western U.S., to quantify the co-variability between į18O values and tagged fraction
(for both TAGY and TAGZ).
3. Simulated į18Op Variability
Previous studies have validated IsoGSM’s ability to capture temporal variations on several
timescales. Yoshimura et al. [38] first compared IsoGSM results with measurements from the Global
Network for Isotopes in Precipitation (GNIP). By comparing IsoGSM and other models against long
37
records (>45 years) from GNIP stations, they concluded that an accurate simulation of atmospheric
winds is necessary to reliably model interannual į18Op variations. These results showed how spectral
nudging constrains the simulated wind and temperature fields, which in turn allows for good
model-data agreement. Within central/coastal California, Yoshimura et al. [30] demonstrated the
ability of the regional version of IsoGSM (IsoRSM) to capture sub-hourly isotopic variations.
Berkelhammer et al. [36] further validated IsoGSM on weekly timescales at 6 sites in California over
the span of 2000–2005. Buenning et al. [37] showed how IsoGSM was able to capture the seasonal
wintertime drop in į18Op at 16 different sites in British Columbia, Washington, Oregon, and
California. However, the lack of long continuous records of į18Op in regions outside of Europe [12]
makes it difficult to validate the model on interannual to multi-decadal timescales, especially in the
western U.S.
To address this issue, studies have used tree cellulose į18O values (į18Oc) as a measure of į18Op
variability [16,19,49–51]. This type of comparison typically involves the use of an offline
biogeochemical model calculation that uses atmospheric model outputs to estimate į18Oc.
Berkelhammer and Stott [16] used IsoGSM output as input for a geochemical model put forth by
Roden et al. [52] and found good model agreement with 2 bristlecone pine trees in the Rocky
Mountains of Colorado. They note that the model’s ability to capture the interannual variations
declines going back in time (especially prior to 1950), which was attributed to a decline in the quality
of the reanalysis forcing data due to less observational constraints. The same type of decline in
IsoGSM model/observation agreement was found in southern California by Kanner et al. [53], who
found that the model performed exceptionally well for the 1979–2004 period, but found a decline in
model agreement from 1953 to 1979. They noted that the discrepancy between the two time periods
could be a result of the increased constraints on the reanalysis data after 1979 (i.e., the satellite era).
4. Attributing Interannual Isotopic Variations
Interannual variance is calculated at each grid cell by computing annual means (centered on the
winter wet season: July through June) for the simulated period; a total of 56 annual means per grid
cell (the first model year is not used). Figure 1a shows a map of the interannual variance of į18Op for
the control simulation (CTRL) over the Pacific and western North America. The largest variance
occurs over the Baja Peninsula of Mexico, and the variance decreases northward along the coast and
inland towards the Southwest U.S. This is consistent with the results of Buenning et al. [37], who
showed that observed seasonal variations along the western U.S. coastline were largest for southern
stations and decreased northward. Figure 1b–g shows the spatial distribution of interannual į18Op
variance for the seven model fractionation experiments. All but two of the experiments showed the
same general pattern of variance as the control simulation; removing equilibrium oxygen isotopic
fractionation during condensation (NOFEQ2) and removing post-condensation exchanges
(NORNEV) being the two exceptions. These results show the importance of isotopic rainout and
post-condensation exchanges in the western U.S., and suggest that other isotope processes (e.g., ocean
evaporation, kinetic fractionation, and the temperature-dependence of Įeq) have less influence on
interannual į18Op variability. The variance was almost completely removed when equilibrium
fractionation during condensation was removed (NOFEQ2), as was the case for the seasonal į18Op
38
cycle [37]. The variance difference between the control simulation (CTRL) and the simulation
that removed post-condensation exchanges (NORNEV) is discussed in more detail in the
subsection below.
4.1. Post-Condensation Exchanges
Post-condensation exchanges were removed by setting raindrop equilibration rates equal to zero,
which not only prevented rain from isotopically equilibrating with the surrounding vapor, but it also
prevented raindrops from fractionating during evaporation [30]. The absolute change in the variance
due to post-condensation exchanges can be quantified by taking the difference in į18Op variance
between the CTRL and NORNEV simulations (Figure 2a). In general, the NORNEV experiment
caused į18Op variance to decrease over land, but increase over the ocean southwest of the Baja
Peninsula (Figure 2a). The fractional change to the variance due to post-condensation exchanges can
be quantified by dividing the variance difference (values in Figure 2a) by the variance from the
control simulation (values in Figure 1a), which is shown in Figure 2b. The fraction of the variance
that is removed by the model experiment (NORNEV) is more uniform over land than the absolute
change, with the highest fractional change occurring over Oregon (Figure 2b). In general
post-condensation exchanges contribute about 60% to the total interannual į18Op variance in the
western U.S., reaching as high as 74% over central Oregon.
What is not apparent in Figure 2a,b is that the removal of post-condensation exchanges
(NORNEV) dampened interannual variations that existed in the control simulation. Figure 3 shows
regional average time series for the CTRL and NORNEV simulations for the Pacific Northwest, the
western US interior, and southern/central California. The curves in Figure 3 show that the variations
produced by both simulation are similar and the inclusion of post-condensation exchanges (in CTRL)
amplify the maxima and minima that would have existed without the processes. To demonstrate that
this is not unique to these three regions, the contribution of post-condensation exchanges to
interannual į18Op variations (defined as the control minus experiment) is correlated to each original
time series from each grid cell from the control simulation (Figure 2c). These calculations produce
high positive correlations over most of the western U.S., ranging from 0.6 to 0.9 (Figure 2c). The
highest correlations were found where the influence of post-condensation exchanges was also high
(e.g., over Oregon). These results reveal that post-condensation exchanges contribute about
60%–75% of the į18Op variance (Figure 2b), mostly by amplifying year-to-year maxima and minima
caused by other processes (Figures 2c and 3).
39
Figure 1. (a) Spatial distribution of variance of į18Op for the control simulation; and
(b)–(h) 7 model experiments. Variance is calculated from interannual time series at each
grid cell. Contour intervals are 0.25‰.
40
Figure 2. (a) Variance difference between the control simulation (CTRL) simulation and
the simulation that removed isotope effects from post-condensation exchanges
(NORNEV); (b) The same variance difference divided by the variance of the CTRL
simulation; (c) Correlation between interannual į18Op variations from the CTRL
simulation and the difference in į18Op between the CTRL simulation and NORNEV
simulation. Open purple rectangles in panel (c) show the boxed regions for Figure 3.
Figure 3. Regional average interannual time series of į18Op for the CTRL and NORNEV
simulations. The panels display the (a) Pacific Northwest; (b) the interior of the western
U.S.; (c) and central/southern California.
41
Figure 3. Cont.
4.2. Condensation Height
Buenning et al. [37] found that į18Op seasonality along the western U.S. coast was primarily
driven by seasonal changes to condensation height. To examine if condensation height is also
influencing interannual į18Op variations, vertical tagging simulations were performed with IsoGSM,
such that the 28 atmosphere levels in IsoGSM were divided into 14 tagged-levels, and 14 separate
tags were added to each of the 14 levels. The tagging was done such that tagged vapor was set equal
to normal vapor within each level at every time step. Tagged vapor is allowed to advect and condense
in the predefined tagged level; however, if the vapor is transported vertically across tagged level
boundaries, it is immediately removed. This approach is similar to the TAGLEV simulation conducted
by Buenning et al. [37], with the exception that the atmosphere is divided into 14 tagged-levels
and not 2.
For most areas, the bottom 14 levels of IsoGSM (the bottom 7 tagged-levels) contributed over
95% of the total precipitation during the course of a typical year. As such, results from the bottom
seven tags are presented here. Figure 4 shows the interannual correlation between the fraction of
precipitation with a vertical tag (i.e., ptag/ptotal) and į18Op for each of the 7 tags at each grid cell in
the western U.S. Though the correlation coefficients do not imply causation, they do quantify
the co-variability between į18Op and condensation height. They will also reveal locations where there
are connections between į18Op and condensation height. For the bottom 3 tags (tags 1–3, hereafter)
correlations are high and positive throughout most of California, with the largest correlation
occurring along the coast of southern and central California (Figure 4a–c). Correlations for tags 1–3
are also high and positive in northern Oregon and southern Washington. The correlations become
weaker for the inland western states and northeastern California.
42
Figure 4. (a)–(g) Correlation between interannual į18Op variations and the fraction of
precipitation from each of the vertical tags. Correlations with the bottommost tag is
shown in panel (a) (tag 1); and correlations with higher level tags are shown in panels
(b)–(g). Contour intervals are 0.1. Absolute values of r above 0.263 are significant above
the 95% confidence level.
43
The correlation changes significantly for tag 4 (Figure 4d) with many regions seeing reduced
correlation, except for parts of southern California, where r-values are at their highest. High negative
correlations exist between į18Op and the fraction of precipitation with higher-level tags (hereafter,
tags 5–7) for many regions in the western U.S. (Figure 4e–g). These shifts from positive to negative
correlations suggests that annual mean į18Op tends to be lower (higher) during years when a larger
fraction of the precipitation was derived from higher (lower) in the atmosphere, where vapor į18O
values are typically more (less) negative. The spatial distributions of the correlation coefficients for
the higher-level tags (5–7) are almost mirror images of the lower level tags (1–3), but of opposite
sign (Figure 4). Like tags 1–3, correlations are high (though negative) for almost all of California,
southern Washington, and northern Oregon. Also, correlations between į18Op and the fraction of
precipitation with tags 5–7 are significantly reduced for the inland western states. The correlations
in Figure 4 reveal evidence of an influence of condensation height on interannual į18Op variations
in some, but not all, regions of the western U.S., similar to the isotope seasonality results of
Buenning et al. [37].
To quantify the contribution of condensation height on interannual į18Op variations, multiple
correlation coefficients were calculated for each location, using the fraction of precipitation from
each tag (ptag/ptotal) as the predictor and į18Op as the dependent variable. Based on the fractional
contribution from each tag (discussed below), only tags 2–6 are used for each calculation. Though
this approach is not a direct measurement of the contribution from condensation height, the r2 value
does give a measure of shared variance, which can be used as a first order estimate of the fraction of
the variance attributable to condensation height. Figure 5a shows the multiple correlation
coefficients, suggesting that condensation height contributes roughly 25%–40% to interannual į18Op
variance in central and southern California (r ranging from 0.5 to 0.63). In coastal northern
California, Oregon, and Washington, the contribution is slightly less at about 10%–15% (r ranging
from 0.3 to 0.4). The r-values in the interior states of the western U.S. are not viewed as a contribution
to the interannual variance because of the sign of the single variable correlations in Figure 4 (which
indicates another effect/process is overriding interannual variations in condensation height in these
regions). Performing the same calculations with į18Op values from the NORNEV simulation resulted
in higher r-values (Figure 5b), which is not surprising since the simulation removed processes that
have already been shown to highly influence interannual į18Op variations. These results show that
condensation height is most influential on interannual į18Op variations in southern and central
California with decreasing influence northward into coastal Washington.
To demonstrate why condensation height drives interannual į18Op variations in some regions more
than others, Figure 6 shows the mean fractional contribution of precipitated tag to the total precipitation
(the mean value used in the correlation calculation) for tags 1 through 7. The contribution of
the bottom tag was only 3%–4% of the total precipitation, but the contribution of tag 2 was higher
and above 10% in most grid cells. Tags 3–5 contributed the highest fraction to the total precipitation
throughout the western U.S., ranging from about 20%–40% for each tag. Further up in the modeled
atmosphere, the fractional contribution decreases sharply for tags 6 and 7, which are similar to
tags 2 and 1, respectively.
44
Figure 5. (a) Multiple correlation between tags 2–6 (ptag/ptotal) and į18Op from the CTRL
simulation; and (b) the NORNEV simulation.
Figure 6. (a–g) The contribution of each vertical tag to the total precipitation. Closed
contour intervals are 0.01.
45
Figure 6. Cont.
Figure 7 shows the mean isotopic composition associated with each vertical tag by amount
weighting (based on tagged vapor) the į18O value of vapor (į18Ov) from the control (CTRL)
simulation. Though the mean į18Ov values of tags 1 and 2 are roughly the same, į18Ov decreases
upward, as would be expected due to vertical isotope gradients in į18Ov [54–56]. Figure 8 shows the
difference in į18O values between tags 3 and 5 (i.e., Figure 7e minus Figure 7c), which are two of
three tags that contributed the most to the total precipitation. The difference in the isotopic
composition of vapor between the two tags clearly shows the vertical gradient (at levels that
contribute the most to total precipitation) is strongest along the coastline and decreases inland.
This feature is partially responsible for the strength of the condensation height effect over most of
Washington, Oregon, and California.
Figure 8 does not reconcile why the influence of condensation height is strongest in southern and
central California and decreases northward along the coast. Figure 9 displays the interannual variance
of the fraction of precipitation with tags 3 and 5. The spatial distribution of the variance is remarkably
similar to Figure 1f (the variance of į18Op from the NORNEV simulation), showing that the largest
variability was near the Baja Peninsula with decreased variance northward along the coast and inland.
Not surprisingly, Figure 9 reveals that the influence of condensation height on į18Op was strongest
in regions where variations in condensation height were also large.
46
Figure 7. (a–g) The isotopic composition of vapor (į18Ov) corresponding with each of
the vertical tags. Mean values are calculated by weighting vertical į18Ov from the CTRL
simulation by the vertical distribution of the given tag. The bottommost tag is shown in
panel a (tag 1), and higher level tags are shown in panels (b)–(g). Contour intervals are 1‰.
47
Figure 8. Difference between the isotopic composition of vapor between vertical tags 3
and 5 (Figure 7e and Figure 7c). More negative values indicate a steeper veritcal gradient
in į18O values of vapor (į18Ov). Closed contour intervals are 0.5‰.
Figure 9. Interannual variance of fraction of precipitation from (a) vertical tag 3; and
(b) vertical tag 5.
4.3. Circulation Effects
4.3.1. Circulation Effects and į18Op
The TAGY simulation was designed to determine how interannual į18Op variations are influenced
by shifts in moisture advection from either the subtropics or middle latitudes. In TAGY, vapor was
continually tagged at all 28 levels of the IsoGSM atmosphere within two boxed regions over the
North Pacific: one in the subtropics and one in the middle latitudes. Outside of these two regions, the
vapor tags are subject to the same atmospheric processes as normal water (i.e., advection, mixing,
condensation, and rainout). As with all atmospheric General Circulation Models, IsoGSM simulates
and accounts for these processes. Furthermore, since both the tagging simulation (TAGY) and the
control simulation (CTRL) are nudged to the same wind and temperature fields, the water
isotopologues and water tags are subject to the same atmospheric and oceanic conditions (e.g.,
48
temperature and rainout histories along air mass trajectories and ocean evaporation conditions).
A fraction of the resulting simulated precipitation that falls in the western U.S. will contain the
middle latitude tag and another fraction will contain the subtropical tag, which is compared to the
simulated į18O values below.
As in Subsection 4.2, interannual correlations coefficients were calculated to quantify covariance
between precipitated tags and į18Op. Figure 10a,b show the interannual correlations between į18Op
and the fraction of precipitation with the middle latitude and subtropical tags. For both tags,
correlations are weak across almost all of the western U.S. This result suggests there is little
covariance between interannual į18Op variations and the fraction of precipitation advected from
either the subtropics or the middle latitudes. Thus, any influence of moisture advection would be
difficult to detect from į18Op at interannual timescales. The exceptions to this include Washington
and Oregon, where correlations with subtropical and middle latitude tagged precipitation fractions
exceed 0.4 and í0.4, respectively. These correlations suggest that the Pacific Northwest might be a
region where precipitation į values adequately trace atmospheric circulation at interannual timescales.
To see if the correlations improve without the influence of post-condensation exchanges and
condensation height, individual interannual time series of į18Op from the NORNEV simulation were
regressed against the fraction of precipitation from the bottom 7 vertical layers from the TAGZ
simulations (values in Figure 6):
δregress = ¦ Ai
7
i=1
pi
+B
p
(1)
In Equation (1), p values are simulated precipitation rates, indices i refer to each of the 7 vertical
tags, and the values of Ai and B (‰) are the regression slopes and intercepts (respectively) used
to fit interannually varying į18Op from the NORNEV simulation (įNORNEV). The residual of the
regression (įresidual) will have removed both the influence of post-condensation exchanges and
condensation height:
δresidual = δNORNEV − δregress
(2)
The interannual variations of įresidual were correlated with the fraction of precipitation from the
middle latitude and subtropical regions of the TAGY simulation. The correlations between įresidual
and the fraction of precipitation advected from the middle latitude-tagged region become slightly
more negative (Figure 10c), which was originally the anticipated relationship. Similarly, the
correlations with the fraction of precipitation with the subtropical tag significantly increased and
were positive at almost all western U.S. locations (Figure 10d). For both tags, the correlations with
įresidual are highest in Washington and Oregon (where the correlations already existed). These results
indicate that the influence of circulation changes on į18Op is secondary and only detectable after the
removal of the primary influences of post-condensation exchanges and condensation height in
California and the western U.S. interior.
49
Figure 10. (a) Correlation between interannual į18Op variations and the fraction of
precipitation from middle latitude tags; and (b) subtropical tags; Panels (c) and (d) show
the same correlations, as (a) and (b) (respectively), but with the influence of
post-condensation exchanges and condensation height removed. Contour intervals are
0.1. Absolute values of r above 0.263 are significant above the 95% confidence level.
4.3.2. Circulation Effects and į18OPW
Results from the TAGY simulation revealed that in most regions of the western U.S. the influence
of circulation changes on interannual į18Op variations is small. However, atmospheric circulation
does affect the isotopic composition of vapor. Correlations between the isotopic composition of
precipitable water (column integrated tropospheric vapor), į18OPW, and the fraction of precipitable
water with the middle latitude and subtropical tags are shown in Figure 11. Figure 11a reveals a
strong negative correlation between į18OPW and the mid-latitude tags throughout much of the western
U.S. The strongest negative correlation in Figure 11a is located at the California/Nevada border near
Lake Tahoe. However, the same figure was generated with additional tagging simulations, and it was
found that the location of the strongest negative correlations was dependent on the location of boxed
regions where the vapor tags were being added (figure not shown). Nonetheless, the additional
simulations also revealed that the negative correlation that exists in the western U.S. is robust.
50
Figure 11. (a) Correlation between interannual į18OPW variations and the fraction of
precipitable water from middle latitude tags; and (b) subtropical tags. Contour intervals
are 0.1. Absolute values of r above 0.263 are significant above the 95% confidence level.
Correlations with the subtropical tags are positive throughout most of the western U.S. The largest
correlation exists in coastal locations within Oregon and Washington, which are the same regions
that had positive correlations with į18Op. The correlations are not as high for the southwest U.S. and
southern California. The additional tagging simulations revealed that the locations of the good and
poor correlations in Figure 11b are robust and not dependent on the location of the boxed regions
where vapor is being tagged. The robustness of both the negative correlation in Figure 11a and the
positive correlation in Figure 11b indicates a strong relationship between the oxygen isotopic
composition of tropospheric vapor and atmospheric circulation in the western U.S. However, this
relationship was not seen in precipitation į values in most regions because of the strong influence of
post-condensation exchanges and condensation height. Indeed, the atmospheric circulation effect
was only transferred to precipitation į values in regions where variations in condensation height are
relatively small like the Pacific Northwest (Figure 9).
4.4. Temperature Effect
Though the NORNEV simulation demonstrated that roughly 60% of the interannual į18Op variations
were due to post-condensation exchanges throughout the western U.S., the tagging simulations
discussed in Sections 4.2 and 4.3 failed to describe the remaining variance in the interior states of
the western U.S. Past studies [16,57] have found or suggested that regions of the interior of the U.S.
show a temperature effect (positive correlations between į18Op and local temperatures). This
relationship is due to increases (decreases) in Rayleigh distillation with lower (higher) temperatures
and subsequently lower (higher) į values of vapor upon arrival at a given location. In this subsection,
IsoGSM’s temperature effect relationship will be quantified within the western U.S.
The correlations between į18Op and surface air temperature were calculated at each grid cell within
the western U.S. for the CTRL simulation. Because the precipitation seasonal cycle peaks in the
summer within the interior states, annual means in this subsection are calculated as the normal
January to December average (though this did not change the general results). Correlations are low
51
in the coastal areas and all of California (Figure 12a), as expected since the modeling results
presented above indicated other processes drive interannual į18Op variations. Relatively high
correlations (r > 0.3) were found in an area of New Mexico spanning northward into Montana and
eastern Washington (Figure 12a). The lack of high correlations between local annual mean
temperatures and į18Op reveals a problem for using climate proxies based on į18Op values to
reconstruct annual mean temperatures in the western U.S. However, when the mean temperatures are
weighted by monthly precipitation the correlations significantly increase over all of the interior states
of the western U.S. (Figure 12b); a result that has been found elsewhere using a different model [15].
Indeed, using precipitation-weighted temperatures increased the correlations from about 0.3 to 0.7
for many regions of the interior western U.S. These results reveal that reconstructed annual mean
temperatures in the western U.S. interior could be biased towards wet season temperatures when
using proxies derived from į18Op.
Figure 12. (a,b) Correlation between į18Op and annual mean temperature; and (c,d)
surface level specific humidity (bottom panels). Right panel annual means of temperature
and specific humidity were weighted by monthly precipitation. Absolute values of r
above 0.263 are significant above the 95% confidence level.
To demonstrate that this temperature effect in the western U.S. was a result of rainout as air masses
moved through the interior of the continent, similar correlation maps were made for specific humidity
(q) and precipitation-weighted specific humidity (Figure 12c,d). Though the correlation maps are not
52
identical to the corresponding temperature plots, they both highlight the interior of the western U.S.
as a region where strong correlations exist, and suggest that the simulated interannual į18Op
variations were driven by the extent of Rayleigh distillation from the coastline to the location where
precipitation occurs. Thus, this rainout process gives rise to the “precipitation-weighted temperature
effect”, as simulated by IsoGSM.
5. Conclusions
The modeling results suggest that interannual variations in precipitation į values in the western
U.S. are primarily caused by a combination of mechanisms, though some atmospheric processes are
detectable at interannual timescales in certain regions. In the western coastal areas and almost all of
California, the variations are largely controlled by changes in condensation height. These induced
variations in į18Op that account for about 20%–40% of the interannual variance, but the year-to-year
changes are amplified by post-condensation exchanges, accounting for about 60% of the variance.
Most of the remaining variance in California and coastal Oregon and Washington is attributable to
subtropical versus middle latitude moisture advection changes (i.e., an atmospheric circulation
effect), which is most prominent and detectable in the Pacific Northwest of the United States. Further
north, Field et al. [58] found a similar relationship between precipitation į values and atmospheric
circulation in western Canada. In California, the atmospheric circulation effect can only be detected
after the influences of post-condensation exchanges and condensation height are removed from
individual time series. However, the model results suggest that atmospheric circulation is a primary
influence on vapor į values throughout almost all of the western U.S.
A temperature effect was found for the inland regions of the western U.S., such that years with
anomalously cooler (warmer) temperatures will lead to more (less) isotopic rainout and lower
(higher) į values of vapor and subsequently precipitation. Like the coastal regions, these isotopic
variations in į18Op are amplified by post-condensation exchanges, contributing about 60% to the
interannual variance. The į18Op-temperature correlations found here were only large after annual mean
temperatures were weighted by precipitation amounts, indicating a problem for reconstructing annual
mean temperatures in the western U.S. interior.
The results presented here will help guide the interpretation of climate proxies in the western U.S.
that are influenced by precipitation į values, such as ice-cores [17,18], speleothems [21,22], tree
cellulose [16,19], leaf wax n-alkanes [24,59], and lacustrine archives [25–28]. The model results
suggest that climate proxy į values from California and parts of Washington and Oregon might
reflect variations in condensation height on interannual timescales. However, Washington and
Oregon were also areas where į18Op correlated with subtropical and mid-latitude tags, indicating
covariance between the vertical and north-south tags in the Pacific Northwest (i.e., moisture
transported from the subtropics typically rains out lower in the atmosphere). In the interior of the
western U.S., proxy į values likely trace precipitation-weighted temperatures changes, which may
differ from annual mean temperatures. The TAGY tagging simulation demonstrated that vapor į
values (į18Ov) trace atmospheric circulation over most regions of the western U.S. This particular
finding may be useful in the context of the results of Helliker et al. [60] who showed through
controlled experiments and model calculations that epiphyte Tillandsia usneoides from CAM plants
53
record į values of vapor. Thus, these plants may be useful in detecting shifts in middle latitude storm
tracks [61,62]. Furthermore, the model results suggest that the Pacific Northwest is a region where
į18Op values trace atmospheric circulation, highlighting an area where shifts in storm tracks may be
detectable through measurements of either precipitation or vapor į values. Additionally, past shifts
may be reconstructed through water isotope based proxy records in the Pacific Northwest. However,
it is important to note that the mean circulation variations that are typical at interannual timescales
(e.g., ENSO variability and the position of the North Pacific High), may differ from the type of
circulation changes on longer timescales, such as multi-decadal and inter-glacial variations.
Detecting shifted storm tracks through isotope measurements should be the focus of future work,
either through direct measurements of isotopes in precipitation or vapor or from climate proxy
reconstructions. This is particularly important because poleward shifted middle latitude storm tracks
are features of a warmer climate that almost all global climate models agree upon [61,62]. Detecting
such shifts is vital for regions, like the southwest U.S., where water resources are already stressed
due to an inadequate supply and growing demand. The results presented here suggest that storm track
shifts can be detected through ongoing monitoring of į18Op in the Pacific Northwest. This stresses
the need to continue station observations of the isotopic composition of precipitation at sites in
Oregon and Washington (e.g., the Olympic site from Buenning et al. [37]). Furthermore, ongoing
measurements of isotopes in vapor may serve useful at detecting storm track shifts in the western
US, either from satellite retrievals [63–65] or ground-based remote sensing [66]. The modeling
results presented here demonstrate how moisture advection can influence the isotopes in atmospheric
moisture in the western U.S.; thus, future work should focus on developing isotope-based
methodologies to monitor and detect storm track variability over the western U.S.
Acknowledgments
Funding for this work was provided by the NOAA/CPO Climate Change & Detection Program:
Paleoclimate Studies (grant NA10OAR4310129). This work was also supported by a grant from the
National Science Foundation (award AGS-1049238). We also thank the helpful comments and
suggestions from two anonymous reviewers.
Conflict of Interest
The authors declare no conflict of interest.
References
1.
2.
3.
Lorius, C.; Jouzel, J.; Raynaud, D.; Hansen, J.; Letreut, H. The ice-core record: Climate
sensitivity and future greenhouse warming. Nature 1990, 347, 139–145.
Vimeux, F.; Ginot, P.; Schwikowski, M.; Vuille, M.; Hoffmann, G.; Thompson, L.G.;
Schotterer, U. Climate variability during the last 1000 years inferred from Andean ice cores: A
review of methodology and recent results. Palaeogeogr. Palaeocl. 2009, 281, 229–241.
McDermott, F. Palaeo-climate reconstruction from stable isotope variations in speleothems:
A review. Quat.Sci. Rev. 2004, 23, 901–918.
54
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Sternberg, L.D.L.O. Oxygen stable isotope ratios of tree-ring cellulose: The next phase of
understanding. New Phytol. 2009, 181, 553–562.
Dansgaard, W. Stable isotopes in precipitation. Tellus 1964, 16, 436–468.
Araguas-Araguas, L.; Froehlich, K.; Rozanski, K. Stable isotope composition of precipitation
over southeast Asia. J. Geophys. Res. Atmos. 1998, 103, 28721–28742.
Yamanaka, T.; Shimada, J.; Hamada, Y.; Tanaka, T.; Yang, Y.H.; Zhang, W.J.; Hu, C.S.
Hydrogen and oxygen isotopes in precipitation in the northern part of the North China Plain:
Climatology and inter-storm variability. Hydrol. Process. 2004, 18, 2211–2222.
Risi, C.; Bony, S.; Vimeux, F. Influence of convective processes on the isotopic composition
(į18O and įD) of precipitation and water vapor in the tropics: 2. Physical interpretation of the
amount effect. J. Geophys. Res. Atmos. 2008, 113, D19306.
Lee, J.E.; Fung, I. “Amount effect” of water isotopes and quantitative analysis of
post-condensation processes. Hydrol. Process. 2008, 22, 1–8.
Lee, J.E.; Risi, C.; Fung, I.; Worden, J.; Scheepmaker, R.A.; Lintner, B.; Frankenberg, C. Asian
monsoon hydrometeorology from TES and SCIAMACHY water vapor isotope measurements
and LMDZ simulations: Implications for speleothem climate record interpretation. J. Geophys.
Res. Atmos. 2012, 117, D15112.
Field, R.D.; Jones, D.B.A.; Brown, D.P. Effects of postcondensation exchange on the isotopic
composition of water in the atmosphere. J. Geophys. Res. Atmos. 2010, 115, D24305.
Buenning, N.H.; Noone, D.C.; Riley, W.J.; Still, C.J.; White, J.W.C. Influences of the
hydrological cycle on observed interannual variations in atmospheric CO18O. J Geophys.
Res. Biogeo. 2011, 116, G04001.
Rozanski, K.; Sonntag, C.; Munnich, K.O. Factors controlling stable isotope composition of
european precipitation. Tellus 1982, 34, 142–150.
Bowen, G.J. Spatial analysis of the intra-annual variation of precipitation isotope ratios and its
climatological corollaries. J. Geophys. Res. Atmos. 2008, 113, D05113.
Sturm, C.; Zhang, Q.; Noone, D. An introduction to stable water isotopes in climate models:
Benefits of forward proxy modelling for paleoclimatology. Clim. Past 2010, 6, 115–129.
Berkelhammer, M.; Stott, L.D. Secular temperature trends for the southern Rocky Mountains
over the last five centuries. Geophys. Res. Lett. 2012, 39, L17701.
Naftz, D.L.; Susong, D.D.; Schuster, P.F.; Cecil, L.D.; Dettinger, M.D.; Michel, R.L.; Kendall, C.
Ice core evidence of rapid air temperature increases since 1960 in alpine areas of the Wind River
Range, Wyoming, United States. J. Geophys. Res. Atmos. 2002, 107, ACL 3.
Schuster, P.F.; White, D.E.; Naftz, D.L.; Cecil, L.D. Chronological refinement of an ice core
record at Upper Fremont Glacier in south central North America. J. Geophys. Res. Atmos. 2000,
105, 4657–4666.
Berkelhammer, M.; Stott, L.D. Modeled and observed intra-ring į18O cycles within late
Holocene Bristlecone Pine tree samples. Chem. Geol. 2009, 264, 13–23.
Berkelhammer, M.B.; Stott, L.D. Recent and dramatic changes in Pacific storm trajectories
recorded in į18O from Bristlecone Pine tree ring cellulose. Geochem. Geophys. Geosy. 2008,
9, Q04008.
55
21. Vacco, D.A.; Clark, P.U.; Mix, A.C.; Cheng, H.; Edward, R.L. A speleothem record of Younger
Dryas cooling, Klamath Mountains, Oregon, USA. Quat. Res. 2005, 64, 249–256.
22. Oster, J.L.; Montanez, I.P.; Sharp, W.D.; Cooper, K.M. Late Pleistocene California droughts
during deglaciation and Arctic warming. Earth Planet Sci. Lett. 2009, 288, 434–443.
23. Ersek, V.; Clark, P.U.; Mix, A.C.; Cheng, H.; Edwards, R.L. Holocene winter climate variability
in mid-latitude western North America. Nat. Commun. 2012, 3, doi: 10.1038/ncomms2222.
24. Feakins, S.J.; Sessions, A.L. Controls on the D/H ratios of plant leaf waxes in an arid ecosystem.
Geochim. Cosmochim. Acta 2010, 74, 2128–2141.
25. Benson, L.V.; Burdett, J.W.; Kashgarian, M.; Lund, S.P.; Phillips, F.M.; Rye, R.O. Climatic and
hydrologic oscillations in the Owens Lake basin and adjacent Sierra Nevada, California. Science
1996, 274, 746–749.
26. Benson, L.V.; Lund, S.P.; Burdett, J.W.; Kashgarian, M.; Rose, T.P.; Smoot, J.P.; Schwartz, M.
Correlation of late-Pleistocene lake-level oscillations in Mono Lake, California, with North
Atlantic climate events. Quat. Res. 1998, 49, 1–10.
27. Li, H.C.; Ku, T.L. Delta C-13-delta O-18 covariance as a paleohydrological indicator for
closed-basin lakes. Palaeogeogr. Palaeoclimatol. 1997, 133, 69–80.
28. Li, H.C.; Ku, T.L.; Stott, L.D.; Anderson, R.F. Stable isotope studies on Mono Lake
(California). 1. Delta O-18 in lake sediments as proxy for climatic change during the last
150 years. Limnol. Oceanogr. 1997, 42, 230–238.
29. Coplen, T.B.; Neiman, P.J.; White, A.B.; Landwehr, J.M.; Ralph, F.M.; Dettinger, M.D.
Extreme changes in stable hydrogen isotopes and precipitation characteristics in a landfalling
Pacific storm. Geophys. Res. Lett. 2008, 35, L21808.
30. Yoshimura, K.; Kanamitsu, M.; Dettinger, M. Regional downscaling for stable water isotopes:
A case study of an atmospheric river event. J. Geophys. Res. Atmos. 2010, 115, D18114.
31. Vachon, R.W.; Welker, J.M.; White, J.W.C.; Vaughn, B.H. Monthly precipitation isoscapes
į18O of the United States: Connections with surface temperatures, moisture source conditions,
and air mass trajectories. J. Geophys. Res. Atmos. 2010, 115, D21126.
32. Ersek, V.; Mix, A.C.; Clark, P.U. Variations of delta O-18 in rainwater from southwestern
Oregon. J. Geophys. Res. Atmos. 2010, 115, D09109.
33. Friedman, I.; Harris, J.M.; Smith, G.I.; Johnson, C.A. Stable isotope composition of waters in
the Great Basin, United States—1. Air-mass trajectories. J. Geophys. Res. Atmos. 2002, 107,
doi: 10.1029/2001JD000565.
34. Friedman, I.; Smith, G.I.; Gleason, J.D.; Warden, A.; Harris, J.M. Stable Isotope Composition
of Waters in Southeastern California. 1. Modern Precipitation. J. Geophys. Res. Atmos. 1992,
97, 5795–5812.
35. Friedman, I.; Smith, G.I.; Johnson, C.A.; Moscati, R.J. Stable isotope compositions of waters in
the Great Basin, United States—2. Modern precipitation. J. Geophys. Res. Atmos. 2002, 107,
doi: 10.1029/2001JD000566.
36. Berkelhammer, M.; Stott, L.; Yoshimura, K.; Johnson, K.; Sinha, A. Synoptic and mesoscale
controls on the isotopic composition of precipitation in the western United States. Clim. Dynam.
2012, 38, 433–454.
56
37. Buenning, N.H.; Stott, L.; Yoshimura, K.; Berkelhammer, M. The cause of the seasonal
variation in the oxygen isotopic composition of precipitation along the western U.S. coast.
J. Geophys. Res. Atmos. 2012, 117, D18114.
38. Yoshimura, K.; Kanamitsu, M.; Noone, D.; Oki, T. Historical isotope simulation using
Reanalysis atmospheric data. J. Geophys. Res. Atmos. 2008, 113, D19108.
39. Reynolds, R.W.; Smith, T.M. Improved global sea-surface temperature analyses using optimum
interpolation. J. Climate 1994, 7, 929–948.
40. Kalnay, E.; Kanamitsu, M.; Kistler, R.; Collins, W.; Deaven, D.; Gandin, L.; Iredell, M.;
Saha, S.; White, G.; Woollen, J.; et al. The NCEP/NCAR 40-year reanalysis project. Bull. Amer.
Meteorol. Soc. 1996, 77, 437–471.
41. Yoshimura, K.; Kanamitsu, M. Dynamical global downscaling of global reanalysis. Mon.
Weather Rev. 2008, 136, 2983–2998.
42. Majoube, M. Fractionation in O-18 between ice and water vapor. J. Chim. Phys. Phys.
Chim. Biol. 1971, 68, 625–636.
43. Majoube, M. Oxygen-18 and deuterium fractionation between water and steam. J. Chim. Phys.
Phys. Chim. Biol. 1971, 68, 1423–1436.
44. Merlivat, L.; Jouzel, J. Global climatic interpretation of the deuterium-oxygen-18 relationship
for precipitation. J. Geophys. Res. 1979, 84, 5029–5033.
45. Jouzel, J.; Merlivat, L. Deuterium and oxygen 18 in precipitation: Modeling of the isotopic
effects during snow formation. J. Geophys. Res. 1984, 89, 11749–11757.
46. Stewart, M.K. Stable isotope fractionation due to evaporation and isotopic-exchange of falling
waterdrops—Applications to atmospheric processes and evaporation of lakes. J. Geophys. Res.
1975, 80, 1133–1146.
47. Bony, S.; Risi, C.; Vimeux, F. Influence of convective processes on the isotopic composition
(į18O and įD) of precipitation and water vapor in the tropics: 1. Radiative-convective
equilibrium and Tropical Ocean-Global Atmosphere-Coupled Ocean-Atmosphere Response
Experiment (TOGA-COARE) simulations. J. Geophys. Res. Atmos. 2008, 113, D19305.
48. Wright, J.S.; Sobel, A.H.; Schmidt, G.A. Influence of condensate evaporation on water vapor
and its stable isotopes in a GCM. Geophys. Res. Lett. 2009, 36, L12804.
49. Feng, X.H.; Reddington, A.L.; Faiia, A.M.; Posmentier, E.S.; Shu, Y.; Xu, X.M. The Changes
in North American atmospheric circulation patterns indicated by wood cellulose. Geology 2007,
35, 163–166.
50. Zhu, M.F.; Stott, L.; Buckley, B.; Yoshimura, K. 20th century seasonal moisture balance
in Southeast Asian montane forests from tree cellulose į18O. Climatic Change 2012, 115,
505–517.
51. Zhu, M.F.; Stott, L.; Buckley, B.; Yoshimura, K.; Ra, K. Indo-Pacific Warm Pool convection
and ENSO since 1867 derived from Cambodian pine tree cellulose oxygen isotopes. J. Geophys.
Res. Atmos. 2012, 117, D11307
52. Roden, J.S.; Lin, G.G.; Ehleringer, J.R. A mechanistic model for interpretation of hydrogen and
oxygen isotope ratios in tree-ring cellulose. Geochim. Cosmochim. Acta 2000, 64, 21–35.
57
53. Kanner, L.; Buenning, N.H.; Stott, L.; Stahle, D. Climatologic and hydrologic influences on the
oxygen isotope ratio of tree cellulose in coastal southern California during the late 20th century.
Geochem. Geophys. Geosy. 2013, Submitted for publiation.
54. Risi, C.; Noone, D.; Worden, J.; Frankenberg, C.; Stiller, G.; Kiefer, M.; Funke, B.; Walker, K.;
Bernath, P.; Schneider, M.; et al. Process-evaluation of tropospheric humidity simulated by
general circulation models using water vapor isotopologues: 1. Comparison between models
and observations. J. Geophys. Res. Atmos. 2012, 117, D05303.
55. Ehhalt, D.H.; Rohrer, F.; Fried, A. Vertical profiles of HDO/H2O in the troposphere. J. Geophys.
Res. Atmos. 2005, 110, D13301.
56. Sayres, D.S.; Pfister, L.; Hanisco, T.F.; Moyer, E.J.; Smith, J.B.; St Clair, J.M.; OތBrien, A.S.;
Witinski, M.F.; Legg, M.; Anderson, J.G. Influence of convection on the water isotopic
composition of the tropical tropopause layer and tropical stratosphere. J. Geophys. Res. Atmos.
2010, 115, D00J20.
57. Vachon, R.W.; White, J.W.C.; Gutmann, E.; Welker, J.M. Amount-weighted annual isotopic
į18O values are affected by the seasonality of precipitation: A sensitivity study. Geophys. Res.
Lett. 2007, 34, L21707.
58. Field, R.D.; Moore, G.W.K.; Holdsworth, G.; Schmidt, G.A. A GCM-based analysis of
circulation controls on į18O in the southwest Yukon, Canada: Implications for climate
reconstructions in the region. Geophys. Res. Lett. 2010, 37, L05706.
59. Romero, I.C.; Feakins, S.J. Spatial gradients in plant leaf wax D/H across a coastal salt marsh
in southern California. Org. Geochem. 2011, 42, 618–629.
60. Helliker, B.R. On the controls of leaf-water oxygen isotope ratios in the atmospheric
crassulacean acid metabolism epiphyte tillandsia usneoides. Plant Physiol. 2011, 155,
2096–2107.
61. Yin, J.H. A consistent poleward shift of the storm tracks in simulations of 21st century climate.
Geophys. Res. Lett. 2005, 32, L18701.
62. Salathe, E.P. Influences of a shift in North Pacific storm tracks on western North American
precipitation under global warming. Geophys. Res. Lett. 2006, 33, L19820.
63. Worden, J.; Bowman, K.; Noone, D.; Beer, R.; Clough, S.; Eldering, A.; Fisher, B.; Goldman, A.;
Gunson, M.; Herman, R.; et al. Tropospheric emission spectrometer observations of the
tropospheric HDO/H2O ratio: Estimation approach and characterization. J. Geophys. Res.
Atmos. 2006, 111, D16309.
64. Worden, J.; Noone, D.; Bowman, K.; Spect, T.E. Importance of rain evaporation and continental
convection in the tropical water cycle. Nature 2007, 445, 528–532.
65. Frankenberg, C.; Yoshimura, K.; Warneke, T.; Aben, I.; Butz, A.; Deutscher, N.; Griffith, D.;
Hase, F.; Notholt, J.; Schneider, M.; et al. Dynamic processes governing lower-tropospheric
HDO/H2O ratios as observed from space and ground. Science 2009, 325, 1374–1377.
66. Wunch, D.; Toon, G.C.; Wennberg, P.O.; Wofsy, S.C.; Stephens, B.B.; Fischer, M.L.; Uchino, O.;
Abshire, J.B.; Bernath, P.; Biraud, S.C.; et al. Calibration of the total carbon column observing
network using aircraft profile data. Atmos. Meas. Tech. 2010, 3, 1351–1362.
58
Effect of Hydrograph Separation on Suspended Sediment
Concentration Predictions in a Forested Headwater with
Thick Soil and Weathered Gneiss Layers
Naoki Kabeya, Akira Shimizu, Jian-Jun Zhang and Tatsuhiko Nobuhiro
Abstract: Two-component hydrograph separation using oxygen-18 concentrations was conducted
at a sediment runoff observation weir installed in a small subcatchment of a forested gneiss catchment
in Japan. The mean soil thickness of this catchment is 7.27 m, which comprises 3.29 m of brown
forest soil (A and B layers) and a 3.98-m layer of heavily weathered gneiss. Data were collected for
a storm on 20–21 May 2003, and the percentage of event water separated by the stable isotope ratio
in comparison with the total rainfall amount was about 1%. This value is within the ratio of a riparian
zone in a drainage area. Temporal variation of suspended sediment concentration exhibited higher
correlation with the event water component than with the total runoff or pre-event water component.
This shows that the riparian zone causes rainwater to flow out quickly during a rain event, and that
this is an important area of sediment production and transportation in a forested headwater with thick
soil and weathered gneiss layers.
Reprinted from Water. Cite as: Kabeya, N.; Shimizu, A.; Zhang, J.-J.; Nobuhiro, T. Effect of
Hydrograph Separation on Suspended Sediment Concentration Predictions in a Forested Headwater
with Thick Soil and Weathered Gneiss Layers. Water 2014, 6, 1671-1684.
1. Introduction
Soil sediment discharge from a forested catchment has been studied for over 60 years [1–3]. These
studies have shown that soil sediment discharge is related to drainage area, rainfall intensity, and
vegetation type and density. Recently, the relationship between suspended sediment (SS)
concentration and runoff amount has been observed in small, forested catchments in various parts of
the world [4–7].
In Japan, attention to research on runoff and SS concentration from forested catchments has
increased remarkably following the Great East Japan Earthquake on 11 March 2011. Radioactive
material, such as Cs, was released because of the accident at the Fukushima Daiichi nuclear
power plant. Although forest ecosystems are thought to exhibit a tendency to retain radioactive Cs,
there is concern that radioactive Cs might flow out from forest ecosystems in steep areas when
subject to considerable rainfall—environments and weather conditions that are common in Japan [8].
Shinomiya et al. [8] observed that radioactive Cs in a small, forested catchment in Fukushima
Prefecture mainly flowed out as suspended matter.
Over the past several decades, research into the rainfall-runoff process for small catchments
has been progressed via the use of tracer information [9–12]. A small, forested headwater is the
beginning of a river and serves as the minimum unit of the water budget and nutrient cycles in a forest
ecosystem [13]. Moreover, as the objective area is relatively small, detailed observations of
geographical features and of the situation generating storm runoff, both in and around a stream
59
channel, are also possible. However, few reports of work exist in which the variations of component
separation by stable isotope ratio and SS concentration have been observed simultaneously by
such research.
Measurement of water and sediment discharge from a headwater catchment is the most elementary
study for soil and water conservation. Williams [14] suggested that there is hysteresis between SS
concentration and runoff from a catchment. This hysteresis has been an obstacle when the temporal
variations of SS concentration and discharge have been modeled simultaneously, and much research
relevant to this topic has been performed [15,16]. To improve the predictive accuracy of models
forecasting SS runoff from a catchment, it is thought necessary to promote the understanding of both
the streamflow mechanism and the SS transport process within a catchment.
Research clarifying the streamflow mechanism from a catchment using a stable isotope tracer is
progressing steadily. For example, by measuring the stable isotope ratio, runoff water and event
rainwater can be separated into two components: pre-event water and event water [17–19]. Many
reports describe the simultaneous observations of dissolved ion concentration and stable isotope ratio
in runoff water [20,21]; however, only a few studies report simultaneous observations of SS
concentration and stable isotope ratio in runoff water.
This research simultaneously observed the temporal change of the stable isotope ratio of stream
water and SS concentration during storm runoff in a small, forested catchment. By performing
hydrograph separation using a stable isotope tracer, the relation between each component of runoff
and SS concentration was investigated. The runoff mechanism and the SS transport process were
considered by comparing these results with information relating to geographical features, especially
the riparian zone. The riparian zone refers to the area relating to the stream bank, and the function of
this zone has been studied recently with respect to both ecological and hydrological aspects [22].
In this research, a riparian zone is defined as the area that is constantly in a state of wet condition
around a spring point and a stream channel between stream banks. Strictly, the spring point and
stream channel are not classified as the riparian zone, but such areas are relatively small and thus,
they are also included within the definition in this paper. On the other hand, the field of hillslope
hydrology provides a related technical term, “source area”. This is defined by the “variable source
area concept” of Kirkby [23], and it means an area within a catchment that contributes to storm runoff.
This paper also follows this definition.
Research is progressing with regard to the function of the riparian zone as a site for the generation
of soil and water movement. The results of this research could be expected to be useful in clarifying
the function of water and soil movement in a riparian zone of a small, headwater catchment with
thick soil and weathered gneiss layers.
2. Experimental
2.1. Study Area
The research was conducted in the Tsukuba Experimental Watershed (TEW), located in southern
Ibaraki Prefecture, Japan (36°20' N, 140°18' E; Figure 1). In 1978, this watershed was established
as an experimental study site for investigations of hydrological processes in a forested mountainous
60
area [24]. The watershed drains 3.79 ha. The main slope direction is to the north and the mean slope
is 25°. In the 10-year period from 1972 to 1981, the average annual air temperature was 14.1 °C at
Kakioka, which is the nearest weather station (36°14' N, 140°12' E; altitude 27.7 m). The annual
rainfall and discharge from 1979 to 1990 (excluding 1988 when data were lacking) were 1429.1 and
641.6 mm, respectively. Although the TEW experiences several snowfalls per year, snow depths
greater than 20 cm are rare and most precipitation is rainfall. Geologically, the watershed is
composed mainly of biotite gneiss overlain by weathered volcanic ash; the soil in this area is brown
forest soil, or Cambisol in the FAO classification [25]. Plantations of Cryptomeria japonica
(common name “sugi” or Japanese cedar) and Chamaecyparis obtusa (“hinoki” or Japanese cypress)
are the main types of vegetation. Pleioblastus chino, a type of bamboo grass, and Aucuba japonica,
an evergreen shrub, both grow on the forest floor. The mean soil thickness is 7.27 m, which comprises
3.29 m of brown forest soil (A and B layers) and a 3.98 m layer of heavily weathered gneiss [26,27].
The stream channel in the TEW originates from three springs (Figure 1) and is one of the
headwaters of the Koise River that flows into Lake Kasumigaura as part of the Tone River system.
To observe the discharge from the three spring subcatchments, 60° V-notch flow-gauging weirs are
operated at the three springs (referred to as A, B, and C). The drainage areas of subcatchments A, B,
and C are 0.60, 0.93, and 0.36 ha, respectively. The discharge amount from the entire TEW is
observed by a 45° V-notch flow-gauging weir at point O (Figure 1). Based on the results of a drilling
investigation, an impermeable wall structure was constructed through the subsurface at point O and
attached to the base rock (6.0 m depth); the weir was then built to measure the amount of watershed
outflow [24]. In contrast, the impermeable walls of weirs A, B, and C are only 1.2 m deep and are
not attached to the rock. In addition to these weirs, a temporary flow-gauging weir was placed at
point S in the TEW to observe flow and to investigate sediment production in the forested catchment;
observations were conducted at this weir for about two years from May 2003, to May 2005. The S
subcatchment drains an area of 2.97 ha. Zhang et al. [28] reported that for the S subcatchment, the
discharge amount of SS was determined mainly by the maximum 10-minute precipitation.
Each rainfall event also greatly contributed to SS discharge. The SS discharge from one storm
accounted for about 5% of the total annual sediment discharge and that from a few storms contributed
about 30% of the yearly SS discharge. In addition, the observation of annual sediment discharge and
the application of a 5-m meshed distributed type sediment discharge model have been conducted
within this catchment [29]. Shimizu et al. [29] showed that the annual sediment discharge of the S
subcatchment was 0.372 t/year. They applied the distributed type sediment yield model to this
catchment and concluded that the area of high sediment production was restricted to the riparian zone
near the stream channel in the catchment.
61
Figure 1. Topography and observation points in the Tsukuba Experimental Watershed.
At spring points A and B, the flat valley bottoms are very narrow (only 0.5 m2), whereas the valley
bottom at spring C is 31.2 m2. The S and O watersheds contain parts of the stream channel (Figure 1),
and the area of the riparian zone of the S subcatchment was calculated by detailed survey as 492.0 m2.
This included the riparian area along the stream channel (stream length (153.4 m) × riparian width
(3.0 m) = 460.3 m2) and the total flat valley bottom area (flat valley bottom of B catchment (0.5 m2) +
flat valley bottom of C catchment (31.2 m2) = 31.7 m2) of the wet zones of the B and C catchments.
Therefore, the percentage of the drainage area occupied by the riparian zone in the S subcatchment
is 1.7%. This riparian zone is including stream channel. In this study, the SS and stable isotope of
the runoff were assessed at the V-notch in the S subcatchment of the Tsukuba Experimental
Watershed (Figure 1). Precipitation was observed by a tipping bucket rain gauge (1 tip = 0.1 mm;
Ikeda Keiki Co., Tokyo, Japan) installed on the roof of the catchment gauging station at O (Figure 1).
In addition, because event water sampling was performed every hour from 17:15 on 17 May 2003,
the precipitation and runoff data were also arranged as hourly values at corresponding times, i.e.,
commencing at 15 min past the hour.
2.2. Storm Runoff Observation and SS Concentration Analysis
The collection of stream water was performed using automatic water sampling equipment (model
6700; Teledyne Isco, Lincoln, NE, USA), and the SS concentration and stable isotope ratios of the
sampling water were analyzed in the laboratory. Water sampling was set up to obtain samples every
62
hour when the rainfall intensity was 1.5 mm/h or more. Thus, the rain event from 20 to 21 May 2003,
was applicable to this research. This 19 h rainfall event produced 52.1 mm of total rainfall with a peak
hourly intensity of 24.6 mm/h (Table 1).
Table 1. Characteristics of the observed rainfall event.
Rainfall characteristic
Total rainfall amount (Ȉ P) (mm)
Start time of the rainfall event
End time of the rainfall event
Duration of rainfall (hours)
A maximum 1-hour rainfall intensity (mm)
The time of a maximum 1hour rainfall intensity
Amount or Time
52.1
16:15 on 20 May 2003
11:15 on 21 May 2003
19
24.6
18:15 on 20 May 2003
After passing a water sample through a 106-ȝm-mesh sieve at the laboratory, 20 cc were isolated
in an airtight, screw-top vial for the stable isotope analysis. Next, it was filtered using the suction
filtration machine equipped with glass filter paper, which was weighed after drying at 105 °C for
3 h. The GF/F filter (d = 0.47 ȝm; Whatman, UK) was used for suction filtration. After filtration,
the glass filter paper was placed in the drier and dried at 80 °C for 48 h, and then weighed with an
electronic balance. The difference in weight before and after drying at 80 °C serves as a measurement
of the amount of SS contained in the sample.
Rainwater was gathered in a 20 L plastic bottle, which was attached to a 21 cm-diameter funnel
installed on the roof of the water level gauging house at O (Figure 1). To prevent the evaporation of
rainwater saved in the bottle and causing a change in the stable isotope ratio, silicone oil was added
to the bottle at the time of commencement of sampling, which made a film on the water’s surface.
Subsequently, the bottle was returned to the laboratory and the oil was removed using a separation
funnel. The rainwater obtained in this way was saved in airtight 20 cc screw-top glass vials.
2.3. Stable Isotope Analysis
A mass spectrometer (MAT252; Thermo Scientific, Waltham, MA, USA) was used for the
oxygen stable isotope analysis of the water samples. The CO2–H2O equilibrium method was used to
measure the hydrogen and oxygen stable isotope ratios. The isotope ratio was expressed as the į
value with respect to that of the Vienna Standard Mean Ocean Water (V-SMOW), which is given as:
·
§ (18 O/ 16 O) sa
δ O sa = ¨¨ 18 16
‰ V - SMOW
− 1¸¸ × 1000 䚷
¹
© ( O/ O) re
18
(1)
where sa and re refer to the sample and standard reference, respectively. V-SMOW is a standard
reference material for measuring stable isotope ratios in water. The standard uncertainties of the į18O
measurements were ±0.02%.
63
2.4. Storm Runoff Hydrograph Separation Using Tracer Information
The runoff component of the stream water is divided into the “event water” (event rainwater)
newly added to the catchment in connection with the rain, and the “pre-event water” (precedent
moisture) already stored in the catchment before the onset of the rain. The contribution of each
runoff component to the rate of total runoff can be calculated using the hydrograph separation
method with a tracer. If it is assumed that a chemical reaction does not arise between the event and
pre-event waters during the observation time, then the following two equations can be formed for
an observed section of a catchment [9]:
Qt = Qevt + Qpre
(2)
Ct Qt = Cevt Qevt + Cpre Qpre
(3)
Here, Q is the discharge; C is the tracer concentration; and subscripts t, evt, and pre express the
total runoff, event water, and pre-event water, respectively. The contribution of the pre-event water
to the total runoff, derived from Equations (2) and (3), is given by the following Equation:
Qpre = [(Ct – Cevt)/(Cpre – Cevt)] Qt
(4)
In Equation (4), Qevt and Qpre are unknowns, whereas Qt could be surveyed as a streamflow rate
and used as the observed runoff from the S subcatchment.
3. Results and Discussion
3.1. Hydrograph Separation by Stable Isotope Ratio
The timing of peak runoff was coincident with the peak of precipitation. Moreover, the peak of
SS concentration was also coincident with the peaks of precipitation and runoff (Figure 2). These
findings were similar to previous researches conducted within this catchment [28,29].
The į18O value of the stream water immediately after the start of the event (17:15 on 20 May 2003)
was í7.94%. Two weeks after the end of the rain event, the streamflow had returned to the baseflow
state. The į18O value of the stream water sampled in the baseflow state was the same as that of
the stream water at the onset of the rainfall event. Therefore, the į18O value of the pre-event water
was assumed fixed at í7.94% during the rainfall event.
Thus, the į18O value of the pre-event water (Cpre) during the event period was set as í7.94%.
The measured value of the isotopic ratio of the rain during the event period was í10.45%. This value
was set as the į18O value of the event water (Cevt). By setting the į18O value of the hourly sampled
stream water to Ct, Equation (3) was used and the runoff at each interval was separated into two
components: event water and pre-event water. Thus, when precipitation and runoff simultaneously
reached their peaks, the contribution rate of the event water was the highest, which was determined
as 60% (Figure 2). The temporal variation of the event water component exhibited a similar pattern
to that of SS concentration. The peak of the pre-event water component was 1 h after the peak of the
event water component and SS concentration.
64
Figure 2. Temporal variations in rainfall, runoff, suspended sediment (SS)
concentration, and oxygen-18 concentration, and the result of hydrograph separation
using oxygen-18 concentration.
The result of hydrograph separation using the stable isotope tracer is shown in Table 2. The ratio
of the runoff component of event water to the total rainfall is 1.0%. This is within the value of 1.7%
for the ratio of the riparian zone to the drainage area, based on a survey result of the S subcatchment
when the riparian zone width was 3.0 m.
65
Table 2. Storm runoff amounts separated by stable isotope tracer information.
Runoff component
Symbol
Total runoff amount
The pre-event component water
The event component water
The ratio of pre-event component water to total runoff amount
The ratio of event component water to total runoff amount
The ratio of total runoff amount to total rainfall amount
The ratio of pre-event component water to total rainfall amount
The ratio of event component water to total rainfall amount
Ȉ Qt
Ȉ Qpre
Ȉ Qevt
Ȉ Qpre/Ȉ Qt
Ȉ Qevt/Ȉ Qt
Ȉ Qt/Ȉ P
Ȉ Qpre/Ȉ P
Ȉ Qevt/Ȉ P
Runoff Amount
( mm)
Percentage
(%)
2.17
1.66
0.50
77
23
4.2
3.2
1.0
The stream and hillslope conditions were checked several times during rainfall events, and no
overland runoff on the hillslope was observed; i.e., the streamflow was flowing only within the
stream channel.
The ratio of event water to total runoff was 23%, and that of pre-event water to total runoff
was 77%. The generation of most of the event component of the water was restricted to a few hours
with strong intensity rainfall. Most of the recession period of the hydrograph comprised the pre-event
component of the water.
3.2. Runoff Generation Mechanism
In the TEW, as in other forested catchments, the infiltration capacity of the soil surface was high
and Horton overland flow was not observed during the rainfall events. However, the topographical
features and soil structure of this catchment are quite different from other catchments. In this
catchment, the watershed has a thick covering of a highly permeable soil layer and a weathered gneiss
layer. However, the valleys around the spring points are steep and narrow, and groundwater exists in
the weathered gneiss layer rather than the soil. For these reasons, changes of the groundwater level
in the source area are small, and a riparian zone can be considered almost the same as the source area.
This differs from other forested catchments where the source area expands more significantly than
the usual riparian zone, for example, where the saturated throughflow dominates runoff generation
in weathered granite catchments [30,31] or in normal vegetated catchments [32]. The spring water is
provided by groundwater flow through the weathered gneiss layer [33]. In subcatchments A and B
that almost do not have a riparian zone, the runoff hardly increased for small-scale rainfall events
of 20 mm or less, and the runoff showed a slow response for large-scale rain events of 100 mm or
more [27]. The hydrographs of subcatchments A and B are similar to the upper weir at CB1 of
Figure 4, in Anderson et al. [34], which represents a steep unchanneled catchment with thick
weathered layers. On the other hand, subcatchments C and S and the entire catchment at O, which
have riparian zone, respond quickly to small-scale rainfall events in the catchment [27]. When a
large-scale rainfall event occurred, the quantity of the baseflow of all subcatchments and the entire
catchment rose significantly and the baseflow took seven months to come back to pre-storm
conditions after a large event [27].
66
This implies that the quick runoff component is related to the size of the riparian zone within
a catchment and the slow runoff component is related to the groundwater flow through the weathered
gneiss layer.
The total rainfall during this study was 52.1 mm, which is a mid-scale rain event for this catchment.
However, almost half of the total rain fell during a 1 h period of peak rainfall. In the hydrograph
separation of the S subcatchment, the runoff peak comprised 60% event water. The runoff of event
water occurred only at the time when rainfall was strong. In this catchment with soils that are deeply
weathered and have generally high infiltration capacities, surface runoff is restricted mainly to the
stream channels, so the storm runoff production must be controlled by subsurface response. Based
on this, it is considered that the runoff component of event water comprised subsurface stormflow
from the riparian zone. Although the source area in a riparian zone is changed for every rainfall event,
it is probably decided by intensity of rainfall and antecedent moisture conditions.
On the other hand, the peak runoff component of pre-event water was 1 h later than the peak of
the event water. Moreover, pre-event water comprised a greater proportion of runoff water at the
time of recession, when the rainfall intensity had weakened. Thus, the runoff component of pre-event
water is mainly formed by groundwater flow.
3.3. Runoff Component and SS Concentration
The SS concentration at each interval was compared with the relation between the total runoff
(Qt), runoff (Qpre) of a pre-event water component, and runoff (Qevt) of an event water component,
and the regression line between each runoff component and the SS concentration was calculated
(Figure 3). The coefficient of determination of the regression line to the SS concentration of Qt was
R2 = 0.6508, the coefficient of determination of the regression line to the SS concentration of Qevt
was R2 = 0.9574, and the coefficient of determination of the regression line to the SS concentration
of Qpre was R2 = 0.1496. Thus, the highest correlation is seen between SS concentration and the
runoff of an event water component. From this, the subsurface stormflow component generated with
event rainwater can be said to play an important role in SS concentration formation during the peak
hour. The width in a stream channel in this catchment was an average of 0.30 m, and this was not
expanded too much even at the time of a heavy rain. Stream bed material was covered with many
stones and soil sediment was stored between the stones. In a heavy rainfall intensity, subsurface
stormflow is generated in the riparian zone, and it flows into the stream channel. As a result, soil
sediment in the stream channel is transported by tractive forces of streamflow.
The event water component is constituted by the subsurface stormflow that occurs in a riparian
zone. At this catchment, the temporal variation of SS concentration depended on the intensity of
rainfall over a short unit of time (10 min) [28], and soil sediment production is only active in the
riparian zone [29].
Thus, the source areas of water runoff and sediment production overlapped in the riparian zone of
this catchment, which is why it is thought that the correlation of temporal variation of subsurface
stormflow and SS concentration increased.
67
Figure 3. Relationship between suspended sediment (SS) concentration and total runoff
(Qt), runoff of pre-event water component (Qpre), and runoff of event water component (Qevt).
Lenzi and Marchi [35] investigated the relation between runoff and SS concentration in
extremely steep mountainous catchments (mean slope: 52°) of the Dolomites in the Italian Alps.
Vegetation cover consisted mainly of herbaceous associations and 14% of the catchment comprised
bare land. They reported that the relation varied, but found a case where the SS concentration peak
appeared after a runoff peak. Following a particle size analysis, they also reported that the origin
of SS was not the stream channel, but the erosion of a slope. Based on their findings and the results
of this research, it can be suggested that the difference in the spatial origin of SS within the
catchment determines the relation between runoff and SS.
4. Conclusions
This research simultaneously observed the temporal change of the stable isotope ratio of stream
water and SS concentration during storm runoff in a small, forested catchment with thick soil and
weathered gneiss layers. By performing hydrograph separation using a stable isotope tracer, the
relation between each component of runoff and SS concentration was investigated. The runoff and
SS concentration during a storm event were shown to peak simultaneously with the maximum
intensity of rainfall, which is similar to the findings of previous researches conducted within this
catchment [28,29]. The percentage of event water to the total rainfall amount, separated by the stable
isotope ratio, was about 1%. This value is within the ratio of a riparian zone within a drainage area.
It is considered that the runoff component of event water comprised subsurface stormflow from the
riparian zone. Although the source area in a riparian zone is changed for every rainfall event, it is
probably decided by intensity of rainfall and antecedent moisture conditions. In a heavy rainfall
intensity, subsurface stormflow is generated in the riparian zone, and it flows into the stream channel.
68
And soil sediment in the stream channel is transported by tractive forces of streamflow. These results
suggest that the riparian zone causes rainwater to flow out quickly during a rain event in a forested
headwater and that it is an important area for sediment production and transportation.
Temporal variation of SS concentration exhibited higher correlation with the event water
component than with total runoff or the pre-event water component. SS discharge correlates well
with event-based runoff, and the runoff mechanisms responsible for bringing event-based water to
the stream channel during storm events are also likely responsible for increasing the sediment load
of the stream channel, thus the runoff mechanism and sources are a critical component to sediment
budgets in headwaters.The SS discharge process, based on the runoff mechanism determined by tracer
information, was examined and it was established as effective to have used the event water extracted
by the stable isotope tracer as a factor of direct SS concentration formation. To reduce the hysteresis
between SS concentration and runoff, it was effective to extract the event water component, as a
direct driving force for transporting SS, using stable isotope tracer information.
The source areas of water runoff and sediment production overlapped in the riparian zone of this
catchment. Such a feature originates in the geographical features and soil structure of this catchment.
Thus, in such a case, it is especially important for sediment management to preserve the riparian zone.
Acknowledgments
We thank Tatsurou Kanazashi, Takashi Yoshitake, Masayuki Sugiyama, and the staff of the Forest
Site Section in the Forestry and Forest Products Research Institute for their help in the field.
We also thank Koji Tamai, the chief of the Forest Hydrology Laboratory in the Forestry and Forest
Research Institute for introducing a suitable cited reference. Part of this work was financially
supported by the Ministry of Agriculture, Forestry and Fisheries, Japan, through a research project
entitled “Development of technologies for mitigation and adaptation to climate change in
Agriculture, Forestry and Fisheries (A-8)”. A part of this work was done at the Lancaster University
under the OECD Co-operative Research Program fellowship 2012. We wish to thank
Nick A Chappell and Wlodek Tych at the Lancaster Environmental Centre of Lancaster University
provided us valuable suggestions.
Author Contributions
The study was design and conceived by Naoki Kabeya and Akira Shimizu. Fieldwork was carried
out in the Tsukuba Experimental Watershed by Naoki Kabeya, Jian-Jun Zhang and Tastuhiko
Nobuhiro under the supervision of Akira Shimizu. Stable isotope analysis was carried out by Naoki
Kabeya with the mass spectrometer of Forestry and Forest Products Research Institute. Hydrological
analysis was carried out by Tatsuhiko Nobuhiro and Naoki Kabeya under the supervision of Akira
Shimizu. Sediment analysis was carried out by Jian-Jun Zhang and Tatsuhiko Nobuhiro under the
supervision of Akira Shimizu. The manuscript was largely written by Naoki Kabeya and Akira
Shimizu but all authors contributed to the writing and review of the manuscript.
69
Conflicts of Interest
The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Anderson, H.W.; Trobitz, H.K. Influence of some watershed variables on a major flood. J. For.
1949, 47, 347–356.
Kawaguchi, T. Suggestion to the elucidation method of influence exerted on the sediment
discharge of a forest. For. Technol. 1952, 124, 9–13. (In Japanese)
Namba, S.; Kawaguchi, T. Influences of some factors upon soil losses from large mountainous
watersheds. Bull. For. For. Prod. Res. Inst. 1965, 173, 93–166 (In Japanese).
Kim, H.; Marutani, T.; Miyazaki, T. Volumetric changes in suspended sediment load and
bed load from mountainous catchments. J. Jpn. Soc. Eros. Control. Eng. 2003, 55, 21–23.
(In Japanese)
Chappell, N.A.; Douglas, I.; Hanapi, J.M.; Tych, W. Sources of suspended sediment within a
tropical catchment recovering from selective logging. Hydrol. Process. 2004, 18, 685–701.
Fukuyama, T.; Onda, Y.; Gomi, T.; Yamamoto, K.; Kondo, N.; Miyata, S; Kosugi, K.;
Mizugaki, S.; Tsubonuma, N. Quantifying the impact of forest management practice on the
runoff of the surface-derived suspended sediment using fallout radionuclides. Hydrol. Process.
2010, 24, 596–607.
Wanga, J.; Edwards, P.J.; Wood, F. Turbidity and suspended-sediment changes from
stream-crossing construction on a forest haul road in West Virginia, USA. Int. J. For. Eng. 2013,
24, 76–90.
Shinomiya, Y.; Tamai, K.; Kobayashi, M.; Ohnuki, Y.; Shimizu, T.; Iida, S.; Nobuhiro, T.;
Sawano, S.; Tsuboyama, Y.; Hiruta, Y. The behavior of radioactive materials in stream water from
a forested watershed. Kanto J. For. Res. 2013, 64, 53–55. (In Japanese)
Pearce, A.J.; Stewart, M.K.; Sklash, M.G. Storm runoff generation in humid headwater
catchments; 1. Where does the water come from? Water Resour. Res. 1986, 22, 1263–1272.
Kendall, C.; McDonnell, J.J. Isotope Tracers in Catchment Hydrology; Elsevier Science:
Amsterdam, The Netherlands, 1998; p. 839.
Soulsby, C.; Malcolm, R.; Helliwell, R.; Ferrier, R.C.; Jenlins, A. Isotope hydrology of the
Allt a’ Mharcaidh catchment, Cairngorms, Scotland: Implications for hydrological pathways and
residence times. Hydrol. Process. 2000, 14, 747–762.
Soulsby, C.; Perty, J.; Brewer, M.J.; Dunn, S.M.; Ott, B.; Malcom, I.A. Identifying and assessing
uncertainty in hydrological pathway: A novel approach to end member mixing in a Scottish
agricultural catchment. J. Hydrol. 2003, 274, 109–128.
Likens, G.E. Biogeochemistry of a Forested Ecosystem, 3rd ed.; Springer-Verlag: New York,
NY, USA, 2013; p. 208.
Williams, G.P. Sediment concentration versus water discharge during single hydrologic events
in rivers. J. Hydrol. 1989, 111, 89–106.
70
15. Kurashige, Y. Mechanism of suspended sediment supply to headwater rivers and its seasonal
variation in West Central Hokkaido, Japan. Jpn. J. Limnol. 1993, 54, 305–315.
16. Eder, A.; Strauss, P.; Krueger, T.; Quinton, J.N. Comparative calculation of suspended sediment
loads with respect to hysteresis effects (in the Petzenknirchen catchment, Austria). J. Hydrol.
2010, 389, 168–176.
17. Kubota, T.; Tsuboyama, Y. Intra-inter-storm oxygen-18 and deuterium variations of rain,
throughfall, and stemflow, and two-component hydrograph separation in a small forested catchment
in Japan. J. For. Res. 2003, 8, 179–190.
18. Gomi, T.; Asano, Y.; Uchida, T.; Onda, Y.; Sidle, R.C.; Miyata, S.; Kosugi, K.; Mizugaki, S.;
Fukuyama, T.; Fukushima, T. Evaluation of storm runoff pathways in steep nested catchments
draining a Japanese cypress forest in central Japan: a geochemical approach. Hydrol. Process.
2010, 24, 550–566.
19. Chaplot, V.; Ribolzi, O. Hydrograph separation to improve understanding of dissolved organic
carbon dynamics in headwater catchments. Hydrol. Process. 2013, doi:10.1002/hyp.10010.
20. Hooper, R.P.; Shoemaker, C.A. A comparison of chemical and isotopic hydrograph separation.
Water Resour. Res. 1986, 22, 1444–1454.
21. Buttle, J.M. Isotope hydrograph separations and rapid delivery of pre-event water from drainage
basins. Prog. Phys. Geogr. 1994, 18, 16–41.
22. Burt, T.; Pinay, G.; Sabater, S. What do we still need to know about the ecohydrology of riparian
zones? Ecohydology 2010, 3, 373–377.
23. Kirkby, M.J. Hillslope Hydrology; Wiley: Chichester, UK, 1978; p. 389.
24. Water Resources Laboratory and Flood Control Laboratory. Statistical reports of hydrological
observation at the Tsukuba Experimental Watershed (May, 1978–December, 1987). Bull. For.
For. Prod. Res. Inst. 1993, 364, 125–168. (In Japanese)
25. Ohnuki, Y.; Yoshinaga, S.; Noguchi, S. Distribution and physical properties of colluvium and
saprolite in unchannelized valleys in Tsukuba Experimental Basin, Japan. J. For. Res. 1999, 4,
207–215.
26. Shimizu, A.; Kabeya, N.; Nobuhiro, T.; Zhang, J.; Kubota, T.; Abe, T. Estimation of subsurface
structure in Tsukuba Forest Experimental Watershed. Kanto J. For. Res. 2007, 58, 153–156.
(In Japanese)
27. Kabeya, N.; Shimizu, A.; Tsuboyama, Y.; Nobuhiro, T.; Zhang, J. Mean residence times of
stream and spring water in a small forested watershed with a thick weathered layer. In Lake
Pollution Research Progress; Miranda, F.R., Bernard, L.M., Eds.; NOVA Science: New York,
NY, USA, 2009; pp. 289–309.
28. Zhang, J.; Shimizu, A.; Kabeya, N.; Nobuhiro, T. Research on suspended sediment of upland
small forest watershed in Japan. J. Beijing For. Univ. 2005, 27, 12–19.
29. Shimizu, A.; Kabeya, N.; Nobuhiro, T.; Tsuboyama, Y.; Abe, T.; Kubota, T. Application of
erosion and sediment yield model with GIS to a small experimental forest watershed. Kanto J.
For. Res. 2006, 57, 299–302. (In Japanese)
71
30. Kabeya, N.; Katsuyama, M.; Kawasaki, M.; Ohte, N.; Sugimoto, A. Estimation of mean
residence times of subsurface waters using seasonal variation in deuterium excess in a small
headwater catchment in Japan. Hydrol. Process. 2007, 21, 308–322.
31. Fujimoto, M.; Ohte, N.; Tani, M. Effects of hillslope topography on hydrological responses in
a weathered granite mountain, Japan: Comparison of the runoff response between the valleyhead and the side slope. Hydrol. Process. 2008, 22, 2581–2594.
32. Steenhuis, T.S.; HrnþíĜ, M.; Poteau, D.; Luna, E.J.R.; Tilahun, S.A.; Caballero, L.A.; Guzman, C.D.;
Stoof, C.R.; Šanda, M.; Yitaferu, B.; et al. A saturated excess runoff pedotransfer function for
vegetated watersheds. Vadose. Zone J. 2013, 12, doi:10.2136/vzj2013.03.0060.
33. Kabeya, N.; Shimizu, A.; Tamai, K.; Iida, S.; Shimizu, T. Transit times of soil water in thick
soil and weathered gneiss layers using deuterium excess modelling. In Conceptual and
Modelling Studies of Integrated Groundwater, Surface Water, and Ecological Systems (IAHS
Publ. 345); Abesser, C., Naütman, G., Hill, M.C., Blöshl, G., Laksmanan, E., Eds.; IAHS Press:
Wallingford, UK, 2011; pp. 163–168.
34. Anderson, S.P.; Dietrich, W.E.; Montgomery, D.R.; Torres, R.; Conrad, M.E.; Loague, K.
Subsurface flow paths in a steep, unchanneled catchment. Water Resour. Res. 1997, 33, 2637–2653.
35. Lenzi, M.A.; Marchi, L. Suspended sediment load during floods in a small stream of the
Dolomites (Northeastern Italy). Catena 2000, 39, 267–282.
72
Water Isotopes as Environmental Tracers for Conceptual
Understanding of Groundwater Flow: An Application for
Fractured Aquifer Systems in the “Scansano-Magliano in
Toscana” Area (Southern Tuscany, Italy)
Marco Doveri and Mario Mussi
Abstract: The “Scansano-Magliano in Toscana” area is characterized by a morpho-structure chiefly
made-up by sandstone and shelly-calcareous lithologies. Generally, these complexes host minor
aquifers in Tuscany, since they have medium to medium-low permeability. In the area under
examination, a sandstone outcrop develops with continuity along the ridge of the structure for several
kilometers and above a shelly substratum. Consequently, this hydrostructural context suggested the
possibility that a significant groundwater body was hosted in the sandstones. In order to verify this
assumption, an isotopic study was carried out taking into account several wells and springs sited on
the sandstone outcrop and its surrounding area; the samples collected over a period of two years were
analyzed to obtain į18O‰, į2H‰ and 3H. A study of the hydrostructural and morphological condition
was also performed, and minor springs were selected. The analyses of this spring-water resulted in
the characterization of the isotopic features of the infiltration water in the studied area, which
represents a fundamental base of work for the interpretation of the data of groundwater points which
drain long flow paths. By means of this approach, the groundwater framework was defined and the
presence of a significant and continuous groundwater body within the sandstone complex was
verified. A preliminary conceptual hydrogeological model was also proposed.
Reprinted from Water. Cite as: Doveri, M.; Mussi, M. Water Isotopes as Environmental Tracers for
Conceptual Understanding of Groundwater Flow: An Application for Fractured Aquifer Systems in
the “Scansano-Magliano in Toscana” Area (Southern Tuscany, Italy). Water 2014, 6, 2255-2277.
1. Introduction
Water isotopes in hydrogeology act as real natural tracers because their signatures are not affected
by water-rock interaction processes up to temperatures of about 200 °C [1]. Their main properties,
such as the variation of values in rainfall both over the different periods of the year and with the
altitude of precipitation [2–4], can consequently be used to define several aspects of the groundwater
flow. By analyzing the ratios 18O/16O, 2H/1H and 3H in springs and/or in water collected from wells,
it is possible to: (i) highlight the existence of different groundwater systems, even when these have
same chemical features; (ii) evaluate the recharge average altitude of groundwater systems; (iii)
achieve information about the hydrodynamic conditions in the aquifer and so on the groundwater
vulnerability; (iv) achieve information about the groundwater residence time.
In fractured and/or karst aquifers, the water isotopes usefulness may be enhanced by the
considerable range of altitude in which these systems normally develop and by the strong heterogeneity
of hydraulic properties that affects isotopic ratios in groundwater, both in space and in time.
As shown by several works of literature (e.g., [5–8]), isotopic applications are a fundamental tool for
groundwater flow understanding in such hydrogeological systems.
73
In this paper, a study mainly based on water isotopes is discussed in order to make observations,
as this kind of survey is able to point out if groundwater framework is characterized by propitious
conditions for water supplying, such as the presence of extensive groundwater bodies. In particular,
the research was carried out on a sandstone aquifer that runs for more than 10 km along the
Scansano-Magliano in Toscana ridge, in southern Tuscany (Figure 1). Few hydrogeological data are
available for this area, because of the lack of previous studies. Nevertheless, the hydrostructural
framework suggests that such fractured aquifer might be a strategic and alternative resource for water
supply, also taking into account the overexploitation and contamination of the nearby alluvial
aquifers. Despite the sandstone is characterized by medium to medium-low permeability and the
wells therein generally have flow rates of a few liters per second, given its extension it can hold a
large volume of water, which might be tapped by a multiple well solution once verified the presence
of a continuity of groundwater circulation. In this context, several springs and wells were sampled in
the Scansano-Magliano in Toscana zone and its surrounding areas, and analyses of į18O‰, į2H‰
and 3H were performed, with the aim to define the groundwater framework and to identify the main
groundwater flow systems, thus verifying whether propitious conditions exist to improve the water
supply of the area.
Figure 1. Geological sketch map and distribution of sampled water points (geological
data from [9], modified).
74
2. Study Area and Methods
2.1. Climate, Geology and Hydrogeology
The Scansano-Magliano in Toscana ridge is located in southern Tuscany between the Albegna
and Ombrone rivers (Figure 1). It develops S-N between 50 and 580 m (a.s.l.) of altitude, starting
from the Albegna plain. The main streamwaters in the studied area are the Patrignone creek and the
Castione ditch, which develop from Scansano toward Magliano in Toscana area, where they flow
into the Albegna River.
Using the data collected by Tuscan Region Administration at the “Scansano raingauge” and
“Manciano thermometric” stations [10], the average annual values of rainfall and temperature of the
area can be assessed as about 880 mm and 14.5 °C, respectively. Based on sixty years of hydrology
data, the maximum value of rainfall is in November (about 125 mm), whereas the minimum is in
July (about 25 mm); the thirteen years of temperature data that are available highlight that the
minimum and maximum values of average temperature occur in January (about 6 °C) and August
(about 23 °C). According to Vittorini (1972) [11] the area belongs to subhumid and subarid
climate classes.
From a geological point of view, the study area fits into an inner portion of the northern Apennine
chain, which is characterized by eastward nappe stacking. In particular, the tectonic evolution is
characterized by early compressional stages, started in the Early-Cretaceous, with the progressive
suture of the Jurassic Ligurian ocean by convergence of Adria and European plates, leading to
continental collision in the Middle Eocene ([12–15] and references therein). The convergence
continued during the post-Oligocene with the eastward overthrusting of the Ligurian Units (derived
from the oceanic domain and its transition to the continental margin) on the detached sedimentary
cover of the continental margin (Tuscan Nappe) [16–18]. From the Middle Miocene, the structural
stacking was affected by extensional processes, related to the development of the Tyrrhenian
Sea [19–21]. Such processes generated Plio-Pleistocene basins, interpreted by many authors as
controlled by longitudinal systems of high angle normal faults, assuming a continuous extensional
regime from the Middle Miocene to the Present [21–23]. The tectonic depressions of the Neogene
Basins have been so affected by a deposition of continental and marine sediments (NeogeneQuaternary age). Moreover, the extensional regime also controlled the emplacement of magmatic
bodies at limited depths and in some case the eruption of volcanic products (e.g., the Mt. Amiata
volcanic rocks; Figure 1).
In this geological framework, the Scansano-Magliano in Toscana zone and its surrounding areas
are characterized by three main units (Figure 1): (1) the youngest one (Neogene-Quaternary age)
mainly consists of alluvial deposits, travertines, volcanic rocks, Pliocenic clays and sandy clays,
and Miocenic conglomerates and sandstones; (2) the Tuscan Nappe Unit, here represented by
sandstones and siltstones (Macigno formation, upper Oligocene–lower Miocene), shales and marls
(Scaglia Toscana formation), calcarenites (Calcareniti di Montegrossi formation), cherty-limestones
and limestones (formations of Calcare Selcifero, Calcare Massiccio), and dolomitic-limestones
(Calcare Cavernoso); (3) the Ligurian Units, consisting of greywackes (Pietraforte Formation,
75
Medium–Upper Cretaceous), shales (Argille a Palombini formation) and of a sequence of clays and
marly-limestones (mainly Argille e Calcari formation).
Based on this geological context, some major hydrogeological units can be defined:
•
•
•
•
•
•
The first unit consists of alluvial deposits, travertine and Miocenic sand, and it is characterized
by medium to high permeability;
The second unit is composed by volcanic rocks, whose permeability is medium-high to high;
The third unit, composed of Pliocene clays, is considered impermeable;
The fourth unit consists of siltstones, shales and of a sequence of clays and marly-limestones
(mainly Ligurian Units and Scaglia Toscana formation), and has medium-low to very
low permeability;
The fifth unit is represented by sandstones (mainly Macigno formation and to a lesser extent
Pietraforte formation). This unit is considered to have medium to medium-low permeability.
The sixth unit, made up by calcarenites, cherty-limestones, limestones and dolomitic-limestones
of the Tuscan Nappe. Its permeability is medium to very high.
In the specific zone of the Scansano-Magliano in Toscana ridge, the most important hydrogeological
complex is represented by the sandstones of the Macigno formation, because it covers with continuity
a surface of about 35 km2 and it is limited, both downward and laterally (eastward and westward),
by medium-low to very low permeability complexes mainly made up by shales and marly-limestones
(“Scaglia Toscana” and “Argille e Calcari” of the Tuscan Nappe and Ligurian Unit, respectively).
Because the sandstone outcrops and is unconfined in this area, a significant groundwater body can
be presumed to reside within this complex, also taking into account the brittle deformation that may
have locally improved the general medium degree of permeability. Indeed, Francese et al. (2009) [24]
highlighted a complex geometry of the fracture network in the Scansano area, which presents
fractures and faults with major NNW-SSE, N-S and E-W trends. Several farm wells already exploit
this resource, which is also drained by some springs with average flowrate of the order of 0.5 L/s.
Only a few springs and wells are present on the marly-limestones outcropping westward and eastward,
showing only minor groundwater flow there.
2.2. Methodology
During the 2004–2005 period, six sampling fields were carried out in order to collect water sample
from 25 springs, 31 wells and 2 stream waters (120 samples in total). Analyses were performed to
achieve the abundance ratios of the water stable isotopes (18O/16O, 2H/1H), which are expressed as
į‰ compared to the V-SMOW standard [25], and the 3H content (expressed as tritium units, TU;
1 TU = 1 3H atom every 1018 total hydrogen atoms). The į18O‰ was analyzed for each collected
sample, whereas į2H‰ values and tritium contents were achieved for 90 and 40 samples, respectively.
The į18O value was determined through analysis of gaseous CO2, previously equilibrated with
water at 25 °C [26]. The mass-spectrometric measurement of the 18O/16O ratio requires correction,
because a fractionation between CO2 and H2O occurs. Analytical precision on į18O‰ values is better
than 0.10‰. The į2H value was measured by reducing the water to elemental hydrogen [27] using
magnesium instead of zinc. Because the totality of water is reduced and all hydrogen is converted to
76
hydrogen gas, the isotopic fractionation does not occur and a correction of the mass-spectrometric
measurement is not necessary. The analytical error for į2H is 1.5‰. The isotopic ratios of CO2 and
H2 were measured by dual inlet mass spectrometer.
Tritium is a short-lived isotope of hydrogen with a half-life of 12.3 years. It is analyzed through
measurement of ȕí decay events in a liquid scintillation counter. Direct liquid-scintillation counting
has a precision of 7 TU. For tritium contents lower than 20 TU, increased precision is gained through
concentration by electrolytic enrichment of 3H in the water before counting, thus reaching a precision
better than 0.8 TU.
As shown in Figure 1, the sampled water points are distributed not only along the Scansano-Magliano
in Toscana ridge but also in its surrounding areas. Such points were selected with the aim to define,
in terms of groundwater flow, the relationships between the sandstone aquifer and the other
hydrogeological complexes, and moreover to obtain a good isotopic characterization of the
rainfall/infiltration water. Indeed, the application of isotopic methods is closely dependent on the
knowledge of some local parameters, such as the distribution of the isotopic composition of
rainwater/infiltration water and vertical isotopic gradient in the area. This basic information is
obtained by examining the į18O‰, į2H‰ and 3H in the rainwater collected every month and at
various altitudes for a period of at least two-three years (e.g., [28]). In the absence of a well
distributed raingauge-network, but also to avoid the long time required by this methodology and to
get information directly on the infiltrated water, it is possible to collect some samples (3–4 each year)
in small springs (low flowrate) that are fed by limited extension basins (e.g., [8,29]). In this way the
stable isotopes of the water are representative of an infiltration average altitudes that is not very
different from the springs’ altitudes and assessable by morphological and hydrogeological
considerations. In order to minimize the error in the assessing of the infiltration average altitudes, it
is preferable to identify springs at the base of small reliefs and in hydrostructural contexts that suggest
a local circulation of groundwater (see the example in Figure 2).
Figure 2. Cross section (A–B) in the zone of the spring “8”.
Based on the latter approach, several springs with low flowrate were identified over the altitude
interval of 40–995 m a.s.l. These springs are representative of a groundwater circulation in aquitard
complexes, or in aquifers complexes that have a limited extension. Despite its high flowrate and wide
recharge area, the main spring draining the Mt. Amiata volcanic aquifer [30] was also included, in
order to achieve isotopic information that are representative of the highest altitudes encountered in
the area surrounding the site under examination. The data derived from all these selected springs
77
allowed fundamental background information to be obtained, e.g., the relationship altitude/į18O‰
(or alternatively altitude/į2H‰), which was then used to assess the average altitude of the feeding
area for the groundwater tapped by wells into the hydrogeological complexes of the ScansanoMagliano in Toscana area.
The tritium values achieved for the water points were also compared with the tritium annual data
of rainfall, which were recorded at the Genova and Pisa stations [29,31,32]. This data can be
considered homogeneous in the whole Mediterranean area and not affected by the precipitation
altitudes [29]. In order to conduct a preliminary assessment of the average residence time of
groundwater flow, the exponential law of radioactive decay was applied to the 3H average annual
values of rainfall occurred in the past years, and the achieved results were compared with the contents
of water points under examination.
3. Results and Discussion
3.1. Isotopic Features of the Infiltration Water
In order to define isotopic features of infiltration water, samples collected at 25 springs, which are
located in the area between Magliano in Toscana and the Mt. Amiata (Figure 1), were analyzed.
Seventeen of such springs were sampled at least two time in different periods, whereas the remaining
springs were collected only once.
The selection of the springs was performed on the base of morphological and hydrostructural
conditions, with the aim to individuate water points which represent local groundwater, whose
average altitudes of feeding are similar to the springs’ altitudes and assessable with a good
approximation (see Figure 2, such as example among the several elaborated hydrogeological
sections). In this way, the relationship between infiltration average altitudes and isotopic values
(in particular that of stable isotopes) can be achieved. The only exception is spring 6, which has an
extended recharge area, whose average altitude was estimated on the base of previous studies [30,33].
Results of isotopic analyses are reported in Table 1. For the springs that were analyzed more than
once, the isotopic composition was observed to be stable over time. Given this general behavior, in
the data processing and interpretation the average isotopic values were considered for such springs;
moreover, also for the springs that were sampled only once, the isotopic values were involved as
representative of the annual average values.
The į2H‰ and į18O‰ values of the springs are compared in Figure 3 with the global meteoric
(GMWL) [34] and the Mediterranean meteoric (MMWL) [35] water lines. The spring water
fall between these two lines and give a regression line equation of į2H‰ = 5.7 × į18O‰ í 2.4
(R2 = 0.97), which can be considered as representative of the local meteoric water line (LMWL). As
previously observed in the southern Tuscany (e.g., [36]), groundwater points dispose between the
GMWL and the MMWL with a slope lesser than 8. Taking also into account that the low and high
altitude springs fall close to GMWL and MMWL, respectively, such behavior likely reflects an
evaporative influence on the isotopic values of rainfall of Mediterranean origin. The wide range of
values observed for į18O and į2H (about 2.5‰ and 15‰, respectively) is nevertheless mainly linked
to the altitude effect.
78
Table 1. Springs isotopic data.
Sampled
Springs
Infiltration į18O‰ (VSMOW)
Spring
Altitudes
(m a.s.l.)
į18O‰ Values
į2H‰ (VSMOW)
į2H‰ Values
Average
(Precision 0.10‰)
(March 04/June–July 04/
(Precision 1.5‰)
(March 04/June–July 04/
Altitudes *
Mean Value or
September 04/May–June 05/
Mean Value or
September 04/May–June 05/
(m a.s.l.)
Single Datum
September 05/October 05)
Single Datum
September 05/October 05)
3
3
H(TU) ± Err
Mean Value or
Single Datum
H(TU) Values
(March 04/June–July 04/
September 04/May–June 05/
September 05/October 05)
1
75
150
í5.58
(í5.53/í5.54/í5.63/ns/í5.62/ns)
í34.2
(í33.6/í33.2/í35.8/ns/na/ns)
2.9 ± 0.7
(na/3.0/2.8/ns/na/ns)
2
260
310
í6.24
(í6.28/í6.18/ns/í6.25/ns/ns)
í38.5
(í38.1/í38.9/ns/na/ns/ns)
4.7 ± 0.7
(na/4.7/ns/na/ns/ns)
3
490
530
í6.68
(í6.59/í6.71/í6.73/í6.67/ns/ns)
í39.4
(í38.3/í40.4/í39.4/na/ns/ns)
na
(na/na/ns/na/ns/ns)
4
380
500
í6.64
(í6.62/í6.65/ns/ns/ns/ns)
í40.3
(í39.4/í41.2/ns/ns/ns/ns)
na
(na/na/ns/ns/ns/ns)
4bis
395
500
í6.74
(í6.79/í6.68/ns/ns/ns/ns)
í40.2
(í38.6/í41.7/ns/ns/ns/ns)
na
(na/na/ns/ns/ns/ns)
5
175
200
í5.57
(í5.54/í5.56/í5.55/í5.61/ns/ns)
í34.1
(í34.4/í34.6/í32.3/na/ns/ns)
4.2 ± 0.6
(na/na/4.2/na/ns/ns)
6
650
1300
í7.89
(í8.00/í7.83/í7.86/í7.86/ns/ns)
í48.3
(í47.9/í47.4/í49.5/na/ns/ns)
6.3 ± 0.7
(na/na/6.3/na/ns/ns)
7
995
1080
í7.42
(í7.52/í7.30/í7.33/í7.51/ns/ns)
í44.8
(í44.8/í43.6/í45.9/na/ns/ns)
5.5 ± 0.7
(na/na/5.5/na/ns/ns)
8
730
800
í6.97
(í7.03/í6.94/í6.96/í6.93/í6.99/ns)
í41.9
(í40.4/í43.0/í42.3/na/na/ns)
5.5 ± 0.8
(na/na/5.5/na/na/ns)
8bis
845
950
í7.17
(í7.29/í7.17/í7.06/ns/ns/ns)
í44.2
(í44.1/í44.5/í43.9/ns/ns/ns)
na
(na/na/na/ns/ns/ns)
9
630
660
í6.66
(í6.67/í6.64/ns/ns/ns/ns)
í40.7
(í39.6/í41.7/ns/ns/ns/ns)
4.7 ± 0.7
(na/4.7/ns/ns/ns/ns)
10
690
730
í6.81
(í6.86/í6.79/í6.73/í6.85/ns/ns)
í41.4
(í40.7/í41.5/í41.9/na/ns/ns)
4.1 ± 0.6
(na/4.2/4.0/na/ns/ns)
21
460
510
í6.69
(ns/í6.69/ns/ns/ns/ns)
í39.7
(ns/í39.7/ns/ns/ns/ns)
na
(ns/na/ns/ns/ns/ns)
31
450
540
í6.53
(ns/ns/í6.50/í6.56/ns/ns)
í41.0
(ns/ns/í41.0/na/ns/ns)
4.8 ± 0.8
(ns/ns/4.8/na/ns/ns)
32
400
430
í6.28
(ns/ns/í6.27/í6.24/í6.32/ns)
í38.6
(ns/ns/í38.6/na/na/ns)
na
(ns/ns/na/na/na/ns)
33
180
250
í6.03
(ns/ns/í5.98/í6.08/ns/ns)
í36.6
(ns/ns/í36.6/na/ns/ns)
na
(ns/ns/na/na/ns/ns)
40
45
150
í5.65
(ns/ns/ns/í5.65/ns/ns)
í33.2
(ns/ns/ns/í33.2/ns/ns)
na
(ns/ns/ns/na/ns/ns)
43
545
630
í6.74
(ns/ns/ns/í6.74/ns/ns)
í40.5
(ns/ns/ns/í40.5/ns/ns)
na
(ns/ns/ns/na/ns/ns)
44
490
520
í6.62
(ns/ns/ns/í6.73/ns/í6.40)
í39.6
(ns/ns/ns/í39.6/ns/na)
3.4 ± 0.6
(ns/ns/ns/3.4/ns/na)
47
515
530
í6.60
(ns/ns/ns/ns/í6.60/í6.60)
í39.6
(ns/ns/ns/ns/í39.6/na)
5.9 ± 0.6
(ns/ns/ns/ns/5.9/na)
48
490
530
í6.43
(ns/ns/ns/ns/í6.46/í6.40)
í39.4
(ns/ns/ns/ns/í39.4/na)
5.8 ± 0.6
(ns/ns/ns/ns/5.8/na)
53
190
240
í5.95
(ns/ns/ns/ns/í5.95/ns)
í37.5
(ns/ns/ns/ns/í37.5/ns)
na
(ns/ns/ns/ns/na/ns)
57
40
50
í5.55
(ns/ns/ns/ns/ns/í5.55)
í34.3
(ns/ns/ns/ns/ns/í34.3)
na
(ns/ns/ns/ns/ns/na)
60
95
130
í5.62
(ns/ns/ns/ns/ns/í5.62)
í34.8
(ns/ns/ns/ns/ns/í34.8)
4.3 ± 0.5
(ns/ns/ns/ns/ns/4.3)
62
140
170
í5.85
(ns/ns/ns/ns/ns/í5.85)
í36.8
(ns/ns/ns/ns/ns/í36.8)
na
(ns/ns/ns/ns/ns/na)
Notes: *: assessed by morphological and hydrogeological considerations. The only exception is for the spring 6 (see text); ns: not sampled; na: not analyzed.
79
Figure 3. į2H‰ vs. į18O‰ for the spring samples.
The regression line showing the į18O‰ values variation with the altitude is achieved in Figure 4
by means of a comparison among the isotopic contents of the springs and the respective estimated
basin average altitudes. The springs 3, 4, 4bis, 21, 44, 47 (located nearby Scansano; Figure 1) and 2
(sited nearby Pereta; Figure 1) are not involved in the computed regression line, because, if compared
to the behavior of the other springs, they show an incompatibility between the estimated feeding
average altitudes and the respective į18O‰ values. In other words, the isotopic contents of these
springs require higher recharge altitudes in respect to the estimated ones. On the other hand such
required altitudes are not detectable on the hilly reliefs neighboring the springs, so the latter seem to
drain not only local groundwater but also more extended flow paths, which will be discussed in the
next paragraph.
Figure 4. Recharge average altitudes vs. į18O‰ for the spring waters (see text for the
methodology through which the recharge altitudes were estimated).
80
The regression line equation is “Altitude (m) = í516 × į18O‰ í 2789” (R2 = 0.98) and will be
used to assess the feeding average altitude of the major groundwater flow systems. The į18O vertical
gradient is about í0.2‰ every 100 meters of altitude increasing, with a į18O‰ of about í5.4 at sea
level. Such values are in agreement with those generally observed along the Tyrrhenian coast of
Italy [28,29,37–39].
Fourteen of the 25 sampled springs were analyzed for tritium contents, achieving values between
2.9 and 6.3 TU (Table 1). In order to achieve preliminary indications on the average age of the spring
water, such contents were compared to the rainfall tritium values, the latter opportunely depleted
according to the decay low (Figure 5). Most of the analyzed springs showed tritium values between
3.4 and 5.5 TU, which are likely linked to the rainfall annual values of the recent years, taking also
into account that such springs drain local flow path from unconfined systems. The 6.3 TU of
the spring 6 indicate an average residence time of about 20 years or longer, in agreement with
the extension and hydrodynamic behavior of the volcanic aquifer which is drained by the same
spring [30,33]. In addition, the values of 5.9 and 5.8 TU, which were detected for the springs 47
and 48, can be representative of an average age of 20 or more years, anyway such springs were
analyzed only once and the relative values may be affected by annual fluctuations that occur in
rainfall tritium contents. Finally, groundwater drained by the spring 1, which showed 2.9 TU
practically constant in two samplings performed at different periods of the same year, seems to have
an average residence time longer than 20 years, which is congruent with the marly lithology in which
the relative aquifer is developed.
Figure 5. Tritium values of the springs in comparison to tritium values of precipitations.
3.2. Groundwater Flow Framework
The isotopic data of the water collected at the wells and streamwaters located in the
Scansano-Magliano in Toscana area are inserted in Table 2. Water from wells generally showed an
81
isotopic stability over time, which indicates groundwater flow in the region is not significantly
affected by the seasonal isotopic variability of the rainfall. Thus, the average isotopic values can be
used for the discussion of the source of groundwater flow. As showed by Figure 6, the wells waters
cluster along the local meteoric line (LMWL). Consequently, isotopic exchange phenomena does not
appear to be an issue for the groundwater at this study site and the observed features are completely
linkable at that of the infiltration water, or, in other words, water stable isotopes work as
environmental tracers.
The range of į18O in the well water was from í6.75‰ to í5.35‰, with the lighter signatures
referred at the Scansano-Pancole zone and the heavier ones registered in the Albegna Plain. A general
increasing of the values is observed moving from Scansano to Magliano in Toscana, with a
substantial agreement with a decrease in altitude. However, this trend and the distribution of
the absolute values across the territory seem to be also affected by the hydro-structural features.
Generally, the lighter signatures were detected for most of the wells that tap into the sandstones
aquifer (Figures 6 and 7), whereas the heavier į18O signatures were found in the groundwater flowing
into the marly/shaly complexes. These features for groundwater flow hosted in the sandstones is
confirmed by the isotopic values of the springs that drain this aquifer in the Scansano zone. Two
groundwater flow systems can be so identified in the Scansano-Magliano in Toscana ridge: (i) the
first concerns the marly/shaly complexes and have heavier isotopic signatures; (ii) the second
develops in the sandstones aquifer and shows lighter isotopic signatures.
Figure 6. į2H‰ vs. į18O‰ for the wells waters. (1) wells into the sandstones; (2) wells
into the other hydrogeological complexes; LMWL is as defined in Figure 3.
82
Table 2. Isotopic data of wells and streamwaters.
Depth of Well
į18O‰(VSMOW)
į18O‰ Values
į2H‰(VSMOW)
į2H‰ Values
(m) or
(Precision 0.10‰)
(March 04/June–July 04/
(Precision 1.5‰)
(March 04/June–July 04/
Name of
Mean Value or
September 04/May–June 05/
Mean Value or
September 04/May–June 05/
Streamwater
Single Datum
September 05/October 05)
Single Datum
September 05/October 05)
415
50
í6.64
(í6.67/í6.61/ns/ns/ns)
í40.1
(í40.4/í39.7/ns/ns/ns)
3.7 ± 0.6
(na/3.7/ns/ns/ns)
545
80
í6.60
(í6.62/í6.58/ns/ns/ns)
í39.8
(í39.2/í40.4/ns/ns/ns)
4.2 ± 0.7
(na/4.2/ns/ns/ns)
13
255
30
í6.19
(í6.23/í6.13/í6.21/ns/ns)
í37.6
(í38.2/í36.9/na/ns/ns)
3.7 ± 0.6
(na/3.6/3.7/ns/ns)
13bis
190
5
í6.18
(í6.18/ns/ns/ns/ns)
í37.7
(í37.7/ns/ns/ns/ns)
na
(na/ns/ns/ns/ns)
14
185
11
í5.82
(í5.79/ns/ns/í5.86/ns)
í35.0
(í35.0/ns/ns/na/ns)
5.9 ± 0.7
(5.9/ns/ns/na/ns)
15
60
40
í5.46
(í5.51/í5.40/ns/ns/ns)
í32.6
(í33.1/í32.0/ns/ns/ns)
6.6 ± 0.7
(na/6.6/ns/ns/ns)
16
26
48
í5.47
(í5.42/í5.51/ns/ns/ns)
í32.6
(í33.8/í31.5/ns/ns/ns)
3.7 ± 0.7
(na/3.7/ns/ns/ns)
19
58
85
í5.81
(í5.81/ns/ns/ns/ns)
í36.2
(í36.2/ns/ns/ns/ns)
na
(na/ns/ns/ns/ns)
Sampled
Altitudes of
Wells or
Ground Level
Streamwaters
(m a.s.l.)
11
12
3
3
H(TU) ± Err
Mean Value or
Single Datum
H(TU) Values
(March 04/June–July 04/
September 04/May–June 05/
September 05/October 05)
22
250
55
í6.05
(í6.00/í6.10/ns/ns/ns)
í38.0
(í37.1/í38.8/ns/ns/ns)
2.6 ± 0.6
(2.5/2.7/ns/ns/ns)
23
400
66
í6.51
(í6.52/í6.46/í6.56/ns/ns)
í39.2
(í38.5/í39.9/na/ns/ns)
na
(na/na/na/ns/ns)
24
365
54
í5.75
(í5.78/í5.72/ns/ns/ns)
í34.2
(í35.2/í33.2/ns/ns/ns)
4.3 ± 0.7
(na/4.3/ns/ns/ns)
25
345
3
í5.96
(í5.93/ns/í5.99/ns/ns)
í35.4
(í35.4/ns/na/ns/ns)
na
(na/ns/na/ns/ns)
26
120
60
í5.82
(í5.79/í5.78/í5.89/ns/ns)
í35.6
(í35.3/í35.9/na/ns/ns)
5.4 ± 0.8
(na/5.4/na/ns/ns)
27
550
83
í6.70
(í6.67/í6.68/í6.75/ns/ns)
í41.2
(í42.7/í39.6/na/ns/ns)
3.7 ± 0.7
(na/3.9/3.5/ns/ns)
28
495
8
í6.50
(ns/í6.42/í6.59/ns/ns)
í38.7
(ns/í37.7/í39.7/ns/ns)
5.0 ± 0.7
(ns/5.0/na/ns/ns)
29
500
50
í6.58
(ns/í6.55/í6.61/ns/ns)
í39.0
(ns/í39.0/na/ns/ns)
2.4 ± 0.5
(ns/2.5/2.3/ns/ns)
30
50
30
í5.93
(ns/í5.92/í5.94/ns/ns)
í35.4
(ns/í35.4/na/ns/ns)
4.6 ± 0.7
(ns/4.1/5.0/ns/ns)
34
42
80
í5.36
(ns/í5.28/í5.44/ns/ns)
í32.8
(ns/í31.7/í33.8/ns/ns)
3.2 ± 0.6
(ns/3.2/na/ns/ns)
35
223
70
í6.01
(ns/í6.09/ns/í5.94/ns)
í36.7
(ns/í36.7/ns/na/ns)
5.6 ± 0.7
(ns/5.6/ns/na/ns)
36
185
32
í6.06
(ns/í6.08/í6.05/ns/ns)
í33.8
(ns/í33.8/na/ns/ns)
0.8 ± 0.4
(ns/0.8/0.9/ns/ns)
37
285
80
í5.91
(ns/í5.81/í6.02/ns/ns)
í35.3
(ns/í35.3/na/ns/ns)
3.8 ± 0.7
(ns/3.8/na/ns/ns)
Table 2. Cont.
Sampled
Altitudes of
Wells or
Ground Level
Streamwaters
(m a.s.l.)
38
41
Depth of Well į18O‰(VSMOW)
į18O‰ Values
į2H‰(VSMOW)
į2H‰ Values
3
3
H(TU) ± Err
H(TU) Values
(m) or
(Precision 0.10‰)
(March 04/June–July 04/
(Precision 1.5‰)
(March 04/June–July 04/
Name of
Mean Value or
September 04/May–June 05/
Mean Value or
September 04/May–June 05/
Streamwater
Single Datum
September 05/October 05)
Single Datum
September 05/October 05)
80
5
í5.56
(ns/ns/í5.56/ns/ns)
í34.7
(ns/ns/í34.7/ns/ns)
na
(ns/ns/na/ns/ns)
49
40
í5.99
(ns/ns/í6.02/í6.07/í5.89)
í37.6
(ns/ns/í37.6/na/na)
na
(ns/ns/na/na/na)
42
49
35
í5.97
(ns/ns/í5.97/ns/ns)
í36.5
(ns/ns/í36.5/ns/ns)
na
(ns/ns/na/ns/ns)
45
346
40
í5.90
(ns/ns/í5.90/ns/ns)
í36.9
(ns/ns/í36.9/ns/ns)
5.3 ± 0.7
(ns/ns/5.3/ns/ns)
49
360
4
í5.70
(ns/ns/ns/í5.70/ns)
í35.7
(ns/ns/ns/í35.7/ns)
4.1 ± 0.6
(ns/ns/ns/4.1/ns)
50
328
38
í6.15
(ns/ns/ns/í6.15/ns)
í37.8
(ns/ns/ns/í37.8/ns)
na
(ns/ns/ns/na/ns)
51
38
50
í5.97
(ns/ns/ns/í5.97/ns)
í37.7
(ns/ns/ns/í37.7/ns)
na
(ns/ns/ns/na/ns)
52
218
60
í6.03
(ns/ns/ns/í6.03/ns)
í36.6
(ns/ns/ns/í36.6/ns)
na
(ns/ns/ns/na/ns)
56
355
50
í5.93
(ns/ns/ns/í5.93/ns)
í36.5
(ns/ns/ns/í36.5/ns)
na
(ns/ns/ns/na/ns)
59
52
40
í5.55
(ns/ns/ns/ns/í5.55)
í33.2
(ns/ns/ns/ns/í33.2)
na
(ns/ns/ns/ns/na)
39
53
Patrignone
í5.53
(ns/ns/í5.80/ns/í5.27)
í34.1
(ns/ns/í34.1/ns/na)
4.5 ± 0.6
(ns/ns/4.5/ns/na)
63
72
Castione
í5.84
(ns/ns/ns/ns/í5.84)
í34.9
(ns/ns/ns/ns/í34.9)
na
(ns/ns/ns/ns/na)
Mean Value or
Single Datum
(March 04/June–July 04/
September 04/May–June 05/
September 05/October 05)
Notes: ns: not sampled; na: not analyzed.
83
84
Figure 7. Thematic map of the į18O‰ values.
Downstream of Magliano in Toscana, where the alluvial system dominates, very different values
of į18O were detected in wells that are close to each other and have a similar depth. The most evident
case regards the well 15, whose average value is í5.4‰, and the wells 30, 41, 42, 51, which are
characterized by values from í5.9‰ to í6.0‰ (Figures 6 and 7). The latter values are congruent
with those detected in the sandstones groundwater flow, thus suggesting a likely groundwater transfer
from the sandy aquifer toward the alluvial aquifer system. This is in agreement with the distribution
of the wells, since the 30, 41, 42, 52 are in front of the zone in which the alluvial sediments overlap
the sandstones, contrarily to the 15 which is westward. The possibility that the Patrignone creek,
given the proximity, affects the isotopic features of the wells 30, 41, 42, 51, seems instead to be
85
unlikely, because in contrast with the homogeneity and stability observed in the wells water values
the streamwater showed a consistent į18O variability, between í5.3‰ and í5.8‰.
An evaluation of the recharge average altitudes for the groundwater flowing in the study area can
be performed by means of the diagram in Figure 8. The latter takes into account the “infiltration
average altitudes/į18O values” relationship, previously achieved (Figure 4), and moreover the
average į18O values of all wells and of the springs which denoted a draining of groundwater not
exclusively fed by local infiltration water (to insert the wells and springs points into the diagram the
altitudes of ground level were considered).
Figure 8. Comparison between the “infiltration average altitudes/į18O values”
relationship (1) and the į18O‰ values achieved in the Scansano-Magliano in Toscana
zone from wells and springs draining sandstones aquifer (3 and 4, respectively), and from
wells draining the other hydrogeological complexes (2).
For the springs (3, 4, 4bis, 21, 44, 47) and wells (11, 12, 23, 27, 29) that drain groundwater from
sandstones in the Scansano-Pancole zone, a recharge average altitude in the range 600–700 m (a.s.l.)
is assessable. Considering that this zone mainly develops between 450 and 550 m (a.s.l.), and that its
highest altitude is about 580 m (a.s.l.) (Figures 1 and 7), the presence of a significant regional
groundwater component is required to justify the above mentioned range of feeding average altitudes.
Indeed, in the surrounding area the nearest zones with altitudes of 600–700 m a.s.l. (and higher ones,
necessarily involved to obtain such range of average values) are present toward NE and at more than
13 km from Scansano.
The water sampled in the Pereta zone (2, 13, 13bis, 36, 50), and again representative of the
sandstones aquifer, have isotopic features that indicate the range 350–450 m a.s.l. as average altitudes
of feeding. These elevations are detectable between Pereta and Scansano at about 4–6 km from the
sampling sites; at any rate, the groundwater flow paths may also be considered more extended,
86
because, being average elevations, a feeding from altitudes higher than 450 m a.s.l., such as that of
the Scansano zone, is likely. A probable hypothesis is that the groundwater of the Scansano zone
represents the starting point of a flow path that develops with continuity southward, and whose į18O
values are influenced by the į18O values of the water that infiltrates into the sandstones aquifer at the
minor altitudes encountered moving in this direction. At least up to Magliano in Toscana, such
sandstones groundwater flow system and its transfer in the alluvial deposits are verified by isotopic
features of the wells 19, 30, 41, 42, 51. In fact, as highlighted in Figure 8, recharge average altitudes
of about 200–300 m a.s.l. are assessable for these water points, and they are consistent with an
involvement of the zones that develop northward from Pereta. All other points plotted into the
diagram, and representative of groundwater hosted in marly, shelly or alluvial complexes, denote a
recharge from average altitudes that are similar to the elevation of the zones in which they are located,
thus suggesting that these groundwater flow systems have generally a local importance. Five wells
(24, 25, 45, 49, 56) even showed į18O values apparently incompatible with the regression line in the
Figure 8, because based on the same line a recharge average altitude below the ground level is
achieved (also taking into account the bottom of the well the incongruence persists). These wells are
all located in a restricted area near Pereta (Figure 7), so the verified behavior seems to be connected
to local causes. Although with the available data it is not possible to reach a conclusion, a probable
influence of hydro-structural conditions on the local groundwater isotopic features can be supposed.
The marly complex, in which the wells tap groundwater, is here interested by a wide fractured zone
between two main sub-vertical faults, for the presence of the nearby tectonic contact with the
sandstones [24]. High rainfall infiltration rates can so occur and locally lead to an isotopic variability
in groundwater during the hydrologic year. Since the mentioned wells were sampled only once or
twice, their isotopic values in such hypothesis would not be representative of the average annual
values in local groundwater and therefore not even of the average altitudes of the recharge area.
In the framework above delineated, Figure 6 might be also viewed as representative of a mixing
in which regional groundwaters and shallower, locally recharged groundwaters are involved. This
would suggest that regional water mixes upwards into the shallower systems. However, further
hydrogeological investigation is required to resolve the appropriate conceptual model.
Tritium was analyzed on 19 of the 31 wells (Table 2), and where the analyses were twice
performed the contents have been stable over time. In order to achieve preliminary indications on the
average age of the well water, the results of analysis are compared in Figure 9 to the annual average
content of rainfall, previously depleted by means of the decay law. Most of the well water is
characterized by values that are congruent with those of the rain of the last 4–5 years, anyway such
values may be also affected by water infiltrated in the period 1980–2000.
The wells 15, 22, 29, 36 are distinguished from other ones. The water of the well 15 (6.6 TU)
seems to be characterized by an average age of 25 or more years, which may be due to lowpermeability conditions for a local prevalence of clayey sediments in the alluvial system. The tritium
contents of 2.6, 2.4 and 0.8 TU, detected for the other three wells, are compatible with average
residence times of 50 or more years. Since on the base of į18O values these water wells belong to
sandstones groundwater flow system and most of water points of this system showed shorter
residence time, local conditions of low-permeability in this aquifer can be supposed.
87
Figure 9. Tritium values of the well water in comparison to tritium values of precipitations.
3.3. Preliminary Elaboration of the Conceptual Hydrogeological Model
Previous discussion has highlighted as in the Scansano zone the sandstones aquifer receives, in
addition to local infiltration water, an input from regional groundwater. Presuming mixing of these
two components as end-members, their mixing can be translated into the following equations, which
refer to į18O and 3H parameters:
− 6.4 × (1 − R ) + X × R = −6.6
(1)
5 × (1 − R) + Y × R = 4
(2)
Where,
•
•
•
í6.4 (‰) is the value of į18O for the local water. It corresponds to the approximation to first
decimal place of the value (í6.37‰) achieved from the “infiltration average altitudes/į18O
values” relationship, considering the average altitude of the Scansano zone (500 m a.s.l.). At
this value of altitude, the upper and lower limit of the 95% confidence interval of regression
line in Figure 4 are í6.32‰ and í6.42‰, respectively;
í6.6 (‰) is the value of į18O for the result of the mixing. It is the approximation to first
decimal place of the value (í6.64‰) achieved by the average of the data of the twelve water
points (wells and springs) which are representative of such mixing. The standard deviation
“ı” of the data set is 0.07;
5 (TU) is the value of 3H for the local water. It corresponds to the approximation of the
average value (5.1 TU; ı = 0.4) of the tritium annual data of rainfall for the years 2004 and
2005 (just before the study period);
88
•
•
•
4 (TU) is the value of 3H for the result of the mixing. It is the approximation of the value
(3.8 TU) achieved by the average of the available data for the water points which are
representative of the mixing (number of points = 4; ı = 0.3);
(1 í R) and R are, respectively, the percentages of local and regional groundwater involved
in the mixing;
X and Y are, respectively, the į18O and 3H values of the regional groundwater.
It should be noted that the differences between the isotopic values inserted in the equations for
the local water and for the mixing result are of the order of the analytical error (for both į18O and
3
H). Even so, taking into account the performed approximations, the ı values and the 95% confidence
interval above mentioned, it is possible to state that such differences are significant.
Solving the Equations (1) and (2) for R, and after simplifications, the following equation
is achieved:
X = − 7 . 4 + 0 . 2Y
(3)
Taking into account the 3H values of the local input and of the mixing result, the 3H value in
regional groundwater is necessarily less than 4.0 TU. A value in the range 0–3 TU can be considered
realistic, given that:
•
•
based on the morphology, the average altitude achieved from the į18O value of the mixing
indicates as the regional groundwater flow is activated more than 13 km from Scansano.
Especially in such hydrostructural context, a resident time of several tens of years is
consequently probable, and in this case the range 0-3 TU is consistent with the tritium curve
showed in Figure 9;
as showed by the evolution of the į18O values occurring in groundwater along the
Scansano-Magliano in Toscana ridge, a significant rate of local infiltration in sandstone
occurs, consequently to modify the value from 5 TU (local groundwater) up to 4 TU (mixing
result) a tritium content abundantly lower than 4 TU is likely for regional groundwater.
Assuming that Y in Equation (3) is 0 or 3 TU, alternatively, we obtain į18O values of í6.8‰ and
í7.4‰, which indicate recharge average altitudes for the regional groundwater within the range
700–1000 m (a.s.l.); at these two extreme conditions of mixing, percentages of regional groundwater
(R) of 50% and 20% are respectively achieved. Such average altitudes, if crossed with the
morphologic and hydrostructural features of the region, suggest that the main recharge area for
regional groundwater could develop among Mt. Labbro, Rocchette, Catabbio and Petricci, in which
calcarenites and limestone outcrop (Figure 1). Indeed, the volcanites of Mt. Amiata (altitudes up to
1700 m a.s.l.) are not involved in regional flow because for this aquifer the infiltration is completely
balanced by perennial yield [33]. Thus, in agreement with the general low permeability of the
lithologies, important contributes from the zone between Scansano and Roccalbegna (Figure 1) may
also be excluded, because in the absence of input from Mt. Amiata they would lead to lower recharge
average altitudes than 700–1000 m (a.s.l.).
A preliminary conceptual model for the groundwater flow in the sandstone aquifer can be at this
point proposed (Figure 10):
89
•
•
regional groundwater, mainly fed from the Mt. Labbro-Rocchette-Catabbio-Petricci zone,
reaches the sandstones aquifer in the Scansano zone by means of an upflow in the fault
systems, and mixes with the local infiltration water. An uprising of regional groundwater
along such fault systems was also suggested by Francese et al. (2009) [24];
groundwater resources resulting from such mixing are partially drained by springs and
withdrawn by wells, and partially take part in a groundwater flow system that with continuity
develops within sandstones at least up to Magliano in Toscana, where a groundwater transfer
toward the alluvial system also occurs.
Figure 10. Groundwater flow conceptual model: (a) simplified map of the
hydrogeological complexes and location of the main recharge area supposed for regional
groundwater; (b) hydrogeological cross section showing hypothesized flow lines of the
components involved in the groundwater flow.
4. Conclusions
The water isotopes, which were analyzed in groundwater, allowed us to define the groundwater
framework in the area of Scansano-Magliano in Toscana ridge and to propose a conceptual
hydrogeological model. Thanks to a detailed isotopic characterization of infiltration water, which
90
was achieved by analyzing minor springs opportunely selected, the data of groundwater from wells
and major springs highlighted the presence of a main groundwater flow system. The latter develops
with continuity within the sandstones complex for more than 12 km, moving from Scansano up to
Magliano in Toscana, where at least partially it transfers groundwater to the alluvial aquifer system.
In the highest part of such groundwater system, nearby Scansano, the isotopes suggest a mixing
process that involves local infiltration water and regional groundwater, whose recharge area, on the
base of į18O and 3H mass balances, seems mainly develops on the calcarenite and limestones
outcropping in the Mt. Labbro-Rocchette-Catabbio-Petricci zone, more than 15 km away from Scansano.
The results discussed in this work underpin that water isotopes help groundwater flow
understanding, especially in fractured aquifer for which the application of conventional methodology
is not always effortless. In the specific case, the isotopic tools pointed out as the sandstone aquifer
of the Scansano-Magliano in Toscana ridge may be a strategic and alternative resource for water
supplying, given the overexploitation and contamination of the nearby alluvial aquifers. In this
context, the sandstone aquifer, in addition to representing an important water source for the isolated
villages that exist on the ridge, might also be tapped to improve, both in quality and quantity, the
water supplying for the villages in the nearby coastal plain. Indeed, in the latter, especially during
the summer period, the high water demand linked to the touristic vocation of the region leads to an
excessive drawdown in the local groundwater body with consequent seawater intrusion occurrence
and worsening of water quality.
Finally, the achieved results encourage and promote more detailed surveys of the chemical,
geophysical, and hydrodynamic type, which should be aimed at defining what zones of the sandy
aquifer are more suitable for groundwater exploitation.
Acknowledgments
The authors wish to thank Sandra Trifiro', Enrico Calvi, Elisa Ferrari, Maurizio Catania,
Caterina Giorgi of the IGG-CNR for their precious work in field sampling and laboratory analyzing.
The authors thank the editor and two anonymous reviewers for their useful comments, which
helped us to improve the manuscript. This research was financially supported by a grant of
AATO6-Ombrone.
Author Contributions
Both Authors planned and carried out this research, and cooperated in the elaboration and
interpretation of the achieved data. The paper was prepared under the direction of Marco Doveri.
Conflicts of Interest
The authors declare no conflict of interest.
91
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Panichi, C.; Gonfiantini, R. Environmental isotopes in geothermal studies. Geothermics 1978,
6, 143–161.
Fontes, J.C. Environmental Isotopes in Groundwater Hydrology. In Handbook of Environmental
Isotope Geochemistry, 1st ed.; Fritz, P., Fontes, J.C., Eds.; Elsevier Scientific Publishing:
Amsterdam, The Netherlands, 1980; Volume 1, pp. 75–134.
Yurtsever, Y.; Gat, J.R. Atmospheric waters. In Stable Isotope Hydrology (Deuterium and
Oxygen-18 in the Water Cycle), 1st ed.; Technical Reports 210; Gat, J.R., Gonfiantini, R., Eds.;
International Atomic Energy Agency (IAEA): Vienna, Austria, 1981; pp. 103–142.
Clark, I.; Fritz, P.; Precipitation. In Environmental Isotopes in Hydrogeology, 1st ed.; Stein, J.,
Starkweather, A.W., Eds.; Lewis Publishers CRC Press: Boca Raton, FL, USA, 1997; pp. 63–78.
Wood, W.W.; Sanford, W.E. Chemical and isotopic methods for quantifying groundwater
recharge in a regional, semiarid environment. Ground Water 1995, 33, 458–468.
Mohammad, Z.I. Application of environmental isotopes in storm-discharge analysis of
two contrasting stream channel in a watershed. Water Res. 1998, 32, 2959–2968.
Marfia, A.M.; Krishnamurthy, R.V.; Atekwana, E.A.; Panton, W.F. Isotopic and geochemical
evolution of ground and surface waters in a karst dominated geological setting: A case study
from Belize, Central America. Appl. Geochem. 2004, 19, 937–946.
Doveri, M.; Menichini, M.; Cerrina Feroni, A. Stable water isotopes as fundamental tool in karst
aquifer studies: Some results from isotopic applications in the Apuan Alps carbonatic complexes
(NW Tuscany). Italian J. Eng. Geol. Environ. 2013, 1, 33–50.
Consorzio Lamma. Territorio: Geologia, Carta Geologica Regionale 1:10,000. Available online:
http://www.lamma.rete.toscana.it/territorio/geologia (accessed on 15 January 2013).
Servizio Idrologico Regionale: Archivio dati. Available online: http://www.sir.toscana.it
(accessed on 21 March 2013).
Vittorini, S. Ricerche sul clima della Toscana in base all’evapotraspirazione potenziale e al
bilancio idrico. Rivista Geografica Italiana 1972, 1, 1–30.
Boccaletti, M.; Elter, P.; Guazzone, G. Plate tectonics models for the development of Western
Alps and Northern Apennines. Nature 1971, 234, 108–111.
Boccaletti, M.; Coli, M.; Decandia, F.A.; Giannini, E.; Lazzarotto, A. Evoluzione
dell’Appennino settentrionale secondo un nuovo modello strutturale. Memorie della Società
Geologica Italiana 1981, 21, 359–373.
Molli, G. Northern Apennine-Corsica orogenic system: An updated review. Geol. Soc. Lond.
Spec. Publ. 2008, 298, 413–442.
Mongelli, F.; Zito, G. Flusso di calore nella regione Toscana. Studi Geologici Camerti 1991, 1,
91–98.
Elter, P. L’ensemble ligure. Bulletin de la Société Géologique de France 1975, 17, 984–997.
Marroni, M.; Treves, B. Hidden terranes in the northern Apennines, Italy: A record of Late
Cretaceous-Oligocene transpressional tectonics. J. Geol. 1998, 106, 149–162.
92
18. Nirta, G.; Pandeli, E.; Principi, G.; Bertini, G.; Cipriani, N. The Ligurian Units of Southern
Tuscany. Bollettino della Società Geologica Italiana 2005, 3, 29–54.
19. Jolivet, L.; Daniel, J.M.; Truffert, C.; Goffé, B. Exhumation of deep crustal metamorphic rocks
and crustal extension in back-arc regions. Lithos 1994, 33, 3–30.
20. Carmignani, L.; Decandia, F.A.; Disperati, L.; Fantozzi, P.L.; Lazzarotto, A.; Liotta, D.;
Oggiano, G. Relationships between the Sardinia-Corsica-Provencal Domain and the Northern
Apennines. Terra Nova 1995, 7, 128–137.
21. Brunet, C.; Monié, P.; Jolivet, L.; Cadet, J.P. Migration of compression and extension in the
Tyrrhenian Sea: Insights from 40Ar/39Ar ages on micas along a transect from Corsica to
Tuscany. Tectonophysics 2000, 321, 127–155.
22. Carmignani, L.; Kligfield, R. Crustal extension in the Northern Apennines: Transition from
compression to extension in the Alpi Apuane core complex. Tectonics 1990, 9, 1275–1303.
23. Brogi, A.; Lazzarotto, A.; Liotta, D.; Ranalli, G. Extensional shear zones as imaged by reflection
seismic lines: The Larderello geothermal field (Central Italy). Tectonophysics 2003, 363,
127–139.
24. Francese, R.; Mazzarini, F.; Bistacchi, A.; Morelli, G.; Pasquarè, G.; Praticelli, N.; Robain, H.;
Wardell, N.; Zaja, A. A structural and geophysical approach to the study of fractured aquifers
in the Scansano-Magliano in Toscana Ridge, southern Tuscany, Italy. Hydrogeol. J. 2009, 17,
1233–1246.
25. Fritz, P.; Fontes, J.C. Introduction. In Handbook of Environmental Isotope Geochemistry,
1st ed.; Fritz, P., Fontes, J.C., Eds.; Elsevier Scientific Publishing: Amsterdam, The
Netherlands, 1980; Volume 1, pp. 1–19.
26. Epstein, S.; Mayeda, T. Variation of O18 content of waters from natural sources. Geochim.
Cosmochim. Acta 1953, 4, 213–224.
27. Coleman, M.L.; Shepherd, T.J.; Durham, J.J.; Rouse, J.E.; Moore, G.R. Reduction of water with
zinc for hydrogen isotope analysis. Anal. Chem. 1982, 54, 993–995.
28. Longinelli, A.; Selmo, E. Isotopic composition of precipitation in Italy: A first overall map.
J. Hydrogeol. 2003, 270, 75–88.
29. Mussi, M.; Leone, G.; Nardi, I. Isotopic geochemistry of natural water from the Alpi
Apuane-Garfagnana area, Northern Tuscany, Italy. Mineral. Petrogr. Acta 1998, 41, 163–178.
30. Doveri, M.; Nisi, B.; Cerrina Feroni, A.; Ellero, A.; Menichini, M.; Lelli, M.; Masetti, G.;
Da Prato, S.; Principe, C.; Raco, B. Geological, hydrodynamic and geochemical features of the
volcanic aquifer of Mt. Amiata (Tuscany, central Italy): An overview. Acta Vulcanol. 2011–
2012, 23–24, 51–72.
31. International Atomic Energy Agency and World Meteorological Organization. Global Network
of Isotopes in Precipitation: The GNIP Database. Available online: http://isohis.iaea.org
(accessed on 24 November 2005).
32. Doveri, M.; Leone, G.; Mussi, M.; Zanchetta, G. Composizione isotopica di acque ipogee
nell’Antro del Corchia (Alpi Apuane, Toscana nord-occidentale). Memorie dell’Istituto Italiano
di Speleologia Serie 2 2005, 18, 119–132.
93
33. Barazzuoli, P.; Bosco, G.; Nante, N.; Rappuoli, D.; Salleolini, M. The aquifer of Mount Amiata:
Evaluation of the perennial yield and its quality. Memorie della Società Geologica Italiana
1994, 48, 825–832.
34. Craig, H. Isotopic Variation in meteoric waters. Science 1961, 133, 1702.
35. Gat, J.R.; Carmi, I. Evolution of the isotopic composition of atmosferic waters in the
Mediterranean area. J. Geophys. Res. 1970, 75, 3039–3048.
36. Minissale, A.; Magro, G.; Vaselli, O.; Verrucchi, C.; Perticone, I. Geochemistry of water and
gas discharges from the Mt. Amiata silic complex and surrounding areas (central Italy). J.
Volcanol. Geotherm. Res. 1997, 79, 223–251.
37. Hauser, S.; Dongarrà, G.; Favara, R.; Longinelli, A. Composizione isotopica delle piogge in
Sicilia. Riferimenti di base per studi idrogeologici e relazioni con altre aree mediterranee.
Rendiconti della Società Italiana di Mineralogia e Petrologia 1980, 36, 671–680.
38. Leone, G.; Mussi, M. Gli isotopi ambientali nelle acque di sorgenti e pozzi e nelle emergenze
sulfuree sottomarine di Capo Palinuro, Salerno. Geologia tecnica e ambientale 2004, 3, 35–55.
39. Celati, R.; Grassi, S.; D’Amore, F.; Marcolini, L. The low temperature hydrothermal system of
Campiglia, Tuscany (Italy): A geochemical approach. Geothermics 1991, 20, 67–81.
Section 2: Multi-Isotope Studies
96
Isotopes as Tracers of Water Origin in and Near a Regional
Carbonate Aquifer: The Southern Sacramento Mountains,
New Mexico
Christopher J. Eastoe and Ryan Rodney
Abstract: High-elevation groundwater sampled in 2003 in the Sacramento Mountains defines a line
resembling an evaporation trend in įD-į18O space. The trend results from recharge of winter
precipitation into fractured limestone, with evaporation prior to recharge in broad mountain valleys.
The same trend occurs in basin groundwater east and west of the range, indicating the high
Sacramento Mountains as the principal regional water source, either direct from the limestone
aquifers or from mountain-derived surface water. Tritium and carbon-14 indicate bulk residence
times of a few decades in the high Sacramento Mountains and at Alamogordo, and of thousands of
years south of Alamogordo and in the artesian aquifer near Artesia. Stable O, H isotope data fail
to demonstrate the presence of Sacramento Mountains water in a saline aquifer of the Hueco
Bolson (Texas).
Reprinted from Water. Cite as: Eastoe, C.J.; Rodney, R. Isotopes as Tracers of Water Origin in and
Near a Regional Carbonate Aquifer: The Southern Sacramento Mountains, New Mexico. Water
2014, 6, 301-323.
1. Introduction
In the Basin and Range province of the southwest USA, stable isotope studies have proved useful
in distinguishing sources of recharge where altitude effects are large, e.g., Tucson Basin [1], or where
isotope effects due to latitude/altitude and evaporation generate river water that is distinctive beside
native basin groundwater [2,3].
The Sacramento Mountains of south-central New Mexico (Figure 1) include a broad area of
forested, well-watered terrain, the source of perennial streams flowing east toward the Roswell Basin
and the Pecos River and of intermittent streams flowing west into the Tularosa Valley.
Pre-development and recent water level data indicate groundwater movement from Tularosa Valley to
the Hueco Bolson [4–6]. Mayer and Sharp [7] suggested regional flow of groundwater from the
Sacramento Mountains to the Texas-New Mexico border, through karst aquifers.
97
Figure 1. Map showing the study areas. Abbreviations are: NM = New Mexico;
TX = Texas; A = Alamogordo; C = Cloudcroft; E = Elk; EP = El Paso; FM = Franklin
Mountains; HB = Hueco Bolson; M = Mayhill; OM = Organ Mountains; P = Piñon;
R = Red Bluff; T = Tularosa; TV = Tularosa Valley; W = Weed. X-X’ is the line of the
geological section in Figure 2.
Groundwater samples for this study were collected in the Sacramento Mountains and the flanking
basins from 2003 to 2008. The aim of the study was to use environmental isotopes to determine the
relationship between water from the Sacramento Mountains and the adjacent basin aquifers. The
relationship between groundwater in the high mountains and that in the Tularosa Valley, the deep
alluvial basin to the west, is the first topic to be addressed. The relationship between groundwater in
the high mountains and that in the hard-rock Roswell artesian basin to the east is the second topic to
be addressed. In both cases, we also attempt to constrain groundwater residence times, and to
determine the seasonality of recharge. Finally, we discuss whether the isotope signature of
Sacramento Mountains groundwater can be recognized as far south as the Hueco Bolson in Texas
(Figure 1).
2. Background
Figure 1 shows the location of the study areas. Area 1, encompassing the Sacramento Mountains
near Cloudcroft and Weed, and the freshwater lens on the western flank of the mountains,
encompasses sites sampled for the first topic described above. Area 2 stretches from the eastern flank
of the mountains to the Pecos River at Artesia, and encompasses sites samples for the second topic.
98
2.1. Topography, Climate and Vegetation
The Sacramento Mountains rise to 2500–2800 m above sea level (m.a.s.l.) in the study area.
A steep western escarpment with deep canyons abuts the Tularosa Valley, a typical fault-bounded
basin of the Basin-and-Range province. Tularosa Valley continues southward into Texas where the
valley is named the Hueco Bolson; Neogene alluvium fills the entire extent of the combined basin to
depths of 600 to 3000 m in the basin center [6]. The eastern flank of the Sacramento Mountains
approximates a dip-slope, and descends gradually toward the Pecos River. No deep alluvial basin is
present on the east side of the range. The climate in the basins is semi-arid; average annual
precipitation is 335 mm at Alamogordo and 340 mm at Artesia. In the high mountains, precipitation
is higher, e.g., 715 mm at Cloudcroft near the range crest [8]. There are two wet seasons, a weak
summer monsoon (June to October) providing 65%–70% of the precipitation, and a winter season of
rain and snow from frontal weather systems [8]. The amounts of both winter and summer
precipitation vary greatly from year to year (Figure 15 of [9]). Vegetation consists of coniferous
forest interspersed with grassy valleys above 2300 m.a.s.l.. At lower elevation, scrubby oak forest
and desert scrub predominate, except along perennial streams where riparian forest is present. Much
of the study area is dry ranch land on which groundwater pumping is essential to the survival of cattle
herds. Large-scale irrigated agriculture, using quarter-section and larger center-pivot and side-roll
equipment, is practiced on the Pecos River flood plain near Artesia. Scattered irrigated plots are
present near Tularosa.
2.2. Geology
The Sacramento Mountains constitute a tilted horst with range-bounding faults on the western
side (Figure 2), and consist of Paleozoic marine sedimentary rocks, mostly Permian-Mississippian
limestone and evaporite, overlying concealed Precambrian basement [9,10]. The surface east of the
range crest approximates a dip slope, with a discontinuous veneer of lower members of the San
Andres formation overlying dolomite and anhydrite of the Yeso formation, both overlain to the east
by the Queen-Grayburg anhydrite and limestone and dolomite of the San Andres formation. Thin
(40 m) Quaternary alluvium overlies Paleozoic strata on the Pecos River flood plain. West of the
range crest, the entire Paleozoic section of the region, mainly carbonate strata, is exposed. Neogene
alluvium fills the Tularosa Valley to depths of 230 to 300 m at Alamogordo.
2.3. Geohydrology
The carbonate strata constitute a regional aquifer system conveying water from the mountains to
the basins, eastward from the range crest through the Yeso formation, and westward through the
highly fractured Paleozoic carbonate section. A map of the potentiometric surface east of the range
crest is available in ([9], Figure 18), and indicates general eastward groundwater flow. West of the
range crest, groundwater levels are less precisely known within area 1, but they decline steeply
towards the west and southwest elsewhere on the escarpment [9]. In the high mountains, the
geohydrology is complex and governed by the detailed lithology of the Yeso formation. The
following geohydrologic features are present ([9], Figure 25): a regional aquifer, locally confined
99
beneath impermeable interbeds of the Yeso formation, and probably continuous with regional
aquifers east of the range; multiple perched aquifers overlying the regional aquifer, some discharging
in small springs controlled by impermeable strata; and vadose zones above and between the perched
aquifers. Large summer rain events in 2006 and 2008 caused rapid water-level response in the
perched aquifers, but slower response in the regional aquifer. In the Roswell artesian basin, a shallow
unconfined aquifer is present in carbonate strata overlying the Queen-Grayburg anhydrite, at depths
less than 100 m below the surface at Artesia. The regional aquifer in this area, at 200–300 m below
the surface, is confined beneath the Queen-Grayburg anhydrite and was artesian at the time of first
development; subsequent pumping has lowered static water levels by tens of meters (Table 1) [11].
Groundwater is present in an unconfined basin-fill aquifer in the alluvium of the Tularosa Valley,
where supply wells pump water from the upper 50 m.
Figure 2. Cross section X-X’ (see Figure 1 for location), after Roswell Geological
Society (1956). SL = sea level. The east slope of the Sacramento Mountains is a
dip-slope with widespread veneer, too thin to depict here, of the Glorieta Sandstone and
overlying members of the San Andres formation.
2.4. Previous Isotope Studies
Stable oxygen and hydrogen isotope and tritium data were collected for rainwater, surface water
and groundwater from Roswell Basin in the late 1970s [12,13], in order to identify sources of
recharge and groundwater residence times. The authors concluded that more detailed sampling was
required, but were able to identify loci of local, rapid recharge using tritium data in the mountain
areas and near Roswell [13]. Stable oxygen and hydrogen isotope data for surface water in the Pecos
River [14,15], have been used to determine the relative contributions of winter snow and monsoon
precipitation to the river in Texas, the authors concluding that the latter predominates [14]. Sulfur
isotopes in Sacramento Mountains groundwater have been utilized to determine the relative inputs
of evaporite gypsum, oxidized sulfides and rain sulfate to the dissolved sulfate inventory [16].
Reference [17] provided stable O and H isotope and 14C data for the well-field supplying water to
the air-force base in Tularosa Valley, and interpreted the data to indicate water residence times
greater than 1000 years. The most detailed recent work is in reference [9], which presented detailed
100
stable isotope data for precipitation and groundwater collected in 2006–2009 between the range crest
and Hope, New Mexico. The authors identified predominant summer recharge in years of heavy
summer rainfall, and used tritium, 14C and CFCs to estimate groundwater residence times of decades
in the high mountains, to thousands of years in the aquifer extending east of the range. The pattern
of stable O and H isotope data in groundwater presented in reference [9] differs markedly from that
in our dataset, allowing for an improved understanding of the hydrology of the mountain range when
both datasets are taken into account. Our study also complements reference [9] in extending spatial
coverage into flanking basins east and west of the Sacramento Mountains.
3. Methods
3.1. Analytical Methods
Samples were taken from domestic, agricultural and municipal production wells, springs, and
surface water in the Peñasco and Pecos Rivers, and from rain gauges near Weed. Isotope measurements
(except accelerator mass spectrometry carbon-14) were performed at the Environmental Isotope
Laboratory, University of Arizona. Stable O, H and C isotopes were measured on a Finnigan
Delta S® dual-inlet mass spectrometer equipped with an automated CO2 equilibrator (for O) and an
automated Cr-reduction furnace (for H). Stable S isotopes were measured on a Thermo Electron
Delta Plus XL® continuous flow mass spectrometer equipped with a Costech® elemental analyzer for
preparation of SO2. Carbon-14 was measured by accelerator mass spectrometry at the NSF-Arizona
Accelerator Facility, University of Arizona. Data generated for this study are listed in Table 1.
Analytical precisions (1ı) are 0.08‰ (O), 0.9‰ (H), 0.15‰ (C) and 0.15‰ (S). Detection limits are
0.6 TU (3H) and 0.2 pMC (14C).
3.2. Correction of Raw Carbon-14 Data
The data lack sufficient detail for chemical balance modeling; therefore a simpler method based
on į13C values is used, following ([18], p. 210). Values of į13C for soil gas are assumed to be í23‰
(corresponding to 100% C3 plant matter input) for the forested mountains, and í19.9‰
(corresponding to 75% C3 input) for desert areas. “Dead” rock carbonate of Guadalupian age has
į13C values from +1 to +5‰ [19,20]; for these strata corrections were calculated for +1‰ and +4‰.
Corrected ages are given in Table 1. In the basin-fill aquifer near Alamogordo, corrections were
calculated for rock į13C from 0‰ to +3‰, representing the entire Paleozoic section.
Table 1. Site information and isotope data.
Well
Site name (Group)
Lat
Long
Site
altitude
Date
m.a.s.l.
Well
depth
m
SWL
į18O
įD
į34S
m.a.s.l.
‰
‰
‰
į13C
DIC
‰
Tritium
C-14
TU
pMC
Corrected
age
yrs BP
Study area 1: high Sacramento Mountains
1-1
Fields (H2)
32.959
í105.525
2270
10 December 2003
121
na
í7.9
í53
12.1
í8.2
0.5
82.0
post-bomb
1-2
Cloudcroft well 4 (H1)
32.9505
í105.7019
2550
8 August 2003
164
2474.5
í9.3
í63
10.3
í9.8
5.1
82.7
post-bomb
78.5
post-bomb
1-3
Ehret (W)
32.945
í105.8405
2007
9 August 2003
206
1861.5
í9.9
í71
1-4
Bearden (W)
32.9601
í105.8844
1627
10 August 2003
9
na
í8.5
í61
1-5
Macon (W)
32.9951
í105.8437
1900
9 August 2003
19
1886.0
í9.7
í66
1-6
Macon spring (W)
32.9951
í105.8437
1900
9 August 2003
na
na
í9.5
í66
2.3
12.3
2.4
2.5
1-7
Williams (W)
32.9905
í105.894
1624
9 August 2003
16
na
í9.6
í66
12.3
1.3
72.6
post-bomb
1-8
Warnock (W)
32.9892
í105.8474
1820
8 August 2003
90
1793.6
í9.9
í66
12.2
1.7
72.5
post-bomb
95.0
post-bomb
1-9
Sect. 22 Water Assoc.Spr. (W)
32.991
í105.871
1760
9 August 2003
í9.6
í66
1-10
Posey spring (H1)
32.793
í105.5779
2450
7 December 2003
í9.3
í64
1-11
Sky Ridge (H1)
32.792
í105.5672
2350
7 December 2003
í9.6
í65
1-12
Sky Ridge spring (H1)
32.794
í105.5783
2355
7 December 2003
í9.4
í64
í9.7
1-13
Scott (H2)
32.798
í105.5521
2230
7 December 2003
48
na
í8.1
í57
í9.4
1-14
Sac. Methodist Academy (H2)
32.794
í105.558
2240
7 December 2003
na
na
í8.6
í60
í9.9
0.7
1-15
Essek (H1)
32.716
í105.5305
2225
7 December 2003
273
na
í10.2
í69
í4.3
2.0
na
na
1-16
Wright (H2)
32.7414
í105.4793
2075
9 December 2003
na
na
í8.1
1-17
Bell (H2)
32.7414
í105.4793
2075
8 December 2003
na
na
í8.2
í57
1-18
Stewart (H2)
32.6953
í105.4219
2035
8 December 2003
252
na
í8.1
í57
Sand spring (H1)
32.713
í105.684
2600
7 December 2003
í10.0
í68
32.713
í105.747
2380
7 December 2003
í9.9
í66
1-19
1-20
Apple Tree Canyon spring
(H1)
A 0.5
9.0
13.0
í10.2
5.7
22.1
2,000–
4,500
2.9
12.2
í7.4
1,800–
0.6
44.3
6.2
92.8
post-bomb
2,800
í9.5
12.1
í10.1
Study area 1: Alamogordo and Tularosa
1-21
Abercrombie
33.091
í106.015
1380
25 January 2005
91
1330.0
í8.5
í59
1-22
Cates
33.075
í106.045
1347
25 January 2005
na
na
í10.5
í67
í8.1
84.4
post-bomb
1-23
Hornback
33.062
í106.063
1326
25 January 2005
42
1305.4
í8.2
í59
í8.1
2.0
79.7
post-bomb
1-24
Cinert
33.040
í106.011
1357
25 January 2005
na
na
í8.4
í59
í4.6
1.2
36.5
post-bomb
1-25
McGinn
32.99
í105.99
1360
25 January 2005
56
na
í9.1
í63
2.1
101
102
Table 1. Cont.
Well
Site
Date
Site name (Group)
Lat
Long
1-26
Dyer
32.9157
í105.9864
1319
8 August 2003
1-27
McDonald
32.9009
í106.0069
1299
8 August 2003
altitude
Well
SWL
į18O
įD
į34S
58
1291.7
í8.9
í62
12.6
91
1279.3
í8.9
í62
12.1
depth
į13C
DIC
Tritium
C-14
Corrected age
3.3
í6.0
1.2
1-28
Dellacorino
32.90
í105.96
1326
25 January 2005
96
na
í9.1
í63
1-29
Noriega
32.8954
í105.9885
1303
9 August 2003
30
na
í8.9
í62
í6.3
3.9
53.6
post-bomb
1-30
City of Alamogordo well 2
32.9681
í105.9369
1440
September 2003
na
na
í8.9
í63
í6.0
1.3
50.6
post-bomb
1-31
City of Alamogordo well 8
32.9681
í105.9369
1440
September 2003
na
na
í9.0
í63
1-32
Harrington
32.9462
í105.9469
1402
8 August 2003
121
na
í9.1
í63
12.6
1-33
Moore
32.83
í105.96
1294
8 August 2003
61
1259.0
í9.3
í64
11.2
í6.0
<0.5
39.8
500–2,200
1-34
Boyle
32.81
í105.99
1253
25 January 2005
46
1234.8
í9.5
í66
1-35
Harrell
32.81
í105.99
1253
25 January 2005
61
1233.3
í9.4
í65
1239.4
1-36
Baca
32.81
í105.99
1253
25 January 2005
76
1-37
Mount
32.74
í105.97
1234
25 January 2005
52
1-38
Wisdom
32.744
í105.966
1237
25 January 2005
49
32.562
í106.025
1230
22 March 2004
na
1.6
A 0.5
í9.2
í64
í9.0
í63
1203.1
í9.2
í64
na
í8.9
í69
í3.8
<0.6
2.9
17,850–21,000
í1.8
<0.4
2.8
12,100–18,500
1.8
75.2
post-bomb
57.5
900–1600
29.2
4,400–5,600
Southeastern Tularosa Valley
1-39
Otero County landfill
1-40
El Paso WU Brine injection site
31.973
í106.106
1269
early 2007
na
na
í9.5
í71
1-40
El Paso WU Brine injection site
31.973
í106.106
1269
early 2008
na
na
í9.5
í70
32.931
í105.282
1750
July 2006
í7.9
í55
Study area 2
2-1
Unnamed spring
2-2
J. Powell windmill
32.955
í105.277
1758
July 2006
73
1694.0
í7.9
í55
2-3
J. Powell well
32.979
í105.248
1748
July 2006
24
1732.8
í8.3
í58
2-4
H. Powell
32.921
í105.252
1740
July 2006
33
1709.5
í8.4
í58
2-5
Orton
32.892
í105.08
1602
July 2006
259
1419.1
í8.1
í56
13.8
2-6
Duncan
32.845
í104.892
1380
July 2006
na
na
í8.3
í56
12.4
2-7
Young
32.840
í104.773
1277
July 2006
na
1086.5
í7.9
í55
2-8
Hope Water Co.
32.810
í104.734
1250
July 2006
na
na
í8.3
í58
2-9
Bannon
32.783
í104.713
1042
July 2006
195
875.3
í8.3
í57
2-10
Jones
32.847
í104.613
1060
July 2006
na
999.0
í8.1
í54
í8.9
12.3
1.5
í8.8
1.0
<0.7
í6.7
14.3
<0.7
Table 1. Cont.
Well
Site name
(Group)
Lat
Long
Site
altitude
Date
Well depth
SWL
į18O
įD
2-11
Lamb
32.843
í104.568
1035
July 2006
na
na
í8.3
í57
2-12
Brown 1 D
32.741
í104.496
1085
July 2006
130
1021.0
í8.3
í58
2-13
Brown 2 D
32.764
í104.534
1085
July 2006
130
na
í7.1
í51
2-14
Brown 3 D
32.712
í104.552
1085
July 2006
na
na
í8.1
í57
į34S
Tritium
C-14
í5.9
<0.6
29.9
2,600–4,000
33.7
1,350–3,000
2-15
Joy 1 S
32.833
í104.378
1020
July 2006
81
1000.2
í7.9
í54
13.0
2-16
Joy 2 D
32.834
í104.368
1020
July 2006
290
na
í8.3
í57
13.2
<0.5
1.3
2-17
Pardue S
32.820
í104.362
1015
July 2006
61
954.0
í8.1
í55
2-18
Rodney S
32.882
í104.424
1045
July 2006
64
990.1
í7.8
í55
2-19
Mayberry 3 S
32.925
í104.412
1032
July 2006
61
1000.0
í7.2
í51
12.8
2-20
Mayberry 4 D
32.925
í104.412
1032
July 2006
304
na
í8.4
í58
13.4
Corrected
į13C DIC
2-21
Mayberry 2 D
32.937
í104.412
1030
July 2006
304
975.1
í8.2
í56
13.8
í6.1
2-22
Mayberry 1 D
32.963
í104.509
1070
July 2006
274
na
í8.3
í58
13.4
í5.4
2-23
Menefee D
32.970
í104.508
1072
July 2006
274
1044.6
í8.2
í57
13.1
Hatfield N well
33.574
í104.482
1105
May 2007
na
na
í6.6
í49
2-25
Hatfield E well
33.572
í104.479
1104
May 2007
na
na
í7.6
í53
2-26
Hatfield artesian
33.574
í104.483
1104
May 2007
na
na
í8.4
í56
Surface Water
32.887
í105.186
1730
July 2006
í8.4
í57
12.7
Rio Penasco
32.886
í104.344
1010
July 2006
í3.3
í34
12.2
Pecos river
32.886
í104.344
1010
December 2006
í6.5
í49
Pecos R.
33.209
í104.395
1041
December 2006
í6.7
í51
Pecos R.
33.382
í104.404
1056
May 2007
í2.7
í35
1-8
32.9892
í105.8474
August 2003
í6.4
í47
4.6
1-16
32.7414
í105.4793
2075
August–October 2003
í10.4
í70
4.7
1-16
32.7414
í105.4793
2075
March 2004
í8.2
í54
7.4
1-13
32.798
í105.5521
2230
August–September 2003
í6.8
í57
3.9
0.5
age
35.6
Roswell
2-24
Pecos R.
Precipitation
103
Notes: S = shallow aquifer, D = Deep (Principal) aquifer in Artesia area; na = not available; A = Apparent tritium; * meters below surface.
104
4. Area 1: High Sacramento Mountains
Samples were collected from wells near La Luz and Fresnal canyons and areas near New Mexico
Route 24 east of the range crest (Figure 3). All samples are from fractured limestone except for 1–4,
which is from shallow alluvium in Fresnal Canyon.
Figure 3. Sample location map for areas 1and 2 (see Figure 1). Stream/canyon names
are abbreviated thus: AC = Agua Chiquita; Bw = Bluewater; Fr = Fresnal; LL = La Luz;
SR = Sacramento River. Town/village names are A = Alamogordo; C = Cloudcroft;
M = Mayhill; T = Tularosa; W =Weed. Site numbers (e.g., 1) correspond to entries in
Table 1, where the corresponding number is 1-1 for area 1, or 2-1 for area 2. Black
circles: sample sites for this study; white circles: sample sites from reference [17].
4.1. O and H Isotopes
On a plot of įD vs. į18O, most of the data fall on a straight line with a slope near 5.6 (Figure 4A),
henceforth called the Sacramento Mountains Trend (SMT). The straight line intersects an estimate
of average winter precipitation at a station at 2790 m.a.s.l. (calculated as arithmetic means (because
amount data are not available) of į18O and įD for three bulk collections in March 2007, 2008 and
2009, and representing the prior 3 months; data from [9]), but does not intersect mean summer
precipitation [9] for that station. In Figure 4B, three groups of į18O values emerge in relation to site
altitude. For the wells, collar altitude is used because static water levels are not available in all cases.
Values of į18O of group W (western slopes) overlap those of group H1 (high elevations), despite the
large altitude difference between the two groups. The difference between groups H2 (high elevations,
but generally lower than H1) and H1 is too great to attribute to altitude. Group H2 sites (1-13, 1-14,
105
1-16, 1-17) are adjacent to broad, flat, canyon bottoms, a typical geomorphic feature of the
Sacramento Mountains. In such places, deep soil (more than 1 m near site 1-16) overlies carbonate
strata, while elsewhere carbonate outcrop is widespread. The data for groundwater in 2003 differ
from data for groundwater in 2006–2009 [9]. The latter occupy a field between the SMT and summer
rain for 2006 and 2008 (Figure 4A), and reflect rapid recharge from heavy monsoon rains in 2006
and 2008. Prior to 2003, there had been no large monsoon rain totals since 1997.
Figure 4. (A) Plot of į18O vs. įD for groundwater samples from the high Sacramento
Mountains. The green line encloses groundwater isotope data from [9]. Seasonal means
for precipitation and the local meteoric water line (LMWL) are for years 2006–2009 [9].
Data plotted as individual points were collected for this study in 2003; (B) Plot of
elevation of well collars vs. į18O for sample sites in the high Sacramento Mountains. The
diagonal lines show the long-term į18O lapse-rates of í1.2‰/1000 m (Tucson Basin [21],
and 1.8‰/1000 m [13]. Site numbers (e.g., 3) correspond to entries in Table 1, where the
corresponding number is 1-3.
-30
įD, ‰
GMWL
-50
SMT
summer
-60
-70
3
-80
-13
-12
H2
H1 &W
winter
-90
-11
-10
-9
-8
-7
į18 O, ‰
A
West slope
Elevation, m.a.s.l.
LMWL
2006-2009 data [9]
-40
High mountains
Seasonal means
2700
2500
H1
H2
2300
2100
1900
W
1700
4
1500
-11
-10
-9
-8
į18O ‰
B
West slope
High mountains
-7
106
4.2. Other Parameters
Groundwater in this area generally has į34S values of 10‰ to 13‰, tritium concentrations of
1 to 3 TU, and 14C in the range 72 to 93 pMC (cf. 0 to 7 TU, and 83 to 93 pMC, for samples from
2006 to 2008 [9]). Corrected 14C data indicate post-bomb water in the west-slope canyons and in two
high-elevation springs, with older groundwater (300–4500 years) at sites 1–15 and 1–18 (Table 1).
4.3. Interpretation
The SMT can be explained as an evaporation trend originating in winter precipitation. Evaporation
prior to infiltration varies in degree, and is greatest in groundwater near the broad canyon bottoms,
(sites 1–13, 1–14, 1–16 and 1–17), where standing water and wet soil are likely to undergo partial
evaporation. Well-mixed high altitude groundwater will plot between groups H1 and H2, and this
isotope signature will be found in groundwater of the limestone aquifer at lower elevations unless
water of different isotope composition is added downgradient. Evaporated runoff from high
elevations may plot on the SMT to the right of group H2. Addition of water from local low-elevation
precipitation would shift groundwater isotopes towards the GMWL.
The difference between the 2003 and 2006–2009 data sets indicates two modes of recharge.
In years with unusually wet summers, (e.g., 2006 and 2008), summer recharge with little evaporation
is the dominant source of recharge. The local meteoric water line (LMWL) in Figure 4A is governed
by rainfall from those years, and may not apply under drier conditions. Following a succession of
dry to average summers, however, winter recharge predominates, even though there is more
precipitation in summer than in winter rain on average. Such was the case from summer 1998 to
2003 when sampling for this study occurred. Under these conditions, evaporation of the infiltrating
water occurs in the broad canyon bottoms east of the range crest, but is not observed between the
range crest and the canyons on the steep west escarpment.
Tritium and corrected 14C contents of high-elevation groundwater indicate the presence of
post-bomb recharge, but tritium levels in 2003 were predominantly lower than average tritium in
post-1992 precipitation (4–7 TU, see Table 1 and [9]; compare a better-constrained average of
5.3 TU for Tucson [22]), indicating mixing with pre-bomb meteoric water. By 2006–2008, more
post-bomb recharge was present, tritium-helium dates were mainly 1–15 years, and CFC ages were
largely 20–30 years [9]. Values of į34S indicate Permian marine gypsum (+12‰ to +13‰) as
the main source of sulfate; lower values most likely reflect oxidation of sulfide present in these
strata [16].
5. Area 1—Alamogordo and Tularosa
Sampling from supply wells in basin-fill alluvium represents lower-TDS water suitable for human
consumption; brackish water is also present >5 km west of the range front. Groundwater in this area
flows west at Alamogordo and Tularosa [4,5], but parallel to the range front south of Alamogordo,
where no major canyons contribute water to the basin.
107
5.1. O and H Isotopes
Most data plot on the SMT (Figure 5), to the right of group W. Samples from Tularosa include
the most and least evaporated of the set. Data for wells south of site 1-33 (į18O between í9.9‰
and í9.5‰ in reference [17]) differ from data collected for the present study in the same area
(į18O between –9.5‰ and –9.0‰). Actual variation in į18O (as opposed to measurement error) is
unlikely in such old groundwater (see below); the earlier data are not used here. In southeastern
Tularosa Valley, sites 1-39 and 1-40 (Figure 1) have groundwater that plots below the SMT
Figure 5. Plot of į18O vs. įD for groundwater samples from basin sediments near
Alamogordo and Tularosa, in relation to samples from La Luz and nearby canyons and
the high Sacramento Mountains. Samples 1-39 and 1-40 are from basin fill more than 40
km south of Alamogordo (see Figures 1 and 3).
-30
-40
GMWL
įD ‰
-50
-60
SMT
22
-70
39
40
-80
-90
-11
-10
-9
-8
-7
į18O ‰
Alamogordo
Tularosa
Group W (La Luz Canyon)
Outlying
5.2. Other Parameters
Tritium is present (1-3 TU) north of site 1-33, and is generally absent (below detection to 0.5 TU)
south of 1-33. 14C generally decreases from near 80 pMC near Tularosa to 20 pMC south of
Alamogordo (Figure 6). Corrected 14C data indicate young groundwater (post-bomb to a few hundred
years) north of Alamogordo and in La Luz canyon, and much older water (500–7500 years,
considering also corrected data from [17]) south of Alamogordo. Values of į34S are near +12‰. At
sites 1–39 and 1–40, tritium is below detection, 14C levels are 3 pMC, and corrected ages are 12,000
to 21,000 years (Table 1).
108
Figure 6. Detail of Figure 3, showing distribution of carbon-14 (pMC) in groundwater
samples. Black circles: this study; white circles: data from [17].
84
T
37
80
51
N
54
70
US
35
A
40
44
43
20,26
17
20 km
5.3. Interpretation
O and H isotope data plotting on the SMT indicate high-elevation precipitation as the source of
groundwater in basin alluvium near Alamogordo and Tularosa. Groundwater from the high
Sacramento Mts. flows to La Luz Canyon sample sites without isotopic shift. The higher degree of
evaporation in samples from the alluvial aquifer could be explained: (1) as mountain-block recharge
combining more-evaporated and less-evaporated recharge from high elevations; or (2) as mountainfront recharge of surface water supplied from high elevations by way of the mountain canyons. The
absence of an evaporation signature in groundwater from carbonate strata in La Luz Canyon, between
the range crest and the basin) argues against the first possibility, while the presence of evaporated
water in the alluvium argues for the second. The higher degree of evaporation of groundwater farther
from the range front (Tularosa, 12–15 km from the range front), in contrast to groundwater nearer to
the range front (Alamogordo, within 6 km), suggests that the sites of infiltration of surface water
extend into the basin, rather than being confined to a narrow zone at the range front. This is
particularly evident in the case of site 1-22 at Tularosa, (Figure 5), where the coincidence of low įD
and į18O with high 14C indicates recharge of very recent runoff at a distance of up to 15 km from the
range front. Both mountain-front and mountain-block recharge seem likely, but the data do not
indicate the relative amounts. In the basin fill south of Alamogordo, tritium and 14C data are consistent
with slow southward flow of groundwater, with little recharge from nearby mountain canyons.
109
6. Area 2: Peñasco to Artesia
Samples were collected between Peñasco and Artesia (Figure 3). Near Peñasco, groundwater
samples were taken from a spring and a windmill in limestone, and from wells in the Rio Peñasco
flood plain. East of the range front, as far as site 2-10, an unconfined aquifer (the principal aquifer
of [11]) is present near the boundary of the Yeso and San Andres formations. Recharge to these strata
may occur near the range front. Samples are from domestic and agricultural wells up to 260 m deep,
with static water levels (SWLs) near 190 m below the surface. East of site 2-10, beneath a broad
plain west of the Pecos River, two major aquifers were sampled. An unconfined aquifer with SWLs
from 30 to 60 m below the surface exists in flood-plain sediments near Artesia. The eastward
continuation of the principal aquifer, 275 to 300 m below the surface, is confined beneath the
Queen-Grayburg anhydrite (Figure 2). It was artesian at the time of first exploitation; SWLs at
present range from 20 to 60m below the surface. Surface water samples were collected from the
Peñasco and Pecos Rivers.
From Peñasco to Hope, groundwater flow is east-southeast (Figure 18 of [9]). Allowing for
variation due to pumping, SWLs in the principal aquifer east of Hope are close to 1000 m.a.s.l.
(Table 1). Southward flow is likely in this area.
6.1. O and H Isotopes
All groundwater and surface water samples plot close to the SMT (Figure 7A). Most data cluster
at the intersection of the SMT with the GMWL, where the separation of the lines is less than 2‰ in
įD, and therefore impossible to resolve within measurement error. Data for the principal aquifer from
Mayhill to Hope [9] match the present data set in įD but include lower values of į18O. Two
groundwater samples from near Artesia (2-13, 2-19) plot to the right of the main data cluster. Surface
water from the Pecos River in the reaches between Artesia and Red Bluff ([15] for 1984–1987, [14]
for 2005, and data from this study) largely plot as a linear trend, close to an extrapolated SMT
(Figure 7A,B).
At the range front, groundwater from limestone (sites 2-1, 2-2) is distinct in į18O from
groundwater and surface water in the Rio Peñasco flood plain (sites 2-3, 2-4) (Figure 7, inset). These
two groups of data bracket the į18O range of the principal aquifer to the east. The unconfined aquifer
at Artesia has į18O values >–8.1‰, higher than for the principal aquifer; two of the samples
(2-15, 2-17) may plot on the GMWL, while one other (2-19) plots on the SMT. One principal aquifer
sample (2-10) plots above the GMWL. Three outlying samples (2-24, 25 and 26, locations in
Figure 1) from the principal (artesian) and shallow aquifers near Roswell plot on a trend similar to,
but slightly above, the SMT.
Previous data for the principal aquifer [13] pre-date automated isotope methods, and partially
overlap the main data cluster from the present study. The two data sets correspond in į18O, but the
older įD data have a spread >20‰, apparently spurious, and appear not to be useful. Weighted
precipitation averages from [12] are for į18O alone, and have been plotted on the GMWL in
Figure 7.
110
6.2. Tritium
In 1977-1978, when bomb tritium averaged about 35 TU in local precipitation, surface water and
alluvial groundwater from the Peñasco River flood plain contained about 10 TU, and tritium in the
principal aquifer near Artesia was below detection [13]. In samples collected for this study, tritium
is present at low levels (<2 TU) in groundwater from near the range front (sites 2-2, 2-3, 2-5), at site
2-19 in the shallow aquifer, and in one deep aquifer sample (site 2-21, 0.5 TU); at other sites it is
below detection (Figure 8). Groundwater from the alluvial aquifer beneath the Peñasco River
(site 2-3) contains 1.5 TU, distinctly lower than the average for precipitation, and consistent (cf. [13]),
with a large pre-bomb groundwater contribution to the surface water of the river. Reference [9] listed
tritium contents <2 TU in groundwater between Elk and Hope.
Figure 7. (A) Plot of į18O vs. įD for groundwater and surface water samples from study
area 2. (A) All data from this study. The field of data from [13] is for the principal aquifer
from Artesia to Roswell, and encompasses all but three outlying data points. The inset
shows a magnified view of clustered data. Site numbers (e.g., 1) correspond to entries in
Table 1, where the corresponding number is 2-1; (B) Plot of į18O vs. įD for the Pecos
River between Artesia and Red Bluff, from other studies [14,15], relative to the SMT.
-30
Ave. precipitation, Elk and
Roswell, [12]
A
GMWL
SMT
Principal
aquifer, [13]
-40
24
13
19
25
įD ‰
-50
Pecos R.
26
-50
Penasco R.
-60
15
17
-55
6
Pecos slope
aquifer, [9]
-70
10
1, 2
-60
4 3
-8.5 -8.3 -8.1 -7.9 -7.7
-80
-11
-10
-9
-8
-7
į18O
-6
-5
-4
-3
-2
‰
High mountains
Range front
Penasco to Hope
Artesia deep
Artesia shallow
Surface water
Roswell
10
B
GMWL
ɷD, ‰
-10
SMT
-30
-50
-70
-10
-8
-6
-4
ɷ18O,
Winter + Spring
‰
Summer + Autumn
-2
0
111
Figure 8. A. East-west profile of Area 2 (refer to Figure 2 for location) showing well
depths and static water levels (SWL) in relation to the surface. “Shallow” refers to the
shallow aquifer at Artesia, and “deep” to the deeper artesian aquifer. Also shown are
measurements of carbon-14 (pMC) and tritium (TU). Tritium below detection is
indicated as “bd”.
1800
W Peñasco
Elevation, m.a.s.l.
1600
Artesia E
Hope
75, 1.8
1400
36
58, 1.0
1200
shallow 1.3
deep 34, 0.5
bd
1000
Data shown thus:
pMC, TU
800
deep bd
29, bd
30, bd
600
0
20
40
60
80
100
km from origin
sample site
Surface
Well depth
SWL
6.3. Other Parameters
14
C is higher (78 and 55 pMC, corrected to 115 and 82 pMC) in two samples (2-2, 2-5) from near
the range front, than in four samples between Hope and Artesia, (29–36 pMC, corrected to 50–70
pMC (Figure 8). Reference [9] gave 40–50 pMC in most groundwater between Elk and Hope. Values
of į34S are 12‰ to 14‰.
6.4. Interpretation
Groundwater in the Pecos Slope and artesian aquifers is largely uniform in isotope content over
an east-west extent of about 100 km, and lies on or close to the SMT. A dominant water source in
the high Sacramento Mountains is therefore likely. East of Peñasco, a few samples (2-13, 2-19, 2-24
and 2-25) have isotope data plotting on the SMT, but to the right of the main data cluster; these may
reflect local recharge of evaporated surface water. Site 2-13 is close to the ephemeral lower reach of
the Rio Peñasco, where recharge of evaporated surface water may occur. The other three samples are
from the shallow aquifer beneath irrigated fields, where reflux of evaporated irrigation water is
probable. Addition of local rainwater is likely for site 2-10 (Figure 7).
Bulk groundwater residence times (the corrected versions of the data shown in Figure 8) in the
principal aquifer are 1300 to 5600 years east of Hope, greater than those suggested in [13].
Most of the į18O and įD data for the Pecos River between Artesia and Red Bluff plot on a linear
trend close to an extrapolated SMT, regardless of season (Figure 7B), and can be therefore be
112
generated as a result of evaporation of water like that in the principal aquifer at Artesia. The principal
source of river water in this area is therefore most likely the Sacramento Mountains, either by natural
recharge from the aquifer, or by way of irrigation on the Pecos flood plain. If this is true,
mountain-derived water is discharged with an isotope signature of evaporation into the river near
Roswell and Artesia. This suggests a modification to the modeling, based on deuterium excess, of river
water sources in reference [14].
7. Discussion
7.1. Water Sources in Study Area
Most water sampled for this study plots on the Sacramento Mountains trend (SMT) in įD–į18O
space. The SMT originates in high-altitude winter precipitation. Such precipitation is therefore the
principal and ultimate source of groundwater in the area between Alamogordo and the Pecos River.
Most water in the Pecos River near Artesia also appears to be of that origin. There is scant evidence
for recharge of local meteoric water at low altitudes. The few exceptions include groundwater from
the southeastern part of Tularosa Valley (where ancient water from high elevations appears to be
present), and some unconfined-aquifer samples from Artesia (where local recharge probably occurs).
7.2. Seasonality of Recharge
The heavy monsoon rains of 2006 and 2008 generated recharge of distinctive įD–į18O signature
in groundwater of the high Sacramento Mountains, but in drier years, 2007 and 2009, groundwater
isotopes shifted towards the SMT [9]. Monsoon rainfall comparable to that in 2006 had not occurred
since 1997, and in the 2003 sampling for this study, winter recharge, plotting on the SMT as a result
of local evaporation prior to infiltration, was predominant. Where old groundwater is present (south
of Alamogordo and in the principal aquifer of Roswell Basin), įD and į18O conform largely to the
SMT. In the long term, therefore, recharge in dry to average years contributes the larger volume to
low-elevation aquifers around the Sacramento Mountains. Years with unusually wet summers lead to
a transient (a few years) response in the high-altitude aquifers, but make little contribution to the old
groundwater in basins at the foot of the mountains.
The Sacramento Mountains are therefore an unusual example of a mountain block in which the
dominant season of recharge can change in response to seasonal precipitation amounts, although
winter precipitation, only 35% of total precipitation on average, predominates in the long-term.
Winter recharge is considered predominant in a number of other mountain ranges in the arid western
USA. In the Spring Mountains, Nevada, another carbonate-rock range, winter precipitation is
dominant; summer rain contributes about 30% of annual precipitation, but only about 10% of
recharge [23]. Winter recharge also predominates in the Huachuca Mountains, Arizona, where
summer precipitation contributes 54% of the annual total on average, but winter precipitation
accounts for 65% ± 25% of recharge [24]. In the Santa Catalina Mountains, Arizona [25], and the
ranges delimiting the Verde River watershed, Arizona [26], winter recharge is considered
predominant, contributing 98% of recharge in the latter case.
113
7.3. Sacramento Mountain Carbonate Strata as a Karst Aquifer
A continuum of aquifers exists in carbonate rock [27]. At one extreme, carbonate dissolution leads
to wide solution cavities that self-organize into dendritic drainage networks discharging through
large springs; water flow rates are commonly 102 to 104 m/day over distances of 103 to 104 m. At the
other extreme, solution cavities are narrow and of limited interconnection, generating an aquifer with
lower flow rates and discharge through many small springs. Although small-scale collapse structures
are recognized [9], cavern networks are not developed in the thinly bedded strata, some impermeable,
of the study area. A flow rate of 10 m/day over 30 km between the range crest and the eastern range
front would result in water travel times of about 8 years. The tritium and 14C data imply residence
times >60 years in the mountain aquifers several cases, while surface water in the Rio Peñasco and
associated flood-plain groundwater also contain some pre-bomb precipitation. The isotope evidence
indicates widespread persistence of pre-bomb precipitation in groundwater, and flow rates typically
much lower than 10 m/day. The Sacramento Mountains therefore fall at the latter end of the
continuum of karst aquifers as described above.
Nonetheless, the carbonate strata in and east of the Sacramento Mountains compose a regional
aquifer system over a distance of 130 km. Regional carbonate aquifers of similar extent have been
demonstrated elsewhere in the region on the basis of geochemical modeling, east of the Salt
Basin in West Texas [28], and in the region southwest of the Cuatrocienegas Basin of Coahuila,
Mexico [29].
7.4. Mode of Mountain-Front Recharge to Basin Alluvium
The location of mountain-front recharge relative to the interface between hard-rock mountain
blocks and basin alluvium in the southwest USA has been addressed in several studies. In the middle
Rio Grande Basin (New Mexico), infiltration is thought to occur in a narrow zone along the range
front of the Sandia and Manzano Mountains [30]. In Chino Valley (Arizona) [26] and Tucson Basin
(Arizona) [1], evidence indicates infiltration from stream beds downstream of the mountain fronts,
at distances of 6 to 10 km in the case of Tucson Basin. Groundwater isotope data also indicate
recharge of ponded surface water in the center of the Hueco Bolson (Texas) [31]. The present study
concurs with the possibility of infiltration as far as 15 km downstream of the range front.
7.5. Source of Hueco Bolson Groundwater
The question addressed here is the source of saline water in the center of the Hueco Bolson,
an alluvial basin near El Paso, Texas, 100–130 km southwest of Alamogordo (Figure 1). Subsurface
movement of groundwater from the Tularosa Valley to the Hueco Bolson is physically possible
according to piezometric data [4,6]. An alternative source is recharge from the Organ and Franklin
Mountains (Figure 1), which supply a freshwater aquifer, the Franklin Mountains freshwater lens
(FMFWL) in ancient fluvial deposits at the western edge of the Hueco Bolson [6]. The catchment
for the FMFWL is largely at altitudes between 1300 and 2400 m.a.s.l. in the Organ Mountains, in
contrast to a catchment at 2400–2800 m.a.s.l. for the four large canyons that focus fresh water from
114
the Sacramento Mountains into basin sediment near Alamogordo. H and O isotopes might therefore
discriminate between the two sources, as discussed inconclusively in [31].
Groundwater from the FMFWL plots along the global meteoric water line with į18O values
between –9‰ and –11‰ (Figure 9). The upper end of the data array corresponds to groundwater of
short residence time, while the lower end corresponds largely to groundwater resident for thousands
of years [2]. The SMT and the suggested paleo-SMT (based on samples 1-39 and 1-40) intersect
the FMFWL trend near –9‰ and –11‰, respectively. On the one hand, the į18O and. įD values of
the saline, evaporated water in the center of the Hueco Bolson plot between the SMT and the
paleo-SMT, and could therefore represent mixtures of older and younger water from the Sacramento
Mountains. On the other hand, į18O and. įD values define an evaporation trend that could originate
in older FMFWL water, so that the water could have originated in the Frankin and Organ Mountains,
perhaps as surface water ponded and evaporated in the basin center at a time of cooler, wetter climate.
The stable isotopes fail to distinguish the two possibilities because of the likely presence of
ancient groundwater.
Figure 9. Plot of į18O vs. įD showing: (a) The Sacramento Mountains Trend (SMT, as
in Figure 4) and data for groundwater at Alamogordo; (b) Data for groundwater in the
Hueco Bolson in Texas, from [2] and [24], distinguished according to salinity (HB-saline
at the basin center, and HB-fresh from the Franklin Mountains fresh water lens on the
western side of the basin); (c) A suggested paleo-SMT based on two samples of ancient
water in the southeastern part of Tularosa Valley.
-45
GMWL
-50
įD, ‰
-55
SMT
-60
-65
-70
1-39
paleo-SMT
1-40
-75
-80
-11
-10
-9
-8
-7
į18O‰
HB saline
HB fresh
SE Tularosa ancient water
7.6. Implications for Groundwater Management
High-elevation winter recharge is the principal source of groundwater over the long term in the
aquifers of the Sacramento Mountains and the flanking basins. If winter precipitation declines, for
instance in response to global climate change, groundwater supply will decrease. The effect would
be felt initially in the high mountain communities such as Weed and Cloudcroft (but might be
115
mitigated if occasional wet summers persist) and in areas from La Luz Canyon to Alamogordo and
Tularosa where groundwater is of post-bomb age (Table 1). In the Roswell basin, where artesian
water has been resident for thousands of years, there would be no short-term effect of diminished
winter recharge; over-pumping for irrigation would be of more immediate concern.
8. Conclusions
Stable O and H isotopes have proved useful as environmental tracers in determining the
relationships among various occurrences of groundwater in the study area, and the seasonality of
recharge. Tritium and 14C have provided valuable constraints on groundwater residence times.
A. Relationship between groundwater in the Sacramento Mountains and in flanking basins.
Groundwater sampled from the high Sacramento Mountains in 2003 has a characteristic isotope
signature. On a į18O vs. įD diagram, it plots on an evaporation trend (the Sacramento Mountains
trend, SMT) of slope near 5.6. Recharge in subsequent years of high summer precipitation plots
above the SMT, and has the isotope signature of summer rain [9]. The SMT signature is found in
carbonate and basin-fill aquifers west and east of the Sacramento Mountains, indicating winter
precipitation in the high mountains as the principal long-term source of groundwater in those basins.
Water derived from high elevations is supplied to aquifers at lower elevations by a combination of
flow through the carbonate aquifers, and mountain-front recharge of surface water showing the isotope
effect of evaporation.
Recharge of local low-altitude meteoric water and irrigation reflux occurs in the shallow aquifer
at Artesia. The principal (artesian) aquifer of the Roswell Basin receives little or no recharge east of
the range front (near Peñasco).
B. Groundwater residence times. Short residence times, a few decades, are characteristic of the
high Sacramento Mountains (cf. [9]). Bulk residence times for groundwater near Hope and Artesia
range from 1000 to 5000 years. In Tularosa Valley, bulk residence times are a few decades near
Alamogordo and Tularosa, hundreds to thousands of years immediately south of Alamogordo, and
up to 20,000 years at distances of 50 or more km south of Alamogordo. The oldest water has į18O
and įD values lower than those on the SMT.
C. Recharge seasonality. Groundwater plotting on the SMT is the result of winter recharge.
However, both winter recharge and summer recharge can occur in the carbonate rock of the high
Sacramento Mountains. Summer recharge is contributes greatly to mountain groundwater during
years of unusually high monsoon rainfall [9], but winter recharge is predominant at other times.
D. Origin of waters more distant from the mountains. Surface water in the Pecos River between
Artesia and Red Bluff has isotope compositions consistent with a predominant origin in the principal
artesian aquifer of Roswell Basin. Groundwater of the central area of the Hueco Bolson near El Paso,
Texas, may have originated from the Sacramento Mountains or from the Organ and Franklin
Mountains. Stable H and O isotopes cannot distinguish the two sources.
116
Acknowledgements
Richard Warnock, Russell Wright and Elaine Wright of the Sacramento Mountains Watershed
Restoration Corporation, and Monroe Curtis of Otero County introduced us to numerous well-owners
who kindly permitted us to take samples. Bob Mayberry, A.J. Posey, Elaine Wright, Bobbie Melton
and Richard Warnock accompanied us in the field. Lynwood Hume kindly provided river and
groundwater samples from Roswell. The authors gratefully acknowledge the contributions of three
reviewers, whose comments greatly improved the article. The study was funded though SAHRA
(Sustainability of semi-Arid Hydrology and Riparian Areas) under the STC Program of the National
Science Foundation, Agreement No. EAR-9876800, except for Study Area 2, where work was
funded by the Environmental Isotope Laboratory at the University of Arizona.
Conflicts of Interest
The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
Eastoe, C.J.; Gu, A.; Long, A. The Origins, Ages and Flow Paths of Groundwater in Tucson
Basin: Results of a Study of Multiple Isotope Systems. In Groundwater Recharge in a Desert
Environment: The Southwestern United States, Water Science and Applications Series;
Hogan, J.F., Phillips, F.M., Scanlon, B.R., Eds.; American Geophysical Union: Washington,
DC, USA, 2004; Volume 9, pp. 217–234.
Eastoe, C.J.; Hibbs, B.J.; Olivas, A.G.; Hogan, J.; Hawley, J.; Hutchison, W.R. Isotopes in the
Hueco Bolson Aquifer, Texas (USA) and Chihuahua (Mexico): Local and general implications
for recharge sources in alluvial basins. Hydrogeol. J. 2008, 16, 737–747.
Eastoe, C.J.; Hutchison, W.R.; Hibbs, B.J.; Hawley, J.; Hogan, J.F. Interaction of a river
with an alluvial basin aquifer: Stable isotopes, salinity and water budgets. J. Hydrol. 2010, 395,
67–78.
Meinzer, O.E.; Hare, F.R. Geology and Water Resources of Tularosa Basin, New Mexico;
U.S. Geological Survey Water-Supply Paper 343; U.S. Government Printing Office:
Washington, DC, USA, 1915; p. 317.
Huff, G.F. Simulation of Ground-Water Flow in the Basin-Fill Aquifer of the Tularosa Basin,
South-Central New Mexico, Predevelopment through 2040; Scientific Investigations Report;
U.S. Geological Survey: Reston, VA, USA, 2005; pp. 2004–5197.
Hibbs, B.J.; Boghici, R.N.; Hayes, M.E.; Ashworth, J.B.; Hanson, A.N.; Samani, Z.A.;
Kennedy, J.F.; Creel, R.J. Transboundary Aquifers of the El Paso/Ciudad Juarez/Las Cruces
Region; Contract Report; Texas Water Development Board, Austin & New Mexico Water
Resources Research Institute: Las Cruces, NM, USA, 1997; p. 148.
Mayer, J.R.; Sharp, J.M., Jr. Fracture control of regional groundwater flow in a carbonate aquifer
in a semi-arid region. Geol. Soc. Am. Bull. 1998, 110, 1657–1671.
117
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
National Oceanic and Atmospheric Administration. Climatography of the United States No. 81,
Monthly Normals of Temperature, Precipitation and Heating and Cooling Degree Days,
1971–2000. Available online: http://cdo.ncdc.noaa.gov/climatenormals/clim81/NMnorm.pdf
(accessed on 15 January 2013).
Newton, T.B.; Rawling, G.C.; Timmons, S.S.; Land, L.; Johnson, P.S.; Kludt, T.J.; Timmons,
J.M. Sacramento Mountains Hydrogeology Study; New Mexico Bureau of Geology and Mineral
Resources Open-File Report 543; New Mexico Bureau of Geology and Mineral Resources:
Socorro, NM, USA, 2012; p. 78.
Roswell Geological Society; Stratigraphic Research Committee. West-East Correlation Section,
San Andres Mts. to N.Mex.-Texas Line, Southeastern New Mexico; Roswell Geological Society:
Roswell, NM, USA, 1956.
Havenor, K.C. The Hydrogeologic Framework of the Roswell Groundwater Basin, Chaves,
Eddy, Lincoln, and Otero Counties, New Mexico. Ph.D. Thesis, University of Arizona, Tucson,
AZ, USA, 1996.
Gross, G.W.; Hoy, R.N.; Duffy, C.J.; Rehfeldt, K.R. Isotope Studies of Recharge in Roswell
Basin. In Isotope Studies in Hydrologic Processes; Perry, E.C., Jr., Montgomery, C.W., Eds.;
Northern Illinois University Press: DeKalb, IL, USA, 1982; pp. 25–33.
Hoy, R.N.; Gross, G.W. A Baseline Study of Oxygen 18 and Deuterium in the Roswell, New
Mexico, Groundwater Basin; New Mexico Water Resources Research Institute: Las Cruces,
NM, USA, 1982; Volume 144, p. 98.
Yuan, F.; Miyamoto, S. Characteristics of oxygen-18 and deuterium composition in waters from
the Pecos River in American Southwest. Chem. Geol. 2008, 255, 220–230.
Coplen, T.B.; Kendall, C. Stable Oxygen and Hydrogen Isotope Ratios for Selected Sites of the
U.S. Geological Survey’s NASQAN and Benchmark Surface Water Networks; U.S. Geological
Survey Open-File Report 00-160; U.S. Geological Survey: Reston, VA, USA, 2000; p. 409.
Szynkiewicz, A.; Newton, B.T.; Timmons, S.S.; Borrok, D.M. The sources and budget for
dissolved sulfate in a fractured carbonate aquifer, Southern Sacramento Mountains, New
Mexico, USA. Appl. Geochem. 2012, 27, 1451–1462.
Huff, G.F. Apparent Age of Ground Water Near the Southeastern Margin of the Tularosa Basin,
Otero County, New Mexico. In Geology of White Sands: New Mexico Geological Society
Guidebook; Giles, K.A., Lueth, V.W., Lucas, S.G., Kues, B.S., Myers, R., Ulmer-Scholle, D.S.,
Eds.; New Mexico Geological Society: Alamogordo, NM, USA, 2002; Volume 53,
pp. 303–307.
Clark, I.; Fritz, P. Environmental Isotopes in Hydrogeology; Lewis Publishers: Boca Raton, FL,
USA, 1997; p. 328
Veizer, J.; Hoefs, J. The nature of 18O/16O and 13C/12C secular trends in sedimentary carbonate
rocks. Geochim. Cosmochim. Acta 1976, 40, 1387–1395.
Colgan, R.E.; Scholle, P.A. Carbon isotopic stratigraphy of the San Andres Formation; A
possible correlation tool? (Abstract). Am. Assoc. Pet. Geologists Bull. 1991, 75, 555.
118
21. Wright, W.E. įD and į18O in Mixed Conifer Systems in the U.S. Southwest: The Potential of
į18O in Pinus Ponderosa Tree Rings as a Natural Environmental Recorder. Ph.D. Thesis,
Department of Geosciences, The University of Arizona, Tucson, AZ, USA, 2001.
22. Eastoe, C.J.; Watts, C.J.; Ploughe, M.; Wright, W.E. Future use of tritium in mapping pre-bomb
groundwater volumes. Ground Water 2011, 50, 87–93.
23. Winograd, I.J.; Riggs, A.C.; Coplen, T.B. The relative contributions of summer and cool-season
precipitation to groundwater recharge, Spring Mountains, Nevada, USA. Hydrogeol. J. 1998, 6,
77–93.
24. Wahi, A.K.; Hogan, J.F.; Ekwurzel, B.; Baillie, M.N.; Eastoe, C.J. Geochemical quantification
of semiarid mountain recharge. Ground Water 2008, 46, 414–425.
25. Cunningham, E.E.B.; Long, A.; Eastoe, C.J.; Bassett, R.L. Migration of recharge waters
downgradient from the Santa Catalina Mountains into the Tucson Basin aquifer. Hydrogeol. J.
1998, 6, 94–103.
26. Blasch, K.W.; Bryson, J.R. Distinguishing sources of ground water recharge by using į2H and
į18O. Ground Water 2007, 45, 294–308.
27. Worthington, S.R.H.; Ford, D.C. Self-organized permeability in carbonate aquifers. Ground
Water 2009, 47, 326–336.
28. Uliana, M.M.; Sharp, J.M. Tracing regional flow paths to major springs in Trans-Pecos Texas
using geochemical data and geochemical models. Chem. Geol. 2001, 179, 53–72.
29. Wolaver, B.D.; Sharp, J.M.; Rodriguez, J.M.; Ibarra, F. Delineation of regional karstic aquifers:
An integrative data approach. Ground Water 2008, 46, 396–413.
30. Plummer, L.N.; Bexfield, L.M.; Anderholm, S.K.; Sanford, W.E.; Busenberg, E. Geochemical
Characterization of Ground-Water Flow in the Santa Fe Group Aquifer System, Middle Rio
Grande Basin, New Mexico; U.S.G.S. Water-Resources Investigation Report 03-4131;
U.S. Geological Survey: Reston, VA, USA, 2004; 1–395.
31. Druhan, J.L.; Hogan, J.F.; Eastoe, C.J.; Hibbs, B.; Hutchison, W.R. Hydrogeologic controls on
groundwater recharge and salinization: A geochemical analysis of the Northern Hueco Bolson
Aquifer, El Paso, Texas. Hydrogeol. J. 2008, 16, 281–296.
119
A Combined Radio- and Stable-Isotopic Study of a California
Coastal Aquifer System
Peter W. Swarzenski, Mark Baskaran, Robert J. Rosenbauer, Brian D. Edwards
and Michael Land
Abstract: Stable and radioactive tracers were utilized in concert to characterize geochemical
processes in a complex coastal groundwater system and to provide constraints on the kinetics of
rock/water interactions. Groundwater samples from wells within the Dominguez Gap region of Los
Angeles County, California were analyzed for a suite of major cations (Na+, K+, Mg2+, Ca2+) and
anions (Clí, SO42í), silica, alkalinity, select trace elements (Ba, B, Sr), dissolved oxygen, stable
isotopes of hydrogen (įD), oxygen (į18O), dissolved inorganic carbon (į13CDIC), and radioactive
isotopes (3H, 222Rn and 223,224,226,228Ra). In the study area, groundwater may consist of a complex
mixture of native groundwater, intruded seawater, non-native injected water, and oil-field brine
water. In some wells, Clí concentrations attained seawater-like values and in conjunction with
isotopically heavier į18O values, these tracers provide information on the extent of seawater intrusion
and/or mixing with oil-field brines. Groundwater 3H above 1 tritium unit (TU) was observed only in
a few select wells close to the Dominguez Gap area and most other well groundwater was aged
pre-1952. Based on an initial 14C value for the study site of 90 percent modern carbon (pmc),
groundwater age estimates likely extend beyond 20 kyr before present and confirm deep circulation
of some native groundwater through multiple aquifers. Enriched values of groundwater į13CDIC
in the absence of SO42í imply enhanced anaerobic microbial methanogenesis. While secular
equilibrium was observed for 234U/238U (activity ratios ~1) in host matrices, strong isotopic fractionation
in these groundwater samples can be used to obtain information of adsorption/desorption kinetics.
Calculated Ra residence times are short, and the associated desorption rate constant is about three
orders of magnitude slower than that of the adsorption rate constant. Combined stable- and
radio-isotopic results provide unique insights into aquifer characteristics, such as geochemical
cycling, rock/water interactions, and subsurface transport and mixing.
Reprinted from Water. Cite as: Swarzenski, P.W.; Baskaran, M.; Rosenbauer, R.J.; Edwards, B.D.;
Land, M. A Combined Radio- and Stable-Isotopic Study of a California Coastal Aquifer System.
Water 2013, 5, 480-504.
1. Introduction
Widespread demands on the groundwater resources of Los Angeles County, California during the
early 20th century have resulted in substantial groundwater-level declines, as well as associated
coastal seawater intrusion and deteriorating water quality [1,2]. In an effort to stave off saltwater
intrusion in the 1950s to the early 1970s, three series of injection wells were installed along the coast
where injected water could create artificial hydraulic barriers, named West Coast Basin, Dominguez
Gap, and Alamitos Gap barrier projects [3,4]. Over the past decade, ~26 to 37 million m3 of water
are annually injected into these three barriers and while seawater intrusion has been generally
120
reduced, at Dominguez Gap saltwater intrusion is still occurring [4]. This inefficiency in halting the
seawater intrusion is due in part to an incomplete characterization of the stratigraphic architecture
that partially controls the groundwater flow in this region. In an effort to better understand the
geochemical character of this regional groundwater and to better predict future groundwater quality
change, a study incorporating stable isotopes was initiated [4–6] and complemented with select U/Th
radionuclide work.
Geochemical and isotopic tracers have been widely used to investigate rock/water interactions,
recharge rates of meteoric water, evaporation effects, and groundwater transport phenomena [7,8].
For example, stoichiometric ratios of various cations and anions may provide insight into the
chemical weathering rates of source minerals [9]. Oxygen isotopes in groundwater have been used
to identify source water and to assess evaporation kinetics [10,11]. Furthermore, a fingerprint of
historic fluctuations in water vapor and air mass trajectories can be preserved in the isotopic
composition of meteoric groundwater [12,13]. The presence of 3H, with a half-life of 12.3 years, in
groundwater can provide information on recharge rates and vertical flow velocities [14]. Stable
isotopic composition of carbon in the dissolved inorganic matter in groundwater may yield
information on the carbonate equilibrium, infiltration of atmospheric CO2, as well as the microbial
degradation of organic matter. In principal, the 14C concentration in groundwater can yield
information on age data from a few thousand years to 45,000 years that can be used to help
constrain groundwater flow velocities and direction, recharge rates, hydraulic conductivities, and
effective porosity.
Select U/Th series radionuclides have been utilized to characterize an aquifer’s physicochemical
properties, such as adsorption-desorption rate constants and aquifer retardation factors [15–21].
Because the geochemical behavior of many contaminants of interest is quite similar to the
geochemistry of select members of the U- and Th-series radionuclides, these nuclides can provide
unique information on the rates of some of these subsurface processes [22]. The movement of a
dissolved species in groundwater can be retarded by several processes, such as ion exchange,
adsorption, diffusion into blind pores, chemical precipitation, and membrane filtration [15,23,24].
The retardation of these species depends on particular aquifer characteristics (e.g., lithology, water
chemistry, residence time), but parameters that control the retardation factor and absorption-desorption
rate constants are not fully understood [17–20,25]. In particular, the dependence of these parameters
on aquifer characteristics that are complexly mixed with diverse source waters, warrants
further study.
121
To apply our combined radio- and stable-isotope approach, a suite of groundwater samples from
the Dominguez Gap region of the southwest Los Angeles Basin, California were collected and
analyzed for major and minor ions, select trace elements (Ba, B, and Sr) and stable isotopes (į18O,
įD, į13CDIC), tritium (3H) and a suite of U- and Th-series radionuclides (223Ra, 224Ra, 226Ra, 228Ra,
and 222Rn). The study area is tectonically active [26–29], and as a result, uplift and erosion have
winnowed many of the fine-grained confining units that serve to protect the underlying aquifers from
seawater intrusion. As a consequence, there exists the potential for enhanced vertical and horizontal
migration of seawater into the producing aquifers and subsequent landward migration of intruded
waters beneath the Dominguez Gap Barrier Project [4,30,31]. The interaction between seawater and
the aquifer system makes the sorption characteristics study of this aquifer of interest due to its
relevance to other tectonically-active or structurally complex coastal areas.
2. Geographic Setting
Natural and artificial recharge to the Los Angeles Basin occurs through local precipitation and
infiltration, seawater intrusion along the coast, injection of non-native water into the barrier wells,
and regional groundwater flow from adjacent basins. Each water-mass end member is geochemically
distinct and has been identified with a suite of isotopes and geochemical parameters [6]. Groundwater
samples for this study were collected from the Dominguez Gap region of the southwest Los Angeles
Basin, located along the coastal plain of Los Angeles County adjacent to San Pedro Bay (Figure 1).
The Dominguez Gap denotes a hydrologic “gap” that occurs just south of the Dominguez Hills,
where the Los Angeles River transverses the Newport-Inglewood Uplift. In general, exposed Late
Pleistocene alluvial deposits cover large parts of the coastal plain except close to the Los Angeles
River, where as much as ~10 m of Holocene-aged fluvial and marine sediment has filled in parts of
the paleo-river channel of the Los Angeles River during a lower sea level stand. The hydrogeology
of the Dominguez Gap region has been well studied [26,27,32] and is summarized in a
shore-perpendicular cross-section (Figure 2). Injection of fresh water in the West Coast Basin
and Dominguez Gap Barrier Projects is a significant source of recharge to the West Coast
Groundwater Basin [6]. Ponti et al. [30] developed a sequence stratigraphic model of the Dominguez
Gap area that further refined these water-bearing depositional systems relative to sediment supply,
sea level, and accommodation space. Their work also identified the Pacific Coast Highway (PCH)
Fault, which may play an important role in the mixing of seawater into deep aquifers.
122
Figure 1. Base map of the well sites and the prominent geologic and hydrologic features
of the study area, Los Angeles County, California. The Huntington Park monitoring well
site is located in the adjacent Central Basin, 6.5 km south of downtown Los Angeles, and
in an area known as the Los Angeles forebay. The forebay is an area of groundwater
recharge for some of the water contained in the West Coast Basin. Note the following
abbreviations: LWEB = Webster; LBCH = Cabrillo; LBPC = Pier C; LBPF = Pier F are
used throughout the text, figures, and tables.
123
Figure 2. An idealized hydrogeologic cross-section of the Dominguez Gap area, also
showing the sequence boundaries as defined by Ponti et al. [30]. Geographic location of
the A–A' transect is shown in Figure 1.
3. Groundwater Radionuclide Transport
The fate of U/Th series radionuclides in groundwater continues to be an active area of research
primarily to better understand and predict subsurface contaminant actinide transport [21]. Many of
the early advances in modeling naturally-occurring radionuclides in groundwater were pioneered by
Rama and Moore [25] and Krishnaswami et al. [15] and later summarized by Ku et al. [20] and
Porcelli and Swarzenski [21]. The following model stems largely from the original Krishnaswami
et al. [15] formulations. The dominant processes that can control the fate of U/Th series radionuclides
in groundwater include both input terms, such as, (1) recoil mechanisms, (2) congruent dissolution
within an aquifer solid, (3) desorption reactions from solid surfaces, and (4) in-situ radioactive decay
of a dissolved parent nuclide, as well as removal terms, such as (1) chemical precipitation, (2)
radioactive decay, and (3) reversible sorption onto particle surfaces. If we assume the kinetics of
adsorption and desorption to be first order [20,21], then the steady-state mass balance reactions for
a radionuclide in an aquifer can be reduced to the following equations [15]:
Dissolved phase: P + k2Ns = ȜNd + k1Nd
(1)
Solid phase: k1Nd = ȜNs + k2Ns
(2)
Here P (atoms per second per volume of water) defines the production rate of a nuclide in solution
by such processes as chemical dissolution (i.e., weathering), in-situ production and recoil, Ȝ is the
decay constant (0.693/t1/2) of a radionuclide, k1 and k2 are first-order adsorption and desorption rate
constants, respectively, and Nd and Ns describe the respective concentration of a nuclide in water
(atoms per volume of water) and adsorbed onto an aquifer matrix. The recoil term is the dominant
124
supply term for short-lived radionuclides (i.e., 224Ra, 223Ra, 228Ra, with a mean-life < 10 years), although
there could be some contribution from congruent weathering for 226Ra (mean-life = 2309 years).
The ratio (ȍ) of the activity of a nuclide (= ȜNd) to its production (P) in solution can be calculated
from Equations (1) and (2), as follows:
ȍ = ȜNd/P = (k2 + Ȝ)/(k1 + k2 + Ȝ)
(3)
Assuming negligible isotopic fractionation [21], the adsorption (k1) and desorption (k2) rate
constants are expected to be the same for two isotopes of the same element. If we assume “i” and “j”
to describe two isotopes of one element, then the mass balance equations of each of these nuclides
can be combined and solved for k1 and k2 [15] as follows:
k1 = [(Ȝi í Ȝj) (1 í ȍi) (1 í ȍj)]/(ȍi í ȍj)
(4)
k2 = [ȍiȍj (Ȝj í Ȝi) + Ȝiȍj í Ȝjȍi)]/(ȍi í ȍj)
(5)
and
From the measured groundwater 222Rn, 224Ra, and 228Ra activities and their radiogenic parents,
230
Th, 228Th, and 232Th activities, in the solid phase, we can determine k1 and k2 using Equations (4)
and (5).
4. Materials and Methods
Most samples for this effort were collected as part of a larger U.S. Geological Survey (USGS)
project on the groundwater quality and geochemical character underlying Los Angeles County and
thus more complete sampling protocols and analytical methods are described in detail therein [4–6].
Briefly, each well was sufficiently purged and then sampled using “clean” procedures to avoid
contamination as per standard USGS water quality sampling protocols. Chemically unstable
constituents, such as alkalinity and 222Rn, were processed and/or preserved in the field. Water quality
data including stable and radiogenic isotopes were determined at the USGS Water Quality
Laboratory in Denver, CO [5,6]. Stable isotopes were determined using isotope mass spectrometry
with a gas-source stable isotope mass spectrometer, as per methods described in Epstein and
Mayeda [33] and Coplen et al. [34]. The 2-sigma uncertainty of oxygen and hydrogen isotopic results
is 0.2‰ and 2‰, respectively.
Radon-222 activities were measured in the field using a commercially available Rn-in-air monitor
(RAD7—DURRIDGE, Inc., Billerica, MA, USA) coupled to a RAD-H2O discrete water sampling
kit [35–38]. Radium-223 and 224Ra activities were quantified using delayed-coincidence alpha
counting techniques [37–39]. Briefly, Ra was quantitatively removed from large groundwater
samples (50–100 L) using MnO2 fiber cartridges. The partially dried fiber was subsequently placed
into a closed, recirculating loop and a RaDeCC detector. The 223Ra and 224Ra isotopes were recounted
after ~20 days to correct for supported 224Ra activities (from 228Th), and subsequently decay-corrected
to the mid-point sampling time. Propagated errors for the delayed coincidence counters are typically
<10%. After the counting for short-lived Ra isotopes was completed, the fiber was leached with
a 6M HCl-H2O2-hydroxylamine hydrochloride mixture to quantitatively remove Ra from the Mn
fiber. The Ra was co-precipitated with BaSO4 using Ba(NO3)2-H2SO4 [40] and the BaSO4 precipitate
125
was counted after 20 days (allowing for the in-growth of 222Rn daughters) in a high-purity Ge well
detector coupled to an InSpector gamma spectrometry software package. The 226Ra and 228Ra
activities were quantified using gamma energies of 352 and 609 keV for 226Ra and 338 and 911 keV
for 228Ra.
Seven soil samples from well cuttings that represented a spectrum of geologic material from the
well sites were also analyzed for 238U, 234U, and 230Th using an inductively coupled plasma mass
spectrometer (ICP-MS). Briefly, ~200 mg of dried, pulverized sample was brought into solution
using HF and concentrated HNO3. Blanks and reference standards for radionuclides in sediment,
IAEA 385 (Irish Sea sediment) were prepared and analyzed as quality control measures. The
concentrations of 238U, 234U and 230Th were measured by ICP-MS in the single-collector mode.
5. Results and Discussion
5.1. Major Ion Composition
In addition to two well parameters (approximate horizontal flow path distance, x, and depth to the
top of the screened interval, z), concentrations of the major cations (Na+, K+, Ca2+ and Mg2+) and
anions (Clí, alkalinity (as CaCO3), and SO42í), dissolved oxygen (DO), as well as Ba, B, and Sr are
presented in Table 1. The DO concentration, which can be a useful proxy for oxidation effects during
sampling of reduced groundwater, ranged between <0.1 to 2.6 mg Lí1. Of the 30 samples that were
measured for DO, only one sample (LWEB-4) was slightly above a “hypoxic” condition. While most
chloride concentrations of native groundwater did not exceed 35 mg Lí1 in the Lower aquifer
systems, some wells close to the coast had historic Clí values as high as 90 mg Lí1. Table 2 lists
summary parameters and descriptions of well waters. Water levels of many of these near-shore wells
increased in response to sustained freshwater injection, yet a concomitant decrease in Clí values is
not always observed [41]. For example, elevated Clí concentrations have been measured in several
Upper and Lower aquifer system wells east of the Dominguez Gap Barrier Project; Long Beach 3
and Long Beach 4. In water from the wells, the Na+ concentration varied between 39 (Huntington
Park #1) and 10,800 mg Lí1, while the Clí concentration ranged from 18.5 to 19,900 mg Lí1
(seawater-like value observed at LBPF-2). Excluding LBPF-2 as groundwater here consist mostly of
seawater, a plot of Clí as a function of Na+ (Figure 3A) illustrates that many of the wells are variably
influenced by elevated Clí concentrations. There is an expected [41] strong positive correlation
(R2 = 0.88) between SO42í and Ca2+ concentrations (Figure 3B). Extensive SO42í reduction and cation
exchange reactions result in most native groundwater within the study area having a characteristic
Ca/Na-bicarbonate to Na-bicarbonate composition with very low Clí concentrations, <65 mg Lí1 [6,42].
Non-native water typically exhibits a dominant Ca/Na-sulfate composition, while wells intruded by
seawater or mixed with oil-field brines have a Na-Cl composition [6]. As many Tertiary brine fluids
are also defined by a high Na-Cl composition, it is not easy to separate these from seawater-intruded
waters (e.g., Wilmington-2 #2). See Table 2 for a summary of characteristic geochemical parameters
that define these well waters as well as recent trends in water quality.
126
Table 1. Select well characteristics and water quality data for wells sampled. Well
location for all but Huntington Park sites shown in Figure 1. The Huntington Park site is
located in a recharge area of the adjacent Central Basin, near downtown Los Angeles [6].
x1
z2
DO 3
Well ID
km
Huntington Park #1 (4/9/1997)
m
0.0 271
Huntington Park #2 (4/10/1997) 0.0 210
í1
Ca2+
í1
Mg2+
í1
mg L
mg L
mg L
0.2
59
14
K+
Na+
Alk. 4
Cl-
SO42-
Ba
B
Sr
mg L
mg L
mg L
mg L
g L
g L
3
39
168
21
80
58
132
451
í1
mg L
í1
í1
í1
í1
í1
í1
g Lí1
0.5
59
14
3
40
178
22
82
70
133
470
Carson-1 #1 (1/6/1998)
16.7 279
<0.1
19
4
3
51
141
20
<0.1
12
96
185
Carson-1 #2 (1/5/1998)
16.7 238
0.2
32
7
2
42
169
21
<0.1
39
102
369
Carson-1 #3 (1/6/1998)
16.7 168
0.2
44
12
3
47
164
23
62
58
105
398
Carson-1 #4 (1/6/1998)
16.7 69
0.2
86
21
4
74
204
112
112
199
117
835
Wilmington-1 #1 (4/24/1999)
20.8 279
0.1
50
16
7
106
134
213
<0.1
12
123
371
Wilmington-1 #2 (4/25/1999)
20.8 238
0.2
124
27
6
130
135
337
59
11
175
1,153
Wilmington-1 #3 (4/25/1999)
20.8 168
<0.2
214
47
9
346
173
907
50
27
240
2,129
Wilmington-1 #4 (4/25/1999)
20.8 69
0.1
282
96
13
457
142
1,209
288
121
221
3,688
Wilmington-1 #5 (4/24/1999)
20.8 37
<0.2
85
31
7
145
197
233
140
103
203
1,089
Wilmington-2 #1 (4/21/1999)
23.0 290
<0.1
3
2
5
195
377
56
<0.1
7
653
39
Wilmington-2 #2 (2/18/1999)
23.0 230
<0.1
35
24
13
499
450
513
<0.1
57
1,578
407
Wilmington-2 #3 (2/21/1999)
23.0 165
<0.1
20
7
4
102
180
72
<0.1
23
266
177
Wilmington-2 #4 (4/21/1999)
23.0 119
<0.1
143
67
16
492
308
1,012
29
117
557
1,266
Wilmington-2 #5 (2/18/1999)
23.0 37
<0.1
761
363
31
2,604
202
5,232
595
162
732
6,905
LWEB-1 (3/29/2001)
19.0 411
0.1
11
3.47
3.8
156
392
18.5
1
8.7
372
120
LWEB-2 (3/29/2001)
19.0 304
0.2
15.9
2.73
2.5
61.5
144
18.8
25.3
10
138
191
LWEB-3 (3/28/2001)
19.0 204
0.4
17.9
3.24
2.7
55.8
159
24.8
0.2
13
132
226
LWEB-4 (3/27/2001)
19.0 162
2.6
88.8
24
6.2
72.9
139
213
57.9
31
109
962
LWEB-5 (3/26/2001)
19.0 125
0.2
238
57.3
7.8
98.6
148
619
62.7
109
110
2,530
LBCH-1 (8/27/2003)
21.4 360
<0.1
6.2
3.03
4.8
206
503
22.9
3
6.3
785
77.2
LBCH-2 (8/26/2003)
21.4 198
0.1
13.5
4.08
3.5
107
181
74.4
1.2
4.6
238
171
LBCH-3 (8/26/2003)
21.4 143
<0.1
12.8
4.43
3.2
97.1
177
67.6
4
3.5
223
185
LBCH-4 (8/25/2003)
21.4 110
0.2
463
174
21
615
145
2,100
278
113
419
4,860
370-AJ (6/08/2005)
21.0 66
-
295
109
12
438
129
1,420
184
78
198
2,770
370-AH (6/08/2005)
21.0 20
-
454
335
34
1,440
200
3,810
391
167
361
4,650
LBPC-1 (4/03/2001)
24.2 366
<0.1
9.57
4.63
5.8
259
492
21.4
3.6
15
1,100
134
LBPC-2 (4/04/2001)
24.2 244
<0.1
6.55
7.18
11
422
614
337
1.4
17
1,150
97.5
LBPF-1 (4/24/2002)
27.0 332
0.1
17.8
30.5
26
1,500
1,230
1,650
-
127
8,400
734
LBPF-2 (4/24/2002)
27.0 102
0.1
519
1,220
272
10,800
289
19,900
2,640
82
4,130
8,500
Notes: 1 Approximate distance along flow path [6]; 2 Depth to top of perforation; 3 Dissolved oxygen, mg Lí1; 4 As CaCO3; 5 LWEB = Webster;
LBCH = Cabrillo; LBPC = Pier C; LBPF = Pier F.
Table 2. Summary parameters and description of well water for this study.
Huntington Park #1 Bent Spring
Huntington Park #2 Harbor
Chemical
composition 2
Ca-HCO3
Ca-HCO3
Change in chemical
composition (1998–2011) 3
unchanged
unchanged
Chloride
range 4
low
low
Stable
isotope 5
N
N
Carson-1 #1
Upper Wilmington
Na-HCO3
unchanged
low
N
Carson-1 #2
Upper Wilmington
Na/Ca-HCO3 unchanged
low
N
Carson-1 #3
Harbor
Ca/Na-HCO3 unchanged
low
N
Carson-1 #4
Pacific
Ca/Na-HCO3 mixing
low
N
Wilmington-1 #1
Upper Wilmington
Na-Cl
unchanged
low
N
Wilmington-1 #2
Upper Wilmington
Ca/Na-Cl
mixing
medium
N
Wilmington-1 #3
Upper Wilmington
Na/Ca-Cl
mixing
medium
I-S-N
Wilmington-1 #4
Wilmington-1 #5
Wilmington-2 #1
Harbor
Pacific
Pliocene B
Na-Cl
Ca/Na-Cl
Na-HCO3
variable
variable
variable
medium
low
low
I
I
N
Wilmington-2 #2
Pliocene A
Na-Cl
mixing
medium
S
Wilmington-2 #3
Lower Wilmington
Na-Cl/HCO3
variable
low
N
Wilmington-2 #4
Upper Wilmington
Na-Cl
mixing
medium
I-S-N
Well ID
Stratigraphic unit 1
Relative age
Comment
of water 6
old
Native water of good quality; end member for flow system
old
Native water of good quality; end member for flow system
Native water of good quality; source of recharge similar to
old
Huntington Park
Native water of good quality; source of recharge similar to
old
Huntington Park
Native water of good quality; source of recharge similar to
old
Huntington Park
Gradual decrease in TDS since initial sampling; [Cl] from
old
210 to ~40 mg Lí1
Possible enhanced lateral movement due to intense nearby
old
pumping
Possible enhanced lateral movement due to intense nearby
old
pumping
Inland from Dominguez Gap Seawater Barrier Project;
recent
contains mixture of native water, seawater, and imported
water from overlying unit
recent
Likely mixture of imported and seawater
recent
Likely mixture of imported and seawater
old
Isotopically light water recharged during Pleistocene
Principally isotopic light water (similar to Wilm2 #1);
old
localized saline unit attributed to partial mixing with an
oil-field brine
old
Native, fresh, sodium-bicarbonate water
Significant improvement in TDS likely a result of more
recent
effective injection; [Cl] decreased from ~1000 to 290 mg Lí1
127
128
Table 2. Cont.
Well ID
Stratigraphic unit 1
Wilmington-2 #5
Harbor
LWEB-1
Pliocene A
LWEB-2
Upper Wilmington
LWEB-3
LWEB-4
LWEB-5
LBCH-1
LBCH-2
LBCH-3
LBCH-4
370-AJ
370-AH
LBPC-1
LBPC-2
Upper Wilmington
Bent Spring
Harbor
Lower Wilmington
Upper Wilmington
Upper Wilmington
Bent Spring
Harbor
Dominguez
Pliocene B
Pliocene A
LBPF-1
Pliocene A
LBPF-2
Bent Spring
Relative age
in
chemical Chloride Stable
Change
Chemical
Comment
composition 2 composition (1998–2011) 3 range 4 isotope 5 of water 6
Some imported water is present, though masked by
seawater intrusion. Significant improvement in TDS likely
Na-Cl
mixing
high
S-I
recent
a result of more effective injection; [Cl] decreased from
~5200 to 2600 mg Lí1
Na-HCO3
unchanged
low
N
old
Isotopically light water recharged during Pleistocene
Geochemistry suggests subtle reactions or long-term
mixing
low
N
old
Na-HCO3
pumping effects
Na-HCO3
variable
low
N
old
Native, fresh, sodium-bicarbonate water
Ca-Cl
variable
low
N
old
Native, fresh, sodium-bicarbonate water
Ca-Cl
variable
low
N
old
Native, fresh, sodium-bicarbonate water
Na-HCO3
unknown
low
N
old
Isotopically light water recharged during Pleistocene
Na-HCO3
unknown
low
N
old
Isotopically light water recharged during Pleistocene
Na-HCO3
unknown
low
N
old
Isotopically light water recharged during Pleistocene
Na/Ca-Cl
unknown
low
N-S
recent
Na/Ca-Cl
unknown
medium N-S
recent
Na-Cl
unknown
high
N-S
recent
Na-HCO3
unknown
low
N
old
Isotopically light water recharged during Pleistocene
Na-HCO3/Cl unknown
medium N
old
Isotopically light water recharged during Pleistocene
Old seawater, distinct major ion composition and trace
medium N
old
Na-Cl/HCO3 unknown
element ratios
Na-Cl
unknown
high
S
old
Groundwater consisting mostly of seawater
1
Notes: Nomenclature consistent with most recent model layer assignments [43]; as well as in Ponti et al. [30] and Figure 2; 2,3 Change in chemical composition: where period of record is
available, a general description of water quality over time is given; 4 Chloride range: low = <250 mg Lí1, medium = 250–2500 mg Lí1, and high = >2500 mg Lí1; 5 N, native water; S, seawater;
I, imported water; 6 Relative age: see Section on Tritium; about 1 tritium unit (TU) used for recent/old categorization.
129
5.2. Tritium
The tritium data provide insight as to the relative age or “old” versus “new” groundwater in our
study. Tritium (3H; t1/2 = 12.4 years) is the only radioactive isotope of hydrogen and while it is
naturally present only in minute (<<1%) quantities, it is also produced as a fission product in nuclear
weapons tests and nuclear power reactors. The convention for reporting 3H concentrations is the
tritium unit (TU), which equals 3.2 pCi Lí1 (7.104 dpm Lí1). As tritium is naturally incorporated into
the water molecule and its abundance is only affected by radioactive decay, 3H serves as a useful
tracer for identifying recently recharged water. A pre-fallout (pre-1952) background 3H abundance
in southern California coastal precipitation was ~2 TU [44]. Beginning in 1952, 3H was released into
the atmosphere, reaching a maximum in 1963 [45]. A reconstructed Los Angeles County
precipitation tritium concentration curve [6] identifies a narrow 1963 peak at ~700 TU that rapidly
decreased to <100 TU by 1970. As a consequence, without consideration for complex mixing
scenarios, groundwater with a 3H value less than <1 TU may be considered “older” water that was
recharged prior to 1952. Conversely, groundwater with a tritium content >1 TU can be interpreted as
“recent” water being wholly or partially recharged post-1952. Along the coast, such interpretation
may be more complicated as recent seawater may provide another source of tritium.
Figure 3. (A) Chloride versus Na+ (except LBPF-2) and (B) SO42í versus Ca2+.
130
Of the 31 groundwater samples analyzed for tritium (average 3H value = 4.3 TU), seven samples
had a 3H concentration >1 TU and of these, three had 3H >10 TU (Table 3); each of these wells
perforated the Pacific and Harbor sequences. The large number of low or less than measureable
(0.1 TU) tritium values indicates that most sampled groundwater in the Dominguez Gap region
appears to be older water (pre-1952). Notable exceptions include wells close to the coast
(Wilmington-1 #4, Wilmington-1 #5, and Wilmington-2 #5) that are directly influenced by recent
seawater intrusion and the seawater barrier injection wells that may introduce additional, substantial
3
H as a result of complex mixing scenarios (Table 2; Figure 2).
Table 3. Select stable and radiogenic isotope data for wells sampled in the study site.
Note: pmc = percent modern carbon.
Well ID
Huntington Park #1
Huntington Park #2
Carson-1 #1
Carson-1 #2
Carson-1 #3
Carson-1 #4
Wilmington-1 #1
Wilmington-1 #2
Wilmington-1 #3
Wilmington-1 #4
Wilmington-1 #5
Wilmington-2 #1
Wilmington-2 #2
Wilmington-2 #3
Wilmington-2 #4
Wilmington-2 #5
LWEB-1
LWEB-2
LWEB-3
LWEB-4
LWEB-5
LBCH-1
LBCH-2
LBCH-3
LBCH-4
370-AJ
370-AH
LBPC-1
LBPC-2
LBPF-1
LBPF-2
į18O
įD
3
(‰)
í7.32
í7.23
í7.33
í7.27
í7.30
í7.12
í7.29
í7.13
í7.34
í9.77
í9.58
í8.73
í8.63
í7.84
í7.96
í7.57
í9.17
í7.72
í8.07
í7.32
í7.10
í9.39
í8.37
í8.15
í6.86
í7.12
í6.76
í9.23
í7.54
í7.47
í0.42
(‰)
í47.5
í47.3
í48.4
í46.6
í47.3
í47.0
í46.8
í46.0
í49.4
í77.5
í73.3
í59.7
í55.8
í50.6
í51.3
í57.7
í62.7
í51.1
í53.2
í48.4
í47.4
í64.0
í55.1
í53.6
í45.9
í49.9
í47.3
í61.1
í49.2
í47.1
í3.65
(TU)
<0.1
<0.1
0.1
<0.1
<0.1
0.1
<0.1
<0.1
1.6
19
11.9
<0.1
<0.1
<0.1
1.5
16.9
<0.1
0.1
<0.1
<0.1
<0.1
<0.1
0.1
<0.1
1.8
1.1
0.9
0.2
<0.1
<0.1
<0.1
H
į13CDIC
14
(‰)
í13.6
í13.5
í12.4
í12.3
í14.2
í15.8
í18.4
í0.3
í0.3
í15.0
6.3
í15.8
í18.9
í14.3
í13.9
4.7
í10.6
í16.0
í16.0
3.1
í12.7
í6.4
í12.3
(pmc)
83.7
83.5
27.6
43.6
56.1
29.6
44.5
2.5
5.2
14.4
1.8
24.1
14.1
42.8
55.1
3.5
8.8
9.8
61.4
2.5
4.3
0.8
48.5
C
131
5.3. Isotopic Composition of Oxygen (į18O) and Hydrogen (įD)
The behavior of stable oxygen (į18O) and hydrogen (įD) isotopes in groundwater can provide
insight into the geochemical character, origin, and transport phenomena of groundwater [8].
Reporting convention for both isotopes is expressed in terms of relative difference, per mill (‰),
from the Vienna Standard Mean Ocean Water (VSMOW) value. In general, the predominant source
of precipitation is from evaporation of seawater, and as a result, the observed global composition
of į18O and įD in rainwater is linearly expressed as the global meteoric water line (GMWL;
įD = 8į18O + 10‰; [46,47]).
In the 31 groundwater samples, the į18O composition ranged from í0.42‰ (LBPF-2) to í9.77‰
(Wilmington-1#4), while the įD composition varied between í3.65‰ (LBPF-2) to í77.5‰
(Wilmington-1#4) (Table 3). A plot of įD versus į18O (Figure 4) indicates a strong linear relationship
(R2 = 0.93), with a slope of 7.2—close to that of the global meteoric water line (GMWL; slope = 8).
Notable exceptions of groundwater (i.e., Wilmington-1#4; Wilmington-2 #5) that influence such a
shift below the GMWL include isotopically heavier water that likely consists of a recent mixture of
saline (e.g., seawater) and imported fresh water [4]. Lighter įD values (<í50‰), observed in some
groundwater samples (e.g., LBPF-1,2, Wilmington-1#1-3, Huntington Park 1,2, Carson-1#1-4,
370-AJ, 370-AH, LWEB-4,5) may identify older groundwater with a isotopically unique signature.
Figure 4. įD versus į18O in selected groundwater samples from the study area. The
global meteoric water line is per Craig (1961). Regression results include all groundwater
data. Water recharged from Los Angeles and Montebello Forebays has a į18O signature
of í7.5‰ to í6.7‰ and í9.5‰ to í8.0‰, respectively, and is isotopically distinct from
non-native, seawater, and oil-field brine values.
132
The composition of įD and į18O in precipitation may also be influenced by local air mass and
vapor trajectories, changes in evaporation, and isotope exchange processes below the cloud base [10,47].
Thus, climatic variations may be recorded in the composition of į18O and įD in groundwater [12].
In LBPF-2, where the Clí concentration approaches a seawater-like value (Figure 4), both į18O and
įD isotopic compositions are highly enriched (heavy) (í0.42‰ for į18O and įD for í3.65‰)
compared to other well waters (í6.76 to í9.77‰ for į18O and í45.9 to í77.5‰ for įD). The isotopic
composition of the well waters, coupled with the major ion chemistry, implies variable mixing with
recent seawater. In addition, isotopically lighter water with low Clí content has also been attributed
to Colorado River water [6] that is used as a source of injection water at the barrier wells [48].
Observed high Clí concentrations along with enriched į18O reveal non-native inputs from seawater
and/or oil-field brines (Table 2).
5.4. Isotopic Composition of į13C in Dissolved Organic Carbon (DIC)
Dissolved inorganic carbon (DIC = [CO2aq] + [HCO3í] + [CO32í]) is generally produced in
groundwater by the dissolution of CO2 during plant (i.e., C3 and/or C4) respiration, the microbial
decomposition of organic matter, and the direct dissolution of carbonate minerals [49]. The
composition of į13C, expressed as per mill (‰) relative to the VPDB (Vienna PeeDee Belemnite)
standard, provides a useful tracer to assess the relative contribution of C from these various sources.
Under an open CO2 system, the groundwater į13CDIC should approach ~9‰ by simple hydrolysis
reactions of soil CO2 alone [7]. Conversely, if the groundwater is closed to soil CO2, then the
į13CDIC should approach values of about í13‰. In groundwater that is strongly reducing and
sulfate-poor [50], the composition of į13CDIC can increase to values in excess of 30‰ as a result of
methanogenesis [51]. Observed saturation calculations for portions of the study area, and elsewhere
in the basin, indicate that calcite should precipitate.
In this study, the groundwater į13C in dissolved inorganic carbon ranged from í18.9 (LWEB-3)
to 6.3‰ (LWEB-1) (Table 3), which reflects the contribution of different DIC sources and/or the
evolution from an open to a closed system. Wells sampled for this study, with the exception of
370-AH and possibly Wilmington-1#5 appear to respond within a confined system. Three deep
samples (LWEB-1, LBCH-1, LBPC-1) contain į13CDIC values >0‰ and very little SO42í, which
would suggest a unique carbon source, such as from an incomplete bacterially-mediated
methanogenic pathway [52]. There have been a number of studies indicating a linear relationship
between į13CDIC and 1/dissolved inorganic carbon (DIC) [49,52–55]. In lieu of direct dissolved
inorganic carbon measurements alkalinity (expressed as CaCO3) can serve as a proxy for DIC, under
the condition that CO2 remains constant [56]. If one excludes values from wells closest to the coast
and down-gradient from the PCH Fault (e.g., LBPC-2, LBPF-1,2) there is a good correlation (R2 =
0.88) between į13CDIC and 1/alkalinity (Figure 5). Such a trend, which implies more than simple
carbonate mineral dissolution, is expected in a complexly mixed coastal aquifer undergoing chemical
evolution [11,13,57].
133
Figure 5. į13CDIC versus 1/alkalinity (excluding wells closest to the coast; LBPC-2, LBPF1,2).
5.5. Carbon-14 (14C)
To assess the relative age of select groundwater samples, 14C (t½ = 5730 years) was also
determined. Natural 14C is mainly produced in the atmosphere by interaction of cosmic ray derived
secondary neutrons with 14N. Carbon-14 derived age results are often expressed as percent modern
carbon 14C (pmc) by comparing the 14C activity of a sample to the known activity of an oxalic acid
standard. Carbon-14 age data are generally interpreted within the context of a geochemical
reactions/evolution model that can account for the various sources and sinks of carbon [58].
In the Los Angeles County groundwater samples, the percent modern carbon (pmc) exhibited a
wide range from 0.8 pmc (LBPF-1) to 83.7 pmc (Huntington Park #1) (Table 3) with an average
value of ~30 pmc. Assuming an initial 14C value of 90 pmc [4], the corresponding groundwater age
estimates may extend from recent to beyond 20 kyr before present and suggest that some well water
undergoes deep circulation of native water through multiple aquifer systems. Age estimates,
however, are not corrected for potential exchange reactions of carbon within the aquifer, and thus
may not reflect the true age of the groundwater. The observed variations in the percentage of modern
14
C indicate that the groundwater system is comprised of complex mixtures of diverse waters. A plot
of į13CDIC versus 14C (Figure 6) indicates that in samples with <10 pmc, į13CDIC values fluctuated from
í16‰ to +6‰, while in samples with >10 pmc, the į13CDIC values ranged from í12‰ to í18‰. Select
groundwaters (e.g., Wilmington 2-1, LWEB1,2, LBCH1-3, and LBPC1,2) that likely were recharged
during Pleistocene, are also isotopically light. Nonetheless, 14C data confirm complex water mixing
and transport scenarios involving multiple aquifer systems that reside within a tectonically active
geologic framework.
134
Figure 6. į13CDIC versus percentage of modern
Angeles Basin.
14
C in groundwater samples from Los
5.6. U- and Th-Series Radionuclides
The isotopic systematics of many of the naturally occurring radionuclides in the U- and Th-series
decay series are invaluable in investigating aquifer behavior. Specifically, U, Th, Ra, and Rn are all
ubiquitous in groundwater and are represented by multiple isotopes with very different half-lives
such that groundwater processes can be studied over a large range in time-scales. Within aquifer host
minerals, these radionuclides are generally expected to be in secular equilibrium. However, these
same radionuclides may exhibit strong fractionations with the surrounding groundwaters. Such
disequilibria can be used, for example, to obtain information of radionuclide release from aquifer
host rocks, groundwater flow rates, or age dates.
5.6.1. Specific Activities of 238U, 230Th and 232Th in Well Cuttings
The specific activity of 238U (t½ = 4.47 Gyr) from select well cuttings varied between 0.46 and
0.73 dpm gí1 (2.22 dpm = 1 pCi), while the 234U/238U activity ratios varied between 0.99 and 1.02
(Table 4). Within one sigma, 234U (t½ = 2.5 Gyr) and 238U are in secular equilibrium with one another,
which indicates that the alpha recoil loss [59] of 234U is generally negligible in these samples. In
contrast, the activity of 230Th (t½ = 73 kyr) was considerably greater than that of 238U in host rock,
thus recoil and/or long-term weathering reactions can provide an additional source of 230Th [21,60].
The 232Th/238U activity ratios varied between 1.30 and 2.25, a range considerably higher than the
average crustal value [61].
135
Table 4. Solid-phase activities of 238U, 230Th, 232Th, and 234U/238U activity ratios (AR).
238
Well ID
LWEB, Harbor
LWEB, Bent Spring
LWEB, Upper Wilmington
LWEB, Upper Wilmington
LBPC, Pliocene A
LWEB, Pliocene A
LBPC, Pliocene B
Ua
(dpm gí1)
0.47
0.49
0.46
0.42
0.56
0.73
0.6
230
232
Th
(dpm gí1)
0.72 ± 0.06
0.66 ± 0.06
0.60 ± 0.06
0.60 ± 0.06
0.72 ± 0.06
1.02 ± 0.06
0.90 ± 0.06
Th
(dpm gí1)
0.84 ± 0.06
0.84 ± 0.06
0.60 ± 0.06
0.90 ± 0.06
1.26 ± 0.09
1.44 ± 0.09
0.90 ± 0.06
234
U/238U
AR
1.01 ± 0.01
1.00 ± 0.01
0.99 ± 0.02
1.02 ± 0.01
0.99 ± 0.01
1.02 ±0.01
1.01 ± 0.01
Note: a Analytical error < 3%.
5.7. Concentrations and Activity Ratios of 222Rn and Ra Isotopes
Ra isotopes can provide unique information regarding the production of U/Th series radionuclides
in groundwater and reveal where significant transformations in adsorption or parent element
distribution can occur along a groundwater flow path [62,63]. Values for Ra partitioning coefficients
and retardation factors can also be obtained from the Ra isotopes but only by assuming that 222Rn
provides a reasonable proxy for the recoil production rates of radium.
The concentrations of 222Rn (t½ = 3.825 days) ranged from 142,000 dpm mí3 (LWEB-2) to
442,000 dpm mí3 (370-AJ) (average 222Rn = 260,000 dpm mí3), which are expectedly the highest
values observed as compared to other members of U- and Th-series radionuclides reported here
(Table 5). The large 222Rn concentrations are the result of radon’s inert character as a noble gas
and its resulting inability to participate in any scavenging reactions. The concentration of
226
Ra (t½ = 1600 years) varied between 29 dpm mí3 (Wilmington-2#3) and 1632 dpm mí3
(Wilmington-2#5) (average 226Ra = 257 dpm mí3), which is ~2–4 orders of magnitude lower than
the 222Rn activities. Rn-222 is a direct measure of 226Ra (direct radiogenic parent of 222Rn) in the host
rocks as well as a measure of the relative emanation efficiency from the host rock [64]. A plot of
226
Ra activity as a function of Clí concentration is shown in Figure 7. Elevated Ra follows an increase
in Clí and is attributed to solubilization of chloride complexes and/or through displacement from
clays by ion exchange and desorption reactions.
Figure 7. Ra-226 activity (dpm mí3) versus Clí concentration.
136
Table 5. Activities (dpm mí3) of dissolved 222Rn and four Ra isotopes in select wells
from within the study area.
222
Well ID
Huntington Park #1
Huntington Park #2
Carson-1 #1
Carson-1 #2
Carson-1 #3
Carson-1 #4
Wilmington-1 #1
Wilmington-1 #2
Wilmington-1 #3
Wilmington-1 #4
Wilmington-1 #5
Wilmington-2 #1
Wilmington-2 #2
Wilmington-2 #3
Wilmington-2 #4
Wilmington-2 #5
LWEB-1
LWEB-2
LWEB-3
LWEB-4
LWEB-5
LBCH-1
LBCH-2
LBCH-3
LBCH-4
370-AJ
370-AH
LBPC-1
LBPC-2
LBPF-1
LBPF-2
Rn
í3
(dpm m )
244,000 ± 39,000
142,000 ± 47,000
242,000 ± 47,000
202,000 ± 23,000
195,000 ± 36,000
217,000 ± 11,000
418,000 ± 27,000
336,000 ± 75,000
388,000 ± 39,000
442,000 ± 39,000
347,000 ± 58,000
160,000 ± 24,000
223,000 ± 46,000
172,000 ± 20,000
173,000 ± 11,000
223
Ra
í3
(dpm m )
147.4
44.1
26.2
35.6
56.5
109.8
3.8
11
9.1
73.4
16
14.7
23.4
4.9
4.3
204.5
3.4
4.1
13.5
60.6
55.9
2.9
9.5
18
151.4
52.9
49.2
22.2
10
24.5
54.5
224
Ra
í3
(dpm m )
2,138.8
1,720.4
844.6
2,270.7
1,976.2
3,310.2
111.9
296.6
284.4
1,392.2
466.9
792.2
2,903.1
1,813.5
2,254.6
3,918.8
170
115
277
1,285
1,301
95
152
690
11,351
765
2,976
294
138
1,022
4,489
228
Ra
í3
(dpm m )
457 ± 25
306 ± 24
172 ± 15
311 ± 19
353 ± 23
694 ± 32
199 ± 16
511 ± 26
452 ± 19
825 ± 35
654 ± 30
173 ± 17
177 ± 17
378 ± 18
553 ± 27
2,716 ± 85
84 ± 9
83 ± 10
581 ± 17
826 ± 21
53.6 ± 9.1
55.2 ± 8.6
1,269 ± 24
2,944 ± 58
826 ± 21
86 ± 12
631 ± 23
-
226
Ra
(dpm mí3)
421 ± 9
159 ± 7
76 ± 4
128 ± 5
154 ± 6.7
296 ± 8
76 ± 4
183 ± 6
166 ± 5
455 ± 9
479 ± 10
66 ± 4
145 ± 6
29 ± 1
225 ± 7
1,632 ± 18
51.6 ± 5.0
36.7 ± 4.4
275 ± 7
282 ± 7
31.1 ± 4.2
29.2 ± 4.0
522 ± 9
515 ± 10
282 ± 7
38 ± 4.6
192 ± 6.5
-
The concentration of 223Ra (t½ = 11.4 days) varied between 2.9 dpm mí3 (LBCH-1) and 204.5 dpm
mí3 (Wilmington 2#5) (mean 223Ra = 42 dpm mí3), while the concentrations of 224Ra (t½ = 3.66 days)
and 228Ra (t½ = 5.75 years) varied between 95 dpm mí3 (LBCH-1) and 11,351 dpm mí3 (LBCH-4)
(mean 224Ra = 1665) dpm mí3 and 54 dpm mí3 (LBCH-1) and 2944 dpm mí3 (370-AH) (mean
228
Ra = 606 dpm mí3), respectively. The 223Ra/226Ra activity ratio (AR) varied between 0.02 and 0.37
(Table 6), with a mean value of 0.18, a value substantially higher than the expected value of 0.046.
Because 223Ra and 226Ra are both generated after three Į decays, groundwater should have a 223Ra/226
Ra activity ratio similar to the host rock (235U/238U) activity ratio of 0.046. Higher 223Ra/226Ra activity
137
ratio may be observed in groundwater after a recharge or precipitation event, as
longer half-life, will not yet have reached a steady state concentration [16,65,66].
226
Ra, due to its
Table 6. Dissolved activity ratios (AR) of 228Ra/226Ra, 224Ra/228Ra, 224Ra/223Ra,
224
Ra/222Rn, and 223Ra/226Ra, as well as the model-derived parameters: ȍ, k1, k2, and Rf
in select wells.
Well ID
LWEB-1
228Ra/226Ra
224Ra/228Ra
224Ra/222Rn
í4
223Ra/226Ra
ȍ224
ȍ228
í4
í4
Rf
(× 103)
0.53
1.3
4.1
-
-
-
2.6
0.21
0.56
AR
AR (× 10 )
AR
1.61 ± 0.23
2.04 ± 0.23
50.7 ± 5.7
7.0 ± 1.2
0.065 ± 0.009
4.9
2.4
28.1 ± 3.1
8.1 ± 2.7
-
5.4
20.5 ± 2.3
11.4 ± 2.3
0.367 ± 0.058
8.8
3.34 ± 0.43
í1
(× 10 )
AR
2.25 ± 0.38
k2 (min-1)
k1
í4
AR
LWEB-2
LWEB-3
224Ra/223Ra
(× 10 ) (× 10 ) (min )
3.8
LWEB-4
2.11 ± 0.08
2.21 ± 0.13
21.2 ± 2.4
63.8 ± 8.1
0.220 ± 0.023
50
23
0.05
1.1
0.5
LWEB-5
2.93 ± 0.10
1.57 ± 0.09
23.3 ± 2.6
66.7 ± 12.8
0.198419
57
36
0.06
2.3
0.3
LBCH-1
1.72 ± 0.38
1.78 ± 0.32
32.5 ± 3.6
4.4 ± 0.3
0.094 ± 0.016
2.9
1.7
1
1.7
5.9
LBCH-2
1.89 ± 0.39
2.76 ± 0.45
16.0 ± 1.8
3.7 ± 0.3
0.325 ± 0.055
2.4
0.88
0.85
0.74
11.5
LBCH-3
-
-
38.2 ± 4.3
20.5 ± 4.7
-
-
-
-
-
LBCH-4
-
-
74.9 ± 8.4
293 ± 33
-
-
-
-
-
370-AJ
2.43 ± 0.06
0.60 ± 0.03
14.5 ± 1.6
17.3 ± 1.7
0.101 ± 0.010
15
25
-
-
370-AH
5.72 ± 0.16
1.01 ± 0.05
60.5 ± 6.8
86 ± 15
0.096 ± 0.010
73
72
1.7
121
0.1
LBPC-1
2.93 ± 0.10
0.36 ± 0.02
13.3 ± 1.5
18.4 ± 2.9
0.079 ± 0.008
18
52
-
-
LBPC-2
2.26 ± 0.42
1.61 ± 0.24
13.9 ± 1.6
6.20 ± 1.3
0.262 ± 0.041
3.5
2.2
1
2.1
4.8
LBPF-1
-
-
41.7 ± 4.7
59.5 ± 7.5
0.127 ± 0.013
41
26
0.08
2.1
0.4
LBPF-2
-
-
82.4 ± 9.2
259 ± 21
-
204
-
0.006
-
Notes: ȍ = Ratio of the activity of a radionuclide in solution (ȜNd), to its production rate, P; k1 = first
order adsorption rate constant; k2 = first order desorption rate constant; Rf = Retardation Factor, calculated
from k1/k2.
The 224Ra/228Ra activity ratios provide a measure of the adsorption and desorption rate constants
for radium [15,21]. Within host rocks that are in secular equilibrium, 224Ra/228Ra = 1. The 224Ra/228Ra
activity ratios in the groundwater samples (Table 6) varied between 0.36 (LBPC-1) and 3.34 (LWEB-2).
While in general the fresh groundwater 224Ra/228Ra ARs fall in a reasonably narrow range
(0.5–2.0; [15,60,67]), but much higher values have also been observed [61]. Typically, higher
224
Ra/228Ra ARs occur in groundwaters where steady state conditions have not yet been reached or
in transitional coastal groundwater systems that are variably affected by seawater mixing [38].
In these groundwater samples (e.g., LBCH-2, LWEB-1,4,3), observed elevated 224Ra/228Ra ARs may
identify waters that have recently been mixed with seawater.
Due to their short and similar half-lives (t½ = <4 days), the activities of 224Ra and 222Rn are
expected to be in steady state in most groundwater. From the host rock 232Th/238U activity ratio,
224
Ra/222Rn activity ratios can be used to calculate recoil and sorption rate constants [21].
The observed 224Ra/222Rn activity ratios (Table 6) varied between 1.34 × 10í5 to 3.9 × 10í4
(average = 1.25 × 10í4) and agree well with other reported values (0.2–4.4 × 10í4; [15,60,67]). The
measured range in 224Ra/222Rn activity ratios reflects the natural variability of these radionuclides in
this groundwater system. The activity ratio of the two longest lived Ra isotopes, 228Ra/226Ra, provides
a measure of the relative recoil rates of radionuclides from two decay series [21]. The observed
138
228
Ra/226Ra activity ratios (Table 6) ranged between 1.1 and 5.7 (average = 2.3). Because 226Ra is the
product of three Į decays, while 228Ra is produced by one single Į decay, 226Ra may be more mobile
than 228Ra, resulting in lower 228Ra/226Ra activity ratios. Differences in the activity ratios can be
attributed to variations in the distribution of U and Th in host rocks.
5.8. Adsorption-Desorption Rate Constants and Retardation Factors
The production rates for Ra isotopes were calculated using 222Rn as the recoil flux monitor and is
based on the relation [15]:
Fi = Fr (Qi/Qr) İ
(6)
where Fi and Fr are the recoil supply rates of Ra isotopes (224Ra and 228Ra) and 222Rn to the
groundwater respectively; Qi and Qr are the production rates of Ra isotopes and 222Rn in the aquifer
solids; and İ is the rate of recoil supply of 224Ra and 228Ra relative to the 222Rn recoil supply. The
term İ depends on where the radionuclide is positioned in the decay series, the Į particle energy
released during its production, and the scavenging capability of its immediate radiogenic parent. The
value of İ can vary from a steady state value of ~1.5, if all the 226Ra recoiled into the groundwater
remains in solution, to 0.86 if all the recoiled 226Ra is adsorbed onto the aquifer grain surfaces [15].
For the derivation of Fi and Fr, we assumed a İ value of 1.0.
The calculated values of ȍ224, ȍ228, k1, k2, and Rf are given in Table 6. The adsorption rate
constants (k1) calculated based on the 224Ra and 228Ra concentrations varied between 0.006 to
1.70 miní1, with an average value of 0.55 miní1 and co-varied positively (R = 0.71, n = 7) with Na+
concentration if one excludes the three highest Na+ values (LBPF-1,2, Wilmington-2#5). This
observation is intriguing, as one might expect an inverse correlation, as higher Na+ would imply more
Na+ available for exchange, and thus, longer Ra residence time (or smaller k1 values). More studies
need to be conducted to validate this observation. The corresponding residence times, calculated for
an irreversible adsorption model (1/k1) ranged from 0.59 to 20.0 min (average = 6.62 min).
The desorption rate (k2) constant varied between 0.56 and 121 × 10í4 miní1, with a mean value of
14.8 × 10í4 miní1. The corresponding average residence time with respect to desorption was
expectedly much greater. Faster sorption (k1) of Ra injected into the water and slower desorption (k2)
from the host rock has been previously documented [15,17,18,67]. This range in values is in contrast
with the values reported for subsurface brines, where k1 and k2 values are typically comparable
[16,68]. As k1 is always much greater than k2 for these groundwater samples, the ratio k1/k2 can
be used as a measure of the retardation factor, Rf [15], which here ranged from 0.1 to 11.5 × 103
(average = 3.5 × 103). Such Rf values are on the same order of magnitude as has been reported for
other groundwater systems [15,16,60,67].
6. Conclusions
Stable- and radio-isotopes, as well as a complementary suite of water quality parameters, were
utilized to assess groundwater properties from select wells in the Los Angeles Basin–Dominguez
Gap area. Groundwater resources in this region have been extensively developed and managed since
the late 1800s. In the study area, groundwater resides in multiple aquifer systems that are complexly
139
mixed with seawater, non-native water that is used to stave off saltwater intrusion, and oil-field
related brine waters. Elevated Clí concentrations are observed in some nearshore wells (LBPF-2,
Wilmington-2 #5) and in the Dominguez Gap area. The į18O composition observed in select wells
provides a measure of saltwater and oil-field brine mixing. Tritium data from these wells reveal that
recent (less than 50 years old) groundwater was present only in a few select wells in the Upper and
Lower aquifer systems close to the Dominguez Gap area (Wilmington-1 #4, Wilmington-1 #5, and
Wilmington-2 #5) where the seawater barrier injection wells may introduce additional 3H. The
į13CDIC composition in groundwater ranged from í18.9‰ to + 6.3‰ and provides an indication of
the evolution of this groundwater system; some well water exhibited considerably lower į13CDIC
values and in the absence of SO42í may reflect a contribution of enhanced 13C from microbial
methanogenesis under anoxic conditions. Lastly, į18O (í9.77‰ to í0.42‰) and deuterium (í3.65‰
to í77.5‰) in these well waters provide information on the likely quantification of water balance
and interaction within adjacent groundwaters.
From a series of well cuttings, aquifer host rock 234U and 238U are expectedly in secular
equilibrium. In contrast, U- and Th-series radionuclides exhibit strong fractionation in groundwaters
from select wells. Such radiogenic disequilibria are used to assess fundamental aquifer properties,
such as the retardation factor Rf and adsorption/desorption rate constants, k1and k2, respectively.
While we cannot quantify a relative groundwater age from the Ra isotopes alone, the calculated
adsorption rate constant (k1) is markedly higher than the desorption rate constant (k2), which implies
that the residence time of dissolved Ra must be very short (average ~7 min).
In combination, this suite of isotopic and elemental tracers can provide valuable information on
complex groundwater mixing scenarios and provenance, which are useful to assess groundwater
vulnerability to current and future external stressors, such as groundwater withdrawals,
contamination, and sea level rise.
Acknowledgements
We thank John Haines of the USGS Coastal and Marine Geology (CM&G) Program for continued
support in coastal groundwater studies. Nancy Prouty, Renee Takesue, and Christopher Conaway
provided thoughtful reviews of an earlier version of this manuscript. The authors are also thankful
for the four constructive reviews that greatly improved the manuscript. The use of trade, product, or
firm names in this publication is for descriptive purposes only and does not imply endorsement by
the U.S. government.
References
1.
2.
Mendenhall, W.D. Development of Underground Waters in the Eastern Coastal Plain Region
of Southern California; Water-Supply Paper 137; U.S. Geological Survey: Reston, VA, USA, 1905.
California Department of Water Resources. Planned Utilization of the Groundwater Basins of
the Coastal Plain of Los Angeles Country, Appendix A: Ground Water Geology; Bulletin 104;
California Department of Water Resources: Sacramento, CA, USA, 1961.
140
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Los Angeles County Department of Public Works Web Page. Seawater Barriers. Available
online: http://dpw.lacounty.gov/wrd/Barriers/ (accessed on 1 March 2013).
Reichard, E.G.; Land, M.; Crawford, S.M.; Johnson, T.; Everett, R.R.; Kulshan, T.V.;
Ponti, D.J.; Halford, K.L.; Johnson, T.A.; Paybins, K.S.; et al. Geohydrology, Geochemistry, and
Ground-Water Simulation-Optimization of the Central and West Coast Basins, Los Angeles
County, California; Water-Resources Investigations Report 03-4065; U.S. Geological Survey:
Reston, VA, USA, 2003.
Land, M.; Everett, R.R.; Crawford, S.M. Geologic, Hydrologic, and Water-Quality Data from
Multiple-Well Monitoring Sites in the Central and West Coast Basins, Los Angeles County,
California; Open-File Report 01-277; U.S. Geological Survey: Reston, VA, USA, 2002.
Land, M.; Reichard, E.G.; Crawford, S.M.; Everett, R.R.; Newhouse, M.W.; Williams, C.F.
Ground-Water Quality of Coastal Aquifer Systems in the West Coast Basin, Los Angeles County,
California, 1999–2002; Scientific Investigations Report 2004–5067; U.S. Geological Survey:
Reston, VA, USA, 2004.
Clark, I.; Fritz, P. Environmental Isotopes in Hydrogeology; Lewis Press: Boca Raton, FL,
USA, 1997.
Kendall, C.; McDonnell, J.J. Isotope Tracers in Catchment Hydrology; Elsevier Science:
Amsterdam, The Netherlands, 1998.
Hem, J.D. Study and Interpretation of the Chemical Characteristics of Natural Water,
4th ed.; Water-Supply Paper 2254; U.S. Geological Survey: Reston, VA, USA, 1992.
Gat, J.R.; Gonfiantini, R. Stable Isotope Hydrology—Deuterium and Oxygen-18 in the Water
Cycle; International Atomic Energy Agency: Vienna, Austria, 1981.
Kendall, C.; Mast, M.A.; Rice, K.C. Tracing Watershed Weathering Reactions with į13C. In
Water-Rock Interactions, Proceedings of the 7th International Symposium, Park City, UT, USA,
13–18 July 1993; Kharaka, Y.K., Maest, A.S., Eds.; Balkema: Rotterdam, The Netherlands,
1993; pp. 569–572.
Rozanski, K.; Araguas-Araguas, L.; Gonfiantini, R. Isotopic patterns in modern global
precipitation, in climate change in Continental Isotopic Records. Geophys. Monogr. 1993, 78,
1–36.
Izbicki, J.A., Danskin, W.R., Mendez, G.O. Chemistry and Isotopic Composition of
Groundwater along a Section near the Newmark Area, San Bernardino County, California;
Water-Resources Investigations Report 97-4179; U.S. Geological Survey: Reston, VA, USA,
1998.
Szabo, Z.; Rice, D.E.; Plummer, L.N.; Busenberg, E.; Drenkard, S.; Schlosser, P. Age dating of
shallow groundwater with chlorofluorocarbons, tritium/helium 3, and flow path analysis,
southern New Jersey coastal plain. Water Resour. Res. 1996, 32, 1023–1038.
Krishnaswami, S.; Graustein, W.C.; Turekian, K.K.; Dowd, J.F. Radium, thorium, and
radioactive lead isotopes in groundwaters: Application to the in-situ determination of
adsorption-desorption rate constants and retardation factors. Water Resour. Res. 1982, 18,
1633–1675.
141
16. Krishnaswami, S.; Bhushan, R.; Baskaran, M. Radium isotopes and 222Rn in shallow brines,
Kharaghoda (India). Chem. Geol. 1991, 87, 125–136.
17. Copenhaver, S.A.; Krishnaswami, S.; Turekian, K.K.; Shaw, H. 238U and 232Th series nuclides
in groundwater from the J-13 well at the Nevada test site: Implications for ion retardation.
Geophys. Res. Let. 1992, 19, 1383–1386.
18. Copenhaver, S.A.; Krishnaswami, S.; Turekian, K.K.; Epler, N.; Cochran, J.K. Retardation of
238
U and 232Th decay chain radionuclides in Long Island and Connecticut aquifers. Geochim.
Cosmochim. Acta 1993, 57, 597–603.
19. Hussain, N. Supply rates of natural U-Th series radionuclides from aquifer solids into
groundwater. Geophys. Res. Let. 1995, 22, 1521–1524.
20. Ku, T.-L.; Luo, S.; Leslie, B.W.; Hammond, D.E. Decay-series disequilibria applied to the study
of rock-water interaction and geothermal systems. In Uranium-series Disequilibrium; Ivanovich,
M., Harmon, R.S., Eds.; Clarendon Press: Oxford, England, 1992; pp. 631–688.
21. Porcelli, D.; Swarzenski, P.W. The behavior of U- and Th-series nuclides in groundwater. Rev.
Mineral. Geochem. 2003, 52, 317–361.
22. Olsen, C.R.; Cutshall, N.H.; Larsen, I.L. Pollutant-particle associations and dynamics in coastal
marine environments: A review. Mar. Chem. 1982, 11, 501–533.
23. Dickson, B.L. Radium isotopes in saline seepages, south-western Yilgarn, Western Australia.
Geochim. Cosmochim. Acta 1985, 49, 361–368.
24. Dickson, B.L.; Wheller, G.E. Uranium-series disequilibrium exploration geology. In
Uranium-Series Disequilibrium: Applications to Earth, Marine, and Environmental Sciences;
Ivanovich, M., Harmon, R.S., Eds.; Clarendon Press: Oxford, UK, 1992; pp. 704–730.
25. Rama; Moore, W.S. Mechanisms of transport of U-Th series radioisotopes from solids into
groundwater. Geochim. Cosmochim. Acta 1984, 48, 395–399.
26. Poland, J.F.; Piper, A.N. Groundwater Geology of the Coastal Zone, Long Beach-Santa Ana Area,
California; Water-Supply Paper 1109; U.S. Geological Survey: Reston, VA, USA, 1956.
27. Yerkes, R.F.; McCulloh, T.H.; Schoellhamer, J.E.; Vedder, J.G. Geology of the Los Angeles Basin,
California: An Introduction; Professional Paper 420-A; U.S. Geological Survey: Reston, VA,
USA, 1965.
28. Davis, T.L.; Namson, J.; Yerkes, R.F. A cross section of the Los Angeles area: Seismically
active fold and thrust belt, the 1987 Whittier Narrows earthquake and earthquake hazard.
Geophys. Res. 1989, 94, 9644–9664.
29. Wright, T.L. Structural geology and tectonic evolution of the Los Angeles Basin, California. In
Active Margin Basins; Memoir 52; Biddle, K.T., Eds.; American Association of Petroleum
Geologists: Tulsa, OK, USA, 1991; pp. 35–134.
30. Ponti, D.J.; Ehman, K.D.; Edwards, B.D.; Tinsley, J.C., III; Hildenbrand, T.; Hillhouse, J.W.;
Hanson, R.T.; McDougall, K.; Powell, C.L., II; Wan, E.; et al. A 3-Dimensional Model of
Water-Bearing Sequences in the Dominguez Gap Region, Long Beach, California; Open-File
Report 2007-1013; U.S. Geological Survey: Reston, VA, USA, 2007.
142
31. Nishikawa, T.; Siade, A.J.; Reichard, E.G.; Ponti, D.J.; Canales, A.G.; Johnson, T.A.
Stratigraphic controls on seawater intrusion and implications for groundwater management,
Dominguez Gap area of Los Angeles, California, USA. Hydrol. J. 2009, 17, 1699–1725.
32. Woodring, W.P.; Bramlette, M.N.; Kew, W.S.W. Geology and Paleontology of Pales Verdes
Hills, California; Professional Paper 207; U.S. Geological Survey: Reston, VA, USA, 1946.
33. Epstein, S.; Mayeda, T. Variation of O18 content of water from natural sources. Geochim.
Cosmochim. Acta 1953, 4, 213–224.
34. Coplen, T.B.; Wildman, J.D.; Chen, J. Improvements in the gaseous hydrogen-water
equilibration technique for hydrogen isotope ratio analysis. Anal. Chem. 1991, 63, 910–912.
35. Burnett, W.C.; Taniguchi, M.; Oberdorfer, J.A. Assessment of submarine groundwater
discharge into the coastal zone. J. Sea Res. 2001, 46, 109–116.
36. Burnett, W.C.; Dulaiova, H. Estimating the dynamics of groundwater input into the coastal zone
via continuous radon-222 measurements. J. Environ. Radio. 2003, 69, 21–25.
37. Swarzenski, P.W.; Orem, W.G.; McPherson, B.F.; Baskaran, M.; Wan, Y. Biogeochemical
transport in the Loxahatchee river estuary: The role of submarine groundwater discharge.
Mar. Chem. 2006, 101, 248–265.
38. Swarzenski, P.W. U/Th series radionuclides as tracers of coastal groundwater. Chem. Rev. 2007,
107, 663–674.
39. Moore, W.S.; Arnold, R. Measurements of 223Ra and 224Ra in coastal waters using a delayed
coincidence counter. J. Geophys. Res. 1996, 101, 1321–1329.
40. Moore, W.S. Sampling 228Ra in the deep ocean. Deep Sea Res. Oceanogr. Abstr. 1976, 23,
647–651.
41. Drever, J.I. The Geochemistry of Natural Waters, 2nd ed.; Prentice Hall: Upper Saddle River,
NJ, USA, 1988.
42. Back, W.; Hanshaw, B.B.; Pyler, T.E.; Plummer, L.N.; Weide, E. Geochemical significance of
groundwater discharge in Caleta Xel Ha, Quintana Roo, Mexico. Water Resour. Res. 1979, 15,
1521–1535.
43. Ponti, D. U.S. Geological Survey, Menlo Park, CA, USA. Personal communication, September
2012.
44. Izbicki, J.A. Source, Movement, and Age of Groundwater in a Coastal California Aquifer; Fact
Sheet FS126-96; U.S. Geological Survey: Reston, VA, USA, 1996.
45. Michel, R.L. Tritium Deposition in the Continental United States, 1953–1989; Water-Resources
Investigations Report 89-4072; U.S. Geological Survey: Reston, VA, USA, 1989.
46. Craig, H. Isotopic variations in meteoric waters. Science 1961, 133, 1702–1803.
47. Dansgaard, W. Stable isotopes in precipitation. Tellus 1964, 16, 436–468.
48. Williams, A.E.; Rodoni, D.P. Regional isotope effects and application to hydrologic investigations
in southwestern California. Water Resour. Res. 1997, 33, 1721–1729.
49. Taylor, C.B. On the isotopic composition of dissolved inorganic carbon in rivers and shallow
groundwater: A diagrammatic approach to process identification and a more realistic model of
the open system. Radiocarbon 1997, 39, 251–268.
143
50. Claypool, G.E.; Kaplan, I.R. The origin and distribution of methane in marine sediments.
In Natural Gases in Marine Sediments; Kaplan, I.R., Ed.; Springer: Berlin, Germany, 1974;
pp. 99–139.
51. Grossman, E.L. Stable carbon isotopes as indicators of microbial activity in aquifers. In Manual
of Environmental Microbiology; Hurst, C.J., Ed.; American Society for Microbiology:
Washington, DC, USA, 1997; pp. 565–576.
52. Hellings, L.; van den Driessche, K.; Baeyens, W.; Keppens, E.; Dehairs, F. Origin and fate of
dissolved inorganic carbon in interstitial waters of two freshwater intertidal areas: A case study
of the Scheldt estuary, Belgium. Biogeochemistry 2000, 51, 141–160.
53. Landmeyer, J.E.; Vroblesky, D.A.; Chapelle, F.H. Stable carbon isotope evidence of
biodegradation zonation in a shallow jet-fuel contaminated aquifer. Environ. Sci. Technol. 1996,
30, 1120–1128.
54. Conrad, M.E.; Daley, P.F.; Fischer, M.L.; Buchanan, B.B.; Kashgarian, M. Combined 14C and
į13C monitoring of in situ biodegradation of petroleum hydrocarbons. Environ. Sci. Technol.
1997, 31, 1463–1469.
55. Marfia, A.M.; Krishnamurthy, R.V.; Atekwana, E.A.; Panton, W.F. Isotopic and geochemical
evolution of ground and surface waters in a karst dominated geologic setting: A case study from
Belize, Central America. Appl. Geochem. 2004, 19, 937–946.
56. Nascimento, C.; Atekwana, E.A.; Krishnamurthy, K.V. Concentrations and isotope ratios of
dissolved inorganic carbon in denitrifying environments. Geophys. Res. Lett. 1997, 24, 1511–1514.
57. Rosenthal, E.; Vinokurov, A.; Magaritz, M.; Moshkovitz, S. Anthropogenically induced
salinization of groundwater: A case study from the coastal plain aquifer of Israel. Contamin.
Hydrol. 1992, 11, 149–171.
58. Plummer, L.N.; Parkhurst, D.L.; Thorstenson, D.C. Development of reaction models for
ground-water systems. Geochim. Cosmochim. Acta 1983, 47, 665–686.
59. Kigoshi, K. Alpha recoil 234Th: Dissolution in water and the 234U/238U disequilibrium in nature.
Science 1971, 173, 47–48.
60. Tricca, A.; Wasserburg, G.J.; Porcelli, D.; Baskaran, M. The transport of U- and Th-series
nuclides in a sandy confined aquifer. Geochim. Cosmochim Acta 2001, 65, 1187–1121.
61. Turekian, K.K.; Wedepohl, K.H. Distribution of the elements in some major units of the earth’s
crust. GSA Bull. 1961, 72, 175–192.
62. Swarzenski, P.W.; Reich, C.D.; Spechler, R.M.; Kindinger, J.L.; Moore, W.S. Using multiple
geochemical tracers to characterize the hydrogeology of the submarine spring off Crescent
Beach, Florida. Chem. Geol. 2001, 179, 187–202.
63. Krishnaswami, S.; Seidemann, D.E. Comparative study of 222Rn, 40Ar, 39Ar, and 37Ar leakage
from rocks and minerals—Implications for the role of nanopores in gas transport through natural
silicates. Geochim. Cosmochim. Acta 1988, 52, 655–658.
64. Rama; Moore, W.S. Submicronic porosity in common minerals and emanation of radon.
Nucl. Geophys. 1990, 4, 467–473.
65. Davidson, M.R.; Dickson, B.L. A porous flow model for steady state transport of radium in
groundwater. Water Resour. Res. 1986, 22, 34–44.
144
66. Martin, P.; Akber, R.A. Radium isotopes as indicators of adsorption–desorption interactions and
barite formation in groundwater. J. Environ. Radioact. 1999, 46, 271–286.
67. Luo, S.; Ku, T.-L.; Roback, R.; Murrell, M.; McLing, T.L. In-situ radionuclide transport
and preferential groundwater flows at INEEL (Idaho): Decay-series disequilibrium studies.
Geochim. Cosmochim. Acta 2000, 64, 867–881.
68. Hammond, D.E.; Zukin, J.G.; Ku, T.-L. The kinetics of radioisotope exchange between brine
and rock in a geothermal system. J. Geophys. Res. 1988, 93, 13175–13186.
145
Submarine Groundwater Discharge at a Single Spot Location:
Evaluation of Different Detection Approaches
Michael Schubert, Jan Scholten, Axel Schmidt, Jean François Comanducci,
Mai Khanh Pham, Ulf Mallast and Kay Knoeller
Abstract: Submarine groundwater discharge (SGD) into the ocean is of general interest because it
acts as vehicle for the transport of dissolved contaminants and/or nutrients into the coastal sea and
because it may be accompanied by the loss of significant volumes of freshwater. Due to the
large-scale and long-term nature of the related hydrological processes, environmental tracers are
required for SGD investigation. The water parameters of electrical conductivity and temperature, the
naturally occurring radionuclides of radon and radium as well as the stable water isotopes 18O and
2
H have proven in previous studies their general suitability for the detection and quantification of
SGD. However, individual hydrogeological settings require a site-specific application of this “tool
box”. This study evaluates and compares the applicability of the abovementioned tracers for
investigating SGD from a distinct submarine source in a karst environment at Cabbé, southern
France. The specific advantages and disadvantages of each individual parameter under the given
hydrogeological conditions are discussed. Radon appeared to be the most suitable environmental
tracer in the site specific context. The water temperature was less reliable due to the little temperature
difference between seawater and groundwater and since the diurnal variation of the air temperature
masks potential SGD signals. Radium isotopes are less applicable in the studied region due to the
lack of a well-developed subterranean estuary. The stable water isotopes showed results consistent
with the salinity and radon data; however, the significantly higher effort required for stable isotope
analyses is disadvantageous. A multi-temporal thermal remote sensing approach proved to be a
powerful tool for initial SGD surveying.
Reprinted from Water. Cite as: Schubert, M.; Scholten, J.; Schmidt, A.; Comanducci, J.F.; Pham, M.K.;
Mallast, U.; Knoeller, K. Submarine Groundwater Discharge at a Single Spot Location: Evaluation
of Different Detection Approaches. Water 2014, 6, 584-601.
1. Introduction
Along the coastlines worldwide, terrestrial waters discharge continuously into the coastal sea.
A key role in the related water budget is played by submarine groundwater discharge (SGD), i.e.,
“the flow of water through the continental margin from the seabed to the coastal ocean, occurring
regardless of fluid composition or driving force with scale lengths of meters to kilometres” [1].
Due to two general facts SGD is of major interest for coastal water resources management: (1) it
provides transport for contaminants and/or nutrients, thereby potentially threatening marine
ecosystem health; (2) it may cause a loss of substantial volumes of freshwater to the ocean. With
regard to the first, several studies have shown that dissolved material transport associated with SGD can
be of the same order of magnitude as material transport associated with river discharge and surface
water runoff [2,3]. With regard to the second it can be stated, that particularly spots of focussed SGD
146
have the so far unexploited potential to be used as important freshwater sources in arid climate zones
and/or other areas that are characterized by water scarcity (e.g., islands in the Mediterranean Sea).
The overall water resources management context of SGD is not only of relevance in the coastal ocean
but also linked to the phenomenon offshore fresh groundwater reserves as comprehensively discussed
by Post et al. [4].
In comparison to the investigation of river water discharge into the sea, which allows
straightforward quantification of discharge rates and material budgets, the localization and
investigation of SGD is more complex. Besides the difficulty that SGD can occur as both, diffuse
seeps across wide patches of sea floor and focussed flow emerging from a distinct submarine spring,
it is the high temporal variability that complicates the localisation of SGD areas and the assessment
of the associated environmental impacts. The main drivers causing temporal fluctuations in SGD are
(1) seasonal and long-term, changes in the inland hydraulic head; (2) tidal-driven, i.e., diurnal
changes in pressure gradients; and (3) short-term non-cyclical wave or current-induced pressure
gradients [5].
A considerable number of scientific studies during the last two decades have investigated
occurrences of SGD in different environmental settings [3,6–14]. These studies revealed that indepth understanding of SGD and related processes demands approaches that apply environmental
tracers, i.e., naturally-occurring hydrochemical indicators or dissolved tracers that show substantial
gradients at the groundwater/seawater interface.
Indicators that are frequently used in this regard include the water parameters specific electrical
conductivity (EC)/salinity and the naturally-occurring radioactive noble gas radon. The two
parameters show the advantage of straightforward on-site detectability [10,15]. The radionuclide
radon-222 (222Rn, hereinafter referred to as radon) is of particular suitability in the given context,
since it is short-lived (half-life 3.82 d) and chemically inert, which makes it a preferable dissolved
tracer in the field of hydrological sciences [13,16–20].
Besides the two on-site detectable parameters EC and radon, the four naturally-occurring radium
isotopes 226Ra, 228Ra, 223Ra, and 224Ra are widely used as environmental tracers for the investigation
of groundwater/seawater water interaction and coastal water mixing [21,22]. While radium is mainly
particle-bound in freshwater (i.e., not in solution), it desorbs from the mineral matrix if the groundwater
salinity rises as is the case in the mixing zone between seawater and freshwater, the so-called
“subterranean estuary” [23]. Because the four radium isotopes cover a wide range of half-lives
(223Ra: t1/2 = 11.4 days; 224Ra: t1/2 = 3.7 days; 228Ra: t1/2 = 5.75 years; 226Ra: t1/2 = 1600 years) mixing
processes of discharging groundwater with seawater can be investigated on different spatial and
temporal scales.
Also suitable as SGD indicators are the isotopic signatures of į18O and į2H, which differ in general
significantly between groundwater and surface water bodies [24,25]. Both isotopes can be used as
tracers for describing mixing ratios between waters masses of different origin [26].
The presented study was carried out at Cabbé within the Bay of Roquebrune close to Monaco
(Figure 1). The site is characterized by focussed groundwater discharge from a major karst aquifer
at a distinct submarine spring into the Mediterranean Sea. Cabbé was chosen as study site because it
147
represents an archetypal and easily accessible site representative for coastal zones that are dominated
by karst aquifers (as in large parts of the northern Mediterranean).
Figure 1. Small-scale site morphology; the white circle highlights the vertical submarine
cave structure. Ropes were used for supporting probes and sampling tubes.
In karst environments large portions of the occurring precipitation recharge the local aquifer
directly without large temporal delay, which in turn discharges mainly as focussed SGD into the sea.
Zektser et al. [27] estimated a SGD flux to the Mediterranean Sea of about 68 km³/a, which
corresponds to one third of the overall river discharge to the Mediterranean Sea. At the Cote d´Azur,
France, between Nice and Menton over 50 sites of focussed SGD have been localized, which, based
on local hydrogeological water mass balances, discharge in total about 18.9 × 106 m3aí1 of freshwater
to the sea [28]. However, the significance of this focused SGD for chemical input to the coastal sea
remains so far unclear.
The objective of this study was to evaluate and compare the informative value of the
abovementioned parameters under the specific conditions of a typically focussed SGD from a
submarine karst spring. The investigated tracers included the stable isotopes of water (2H and 18O),
naturally-occurring radioactive isotopes (222Rn, 228, 226, 224, 223Ra) as well as physical (temperature) and
chemical water parameters (salinity). Temporal and spatial patterns of all tracers were recorded over
a 24 h period at Cabbé and along a 1.75 km transect located within the Bay of Roquebrune.
In addition, the applicability of a multi-temporal thermal remote sensing approach [29] was evaluated
for supporting the tracer results (and vice versa).
2. Methods
2.1. Site Setting
The submarine spring Cabbé is located in the north-western Mediterranean Sea about 5 km NE of
Monaco (Figure 2). It is part of a group of several submarine springs mainly occurring along the
coast between Nice and Menton, southern France [30]. At Cabbé groundwater discharges in a water
148
depth of about 4 m from the base of a vertical submarine cave with a diameter of approximately 3 m
(Figure 1). The cave is surrounded by limestone forming a vertical hole that is constantly inundated
by seawater. It is hydraulically connected to a major karst aquifer. Groundwater residence times in
this karst aquifer were estimated to be between eight to nine days [30] (cf. Section 3.1.6). The yearly
average temperature in the region is ~15 °C with an average of 770 mm of precipitation.
Figure 2. Regional setting of the site and the measured transect (A) within the
Mediterranean Sea; (B) in geographical relation to Monaco.
Geologically the area is part of the SE front of a subalpine range called “Arc de Nice”, which
consists of a series of south-verging folds and thrusts involving Mesozoic and Palaeogene sediments.
The sediments consist of marls, clays, dolomites and limestones. Important aquifers with deep karstic
circulation are associated with the latter. Coastal and submarine groundwater discharge spots are
generally situated at the contact between the Jurassic limestone and the Cretaceous marls and clays.
2.2. Materials and Methods
At the base of the submarine vertical cave, i.e., at the submarine spring, the parameters
temperature, salinity, radon concentration, and water depth were monitored using automated
detection equipment. All instruments were working properly from about 16:00 on the first day of the
24 h monitoring campaign and measured continuously for 24 h. A recording gap occurred on the
second day between 03:00 and 05:00 due to technical difficulties. In addition to the recording of 24 h
time series, discrete water samples were taken every 60 min for stable isotope analyses. Four discrete
samples for radium analyses were obtained on the second day covering a period between low and
high tide.
149
Additionally to the time-series measurements at the submarine spring, a survey was undertaken
along a 1750 m transect within the Bay of Roquebrune starting at the submarine cave Cabbé. Along
the transect the parameters salinity, temperature and radon concentration were monitored continuously
while cruising with a speed of about 5 km/h. For stable isotopes and radium measurements discrete
samples were taken along the transect.
For the measurement of conductivity/salinity, temperature, and water depth a CTD probe (YSI
Inc., Yellow Springs, OH, USA) was used. At the spring the CTD was lowered into the cave with
ropes (cf. Figure 1) and fixed close to the bottom of the vertical cave structure. The recorded water
depth was used for monitoring the tidal cycle and for assessing the water turbulence within the cave
(wave intensity). Along the transect the CTD was dragged behind the boat in a water depth of about
2 m for temperature and salinity recording. All CTD data were recorded in 10 second intervals. For
comparison with the other monitored parameters, suitable moving average values were calculated.
For measurement of radon-in-water concentrations the portable radon-in-air monitor RAD-7
(Durridge Company, Billerica, MA, USA) was used. For radon detection, water was pumped
either from the surface sea water (transect) or from the bottom part of the submarine cave. Radon
was extracted from the constant water stream into a closed air loop by means of a RAD-Aqua
extraction-module as described in detail by e.g., Schmidt et al. [15]. In order to achieve radon
measurements with uncertainties of about ±10% two RAD-7s were run in parallel with a 30 min
counting cycle (30 min time interval between actual readings). The radon concentration that was
measured in the air loop was converted into the related radon-in-water concentration by applying
the temperature and salinity dependent water/air partition coefficient Ka/w of radon [31]. Thus
a continuous time series with 30 min detection intervals was achieved.
For the analysis of radium concentrations (223, 224, 226, 228Ra), 70 litres of water were pumped either
from the bottom part of the submarine cave or from surface sea water (~2m water depth) along
the transect. The water was pumped over MnO2-fibre cartridges with a pump rate of about 1 l/min
in order to scavenge the radium quantitatively [8]. The short-lived 223Ra and 224Ra were measured
using the RaDeCC delayed coincidence counting system [32] following the procedure described in
Moore [8] and Scholten et al. [33]. For the long-lived 226Ra and 228Ra measurements the MnO2-fibres
were ashed and prepared for gamma spectrometry measurements [21].
Sampling for stable water isotopes (2H and 18O) was carried out every 60 min from the bottom
part of the submarine spring. Along the transect five discrete samples were taken at equidistant
locations. All stable isotope samples were collected in 30 mL HDPE bottles and were measured using
a continuous flow high temperature pyrolysis unit combined with an isotope ratio mass spectrometer
(IRMS) delta plus XL Thermo Finnigan [34]. The isotopic composition of all samples was
determined by the H2O-H2 equilibration method (for 2H) with an analytical precision of ±1.0‰ and
the H2O-CO2 equilibration method (for 18O) with an analytical precision of ±0.1‰. The results of
the hydrogen and oxygen isotope measurements are expressed as delta notations (δ2H, δ18O) relative
to the Vienna Standard Mean Ocean Water (VSMOW).
150
The applied multi-temporal thermal remote sensing approach was based on eight Landsat ETM+
band 6.2 (high gain) images (path 194/row 30). All images were recorded during the hydrological
summer (May-October) of the years 2000–2009 at 12:00 a.m. local time (GMT+2). Each image,
which exhibits a ground sampling distance (GSD) of 30 m, was co-registered to UTM WGS 84 Zone
32N. For images recorded after the 31 May 2003, the gaps caused by SLC (scan line corrector) failure
of Landsat were filled using the triangulation method implemented in the Envi 4.7 software package.
Further data processing included the conversion from digital numbers to sea-surface temperatures
(SST) including an atmospheric correction of each image. The applied value for water emissivity is
0.98. Values for atmospheric transmissivity, upwelling and downwelling radiances needed for the
atmospheric correction of each image were obtained through the web-based Atmospheric Correction
Tool that is based on MODTRAN [35]. The identification of SGD locations was based on the
assumption that temperatures of discharging groundwater are less variable than ambient seawater
temperatures. Hence the standard deviation of the temperature per pixel of the SST-image time-series
was calculated and used as indicator [29].
3. Results and Discussion
3.1. Cabbé Submarine Spring
3.1.1. Water Turbulence and Tidal Cycle
The water turbulence was recorded because it has a strong impact on the conditions in the vertical
cave, namely on (i) the degree of radon degassing from the water; (ii) the intensity of SGD/seawater
mixing; and (iii) the temperature of the seawater. In particular at monitoring points that are located
close to a rocky shore, as is the case at Cabbé, water turbulence must not necessarily correspond to
the actual wind speed since it mainly results from the surf breaking on the rocky shore.
Figure 3 illustrates the wave intensity by showing the short-time deviation of the one minute
running average water depth recorded in 10 second steps. With regard to water turbulence the 24 h
monitoring period can be divided into two subsections: during the beginning of the sampling
campaign, on the first day between 16:00 and 18:00, very turbulent waters were observed. Thereafter,
in particular between 12:00 and 16:00 on the second day, the sea became significantly calmer. A data
gap between about 03:00–05:00 occurred due to malfunction of the detection equipment.
The recorded water depths were also used for determination of the tidal cycle. The tidal cycle is
a major driver for periodic changes in the SGD rate: during low tide the hydraulic gradient between
land and sea is higher compared to high tide. This steeper gradient allows more groundwater to
discharge from the hydraulically connected aquifer [6]. Figure 4 shows the water depth above the
bottom of the submarine cave during the 24 h monitoring campaign.
151
Figure 3. Wave intensity shown as deviation of the water depth (WL: water depth).
Figure 4. Water column height over the Cabbé submarine spring (i.e., water depth) in
20 min intervals (white diamonds); grey points display individual 10 second readings.
Despite the data gap, tidal periods with low tide at around 16:00 and 04:00 and high tide at around
23:00 and 11:00 can clearly be distinguished. However, even though the data show distinct flood/ebb
periods, the water depths recorded during the first four hours (16:00–18:00) were considerably lower
than during later ebb stages. That effect is believed to be due to two reasons: (1) strong winds coming
off the land at that time pushed water out of the bay resulting in a generally decreased seawater level;
(2) the water in the vertical cave, due to the high turbulence (cf. Figure 3), was highly enriched with
air bubbles reducing the bulk water density and hence the hydrostatic pressure of the water column
over the CTD (the CTD applies a pressure transducer for water depth detection). Thus, water depth
readings for the period between 16:00 and 18:00 are probably too low. The changing conditions
reflected in Figure 3 do affect tracer distribution behavior in the coastal sea in general. Hence
the recorded tracer data show a strong dependence on the water turbulence as it will be revealed in
the following.
152
3.1.2. Water Temperature
As illustrated in Figure 5A a distinct positive correlation between temperature and water
depth was observed during the second day of the sampling campaign (calm sea). A coefficient of
determination of R2 = 0.766 (N = 20) was determined. This relation is caused by the changing
hydrostatic pressure on the aquifer correlated to the changing tides. The discharging groundwater
has a somewhat lower temperature (~17.5 °C; [30]) than the coastal sea water (~22 °C) thus leading
to water temperature decrease in the cave at low tide. However, during the first hours of the campaign
(rough sea; low tide) that correlation was seemingly reversed. In spite of the low water depth
(and the supposedly associated increased discharge of cool groundwater) the water temperature
showed the highest values recorded during the entire campaign. This observation is explained to be
as a result of the rough sea (before 18:00, cf. Figure 3). The strong surf resulted in a substantial
amount of (warm) air bubbles whirling deep in the studied water column, giving rise to a significant
increase of the water temperature during that time (the outside air temperature was measured to be
around 25 °C that afternoon).
In conclusion it can be stated that whereas in several other studies the water temperature could be
applied as a conservative SGD tracer [36], its applicability is limited in regions where the temperature
gradient between groundwater and seawater is too small to allow discrimination between diurnal
fluctuations of the seawater temperature and water temperature fluctuations caused by SGD. Besides
it was shown that water turbulence can have a significant influence on the seawater temperature.
3.1.3. Salinity
As shown in Figure 5B a distinct positive relation between salinity and water depth was observed
during the second part of the sampling campaign (calm sea). As expected, high water depths
(high tides at 23:00 and 11:00) resulted in increased salinities, an effect that can be attributed to the
decreased SGD rate at high tide. However, during the period characterized by rough conditions in
the beginning of the campaign (16:00–18:00) this relationship seems to be reversed. In spite of low
water depths and related increased SGD rates the recorded salinities were rather comparable to the
respective data recorded during the following high tides. It is most likely that the reason for this
reversed relationship is the high turbulence of the water during the first hours of monitoring resulting
in a more intense exchange of seawater with the discharging groundwater in the submarine cave.
Even though the recorded wave heights are maximally 0.1 m, the actual breaking of the waves into
the mouth of the cave caused visibly heavy turbulences in the water column of the cave resulting in
the recorded inverse trend in salinity (and temperature). Another potential explanation is that
seawater, which was pushed into the conduit system earlier during the rough sea conditions,
discharged again at low tide after the sea started to calm down.
In terms of its applicability as SGD tracer, salinity can generally be considered more suitable than
temperature. Besides the fact that it is not influenced by atmospheric conditions, the salinity gradient
at the groundwater/seawater interface is much more distinct than temperature gradients. While offshore seawaters generally show salinities of around 35 (38.06 in the Bay of Roquebrune)
groundwater shows in general negligible or lower salinities (7.2 at Cabbé, see Table 1).
153
Figure 5. (A) Water depth and temperature; (B) Salinity; (C) Radon concentration;
(D) Delta 2H; and (E) Delta 18O during the 24 h observation period. LT = low tide;
HT = high tide.
154
Table 1. Groundwater and off-shore seawater end-members for all applied parameters.
Parameter
Groundwater
Off-shore seawater
salinity
~7.2
38.06
222
Rn
28,000 Bq/m3
<4 Bq/m3
į18O
í4.7‰ *
1.22‰
2
įH
í35.1‰ *
10.4‰
Note: * derived from long-term weighted annual mean rain water composition from the local
GNIP station in Monte Carlo/Monaco (GNIP station code 769001).
3.1.4. Dissolved Radon Concentrations
Due to the lack of radon production, seawater shows generally a radon background of only a
few Bq/m3. Groundwater concentrations, in contrast, can reach levels of up to 20 kBq/m3 and more.
The strong gradient at the groundwater/seawater interface results in an inverse correlation between
seawater depth and radon concentration in the coastal sea, i.e., in high concentrations during low
tides (due to intense SGD) and vice versa [37].
Figure 5C illustrates the radon data recorded at the base of the submarine cave. As displayed, the
expected inverse correlation between radon and water depth was observed from 22:00 until the end
of the monitoring campaign (calm sea). For the second day of the campaign a coefficient of
determination of R2 = 0.624 (N = 20) was found. At the beginning of the campaign however (rough
sea), this relationship appears to be reversed. Low water depths are associated with low radon
concentrations in the water column. The most likely reason for this unusual radon pattern is the high
water turbulence (as mentioned above). While radon is degassing from every open water surface,
this gas evasion is strongly enforced if the water is very turbulent and contains many air bubbles.
(The rough sea resulted also in intense water mixing between seawater and discharging
groundwater). Radon degassing is due to the water/air partition coefficient of radon which is
Kw/a = 0.25 at the given temperature [31]. Comparable to the interpretation of the salinity data another
potential explanation is that seawater, which was pushed into the conduit system earlier, discharged
again after the sea started to calm down.
For the applicability of radon as SGD tracer it can hence generally be stated that radon reveals
robust data due to its distinct gradient at the groundwater/seawater interface, due to its inert behaviour
and due to its straightforward detectability on site. However, in the case of rough seas intense radon
degassing may limit its applicability as a tracer.
3.1.5. Dissolved Radium Activity
At Cabbé, radium isotope concentrations do not show clear variations with tides. Also the isotope
ratios determined along the transect are in the same range as observed at Cabbé (Figure 6). If SGD
at Cabbé were a source of radium one would expect decreasing 224Ra/228Ra and/or 228Ra/226Ra values
with increasing distance from Cabbé.
The lack of radium enrichment in the groundwater discharging at Cabbé is in contrast to several
other SGD studies in the Mediterranean Sea where radium has been successfully applied as SGD
tracer [38–40]. We explain the lack of radium in the discharging groundwater to be due to two main
155
reasons: (1) The SGD site studied is hydraulically connected to a major karst aquifer. In such oxic
groundwaters radium adsorbs on mineral surfaces and therefore dissolved concentrations are very
low. (2) In general radium is released from aquifer mineral surfaces in the mixing zone between
seawater and groundwater (subterranean estuary) where increased salinities are present [41].
In contrast to many other settings in the Mediterranean Sea there is no well-developed subterranean
estuary at Cabbé. The groundwater discharges directly into the cave without an aquifer zone where
radium can be mobilized from the aquifer matrix.
Due to the specific hydrological setting, radium has to be considered less suitable for SGD
investigation at Cabbé. This may also be the case for other focused SGD locations having a poorly
developed subterranean estuary.
Figure 6. Radium isotope ratios at Cabbé and along the transect. (A) 224Ra/228Ra; (B) 228Ra/226Ra.
(A)
(B)
3.1.6. Stable Water Isotopes
As shown in Figure 5D,E the stable isotope signatures of the water show a distinct positive relation
with the tides, i.e., the water depth in the submarine cave. Only during the first hours of the
observation period a seemingly inverse relationship was found (as already discussed for the
other parameters).
The reason for the generally lighter isotopic signatures at low tide is the significant difference in
isotopic signatures between seawater and meteoric groundwater. Lighter signatures at low tide
indicate increased discharge of isotopically lighter groundwater into the sea. A shift to heavier
isotope signatures can be observed at high tides since the water column is less influenced by
groundwater at this stage. The inverse behaviour at the beginning of the observation period is most
likely caused by intense water mixing between the near-shore seawater and the SGD fed water
column in the submarine cave. The strong impacts of the water turbulence and of potentially
discharging seawater that was pushed into the conduit system earlier on the stable isotope signatures
during the first hours of recording are comparable to the impact on salinity and radon as
discussed above.
156
3.2. Transect
Whereas all parameters recorded at the Cabbé SGD location (submarine spring) were discussed
individually as time series, the data recorded along the transect are discussed jointly in the following.
The transect started at a distance of about 50 m from the submarine cave at Cabbé and ended
ca. 1750 m off-shore in the Bay of Roquebrune. Figure 7 displays the continuously recorded salinity
data as 20 min average values. The salinity increases from ~37 at the coastal end of the transect to
~38 at a distance of 600 m off-shore from where it remains constant with increasing distance.
Although the absolute gradient is small, it can be seen that SGD has significant influence on
a rather large area within the Bay of Roquebrune. It has to be mentioned that no further water influx
(surface water, precipitation) occurred during the field campaign, which allows attributing the
decreasing salinity solely to the SGD influence.
Figure 7. Water salinity and temperature along the transect.
The water temperatures recorded along the transect show a rather indistinct picture with
temperatures of about 22.4–22.5 °C close to the shore and later decrease towards the open ocean to
a rather constant 22.0 °C. The temperatures measured close to the shore are higher than the highest
temperature observed during time-series measurements at Cabbé (maximum temperature 22.0 °C).
Thus the temperature variations observed along the transect are most likely not related to SGD at
Cabbé. This suggests that temperature is not applicable as a parameter for SGD evaluation at the
investigated site.
The stable isotope signatures for į18O and į2H display a pattern along the transect comparable to
salinity (Figure 8). While į18O and į2H values are constant off-shore both isotopes show a shift to
lighter signatures from around 1000 m towards the SGD location. These changes are caused by the
discharging groundwater which shows significantly lighter isotopic signatures than the open sea as
discussed above.
Radon shows an inverse behaviour as expected. The concentration decreases from 250 Bq/m3 in
the vicinity of the SGD spot, here clearly reflecting the SGD influence, to 4 Bq/m3 off-shore.
157
Figure 8. Radon concentrations and water stable isotope signatures along the transect;
(A) į18O; (B) į2H.
3.3. SST Evaluation
The sea surface temperature (SST) standard deviations within the Bay of Roquebrune are
illustrated in Figure 9. The image shows values ranging between 1.8 and 6.0 °C. Areas with high
standard deviations indicate varying influence due to e.g., seasonal temperature effects. Areas with
low sea-surface temperature variability located close to the shore indicate spatially and thermally
persistent groundwater inflow, which stabilizes the seawater temperature in the vicinity of the
inflow location.
Figure 9. Sea Surface Temperature (“SST”) standard deviation over time within the Bay
of Roquebrune (note that small STD values in the proximity of the shore indicate
potential Submarine Groundwater Discharge (“SGD”) abundance while lower STD
values offshore indicate a small SST variability only, caused by the water depth and
related high heat capacity of the water column).
158
Within the Bay of Roquebrune several areas with small standard deviation values <~2 °C are
visible along the coastline. The three most prominent ones are located (i) south of Dondéa; (ii) east
of Roquebrune-Cap-Martin at Grimaldi and (iii) at the investigated study site at Cabbé. The latter
area has a lateral extent of ~300 m and, unlike the larger area south of Dondéa, no fringe of higher
SST standard deviation values in between the actual anomaly and the shoreline. This confirms that
groundwater discharges at the shore at Cabbé.
Following the transect from the shore into off-shore direction the SST standard deviation values
remain below ~2.3 °C up to a distance of ~200 m from Cabbé. There they increase first abruptly and
rise thereafter steadily up to a distance of about 600 m off-shore. The elevated off-shore SST standard
deviation values are in accordance with the salinity and radon distribution patterns, which likewise
indicate decreasing influence of SGD with increasing distance from Cabbé. Hence it can be
stated that, even though the satellite-based approach relies on images recorded long before the
actual sampling campaign, the data proves the submarine groundwater discharge at Cabbé to be
spatio-temporal constant and provides thus an independent tool for SGD investigation.
3.4. Quantification of Groundwater/Seawater Ratios at Cabbé
A simple two-component mixing equation between end-member concentrations (off-shore water
and groundwater) of stable isotopes, radon and salinity was applied for quantifying the relative
proportions of groundwater and seawater at Cabbé. For the seawater, end-member values measured
at the end of the transect furthest distant from shore were used. Since no groundwater wells were
accessible in the immediate vicinity of Cabbé, end-members for radon and salinity were obtained
from a well situated directly at the beach at ~500m distance to Cabbé within the same
geological/hydrogeological setting. The terrestrial į18O and į2H end-members were derived from the
long-term weighted annual mean rain water composition for the closest GNIP (Global Network of
Isotopes in Precipitation) station in Monaco (GNIP station code 769001). The calculation of the
annual mean is based on 112 monthly isotope values obtained between 1999 and 2009 [42]. No rain
samples were taken in the course of the sampling campaign since no rain had occurred during the study.
Table 1 summarizes the end-members for all the parameters used to calculate the percentage of
meteoric water in the water column of the Cabbé submarine cave (Figure 10). Generally the results
of the SGD indicators give a consistent picture. The fraction of meteoric groundwater in the vertical
cave at Cabbé varies between about 1.5%–4.5% for the radon, between 2.0% and 4.0% for the
salinity, and between 3.0%–5.0% for the stable isotope data, respectively.
As it can clearly be seen in Figure 10, in particular for the second part of the campaign, the
groundwater portion depends on the tidal stage. Generally high tide (HT) periods are characterized
by a lower fraction of groundwater in the water column of the vertical cave than low tides (LT).
An exception from that generally expected behaviour is the starting period of the campaign, which
was characterized by a rough sea influencing tracer distribution behaviour.
159
Figure 10. Percentage of groundwater in the water in the submarine cave at Cabbé.
4. Conclusions
We studied the applicability and the informative value of a set of environmental tracers, which
have proven in previous studies to be suitable for detecting and quantifying SGD, under the specific
conditions of a focussed SGD from a submarine karst spring at Cabbé. Our results show that
substantial qualitative and semi-quantitative information can be achieved by applying the easily
on-site detectable parameters salinity and radon concentration (222Rn). Radon can be considered as
the most robust environmental tracer for SGD in the given case even though radon degassing from
the water column has to be considered particularly with rough seas. The temperature of the water
column can only be applied as a reliable SGD indicator if the temperature gradient between
groundwater and seawater is sufficiently distinct. If it is too small as in the studied case, the diurnal
variation of the air temperature masks the potential SGD signals. In our study the naturally-occurring
radium isotopes (228, 226, 224, 223Ra), were not suitable for detecting SGD, which is interpreted to be
due to a poorly developed subterranean estuary inhibiting significant radium mobilization from the
aquifer matrix. The stable isotopes of water (2H and 18O) are suitable SGD indicators. In a mass
balance approach, stable isotopes, radon and salinity give comparable estimates of the groundwater
fraction discharging at Cabbé. The disadvantage of using stable isotopes in comparison to salinity
and radon is the significantly higher analytical effort required for stable isotope analyses.
Additionally our study showed that a multi-temporal thermal remote sensing approach applying sea
surface temperature (SST) patterns can be used as a powerful tool for backing up tracer results and,
more importantly as a SGD pre-screening method in order to optimize on-site surveys. Further
research is recommended to evaluate the applied “SGD tool box” for the investigation of offshore
fresh groundwater reserves.
Conflicts of Interest
The authors declare no conflict of interest.
160
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Moore, W.S. The effect of submarine groundwater discharge on the ocean. Annu. Rev. Mar. Sci.
2010, 2, 59–88.
Slomp, C.P.; van Cappellen, P. Nutrient inputs to the coastal ocean through submarine
groundwater discharge: Controls and potential impact. J. Hydrol. 2004, 295, 64–86.
Santos, I.R.; Niencheski, F.; Burnett, W.; Peterson, R.; Chanton, J.; Andrade, L.C.; Milani, I.;
Schmidt, A.; Knoeller, K. Tracing anthropogenic-driven groundwater discharge into a coastal
lagoon from southern Brazil. J. Hydrol. 2008, 353, 275–293.
Post, V.E.A.; Groen, J.; Kooi, H.; Person, M.; Shemin Ge, S.; Edmunds, W.M. Offshore fresh
groundwater reserves as a global phenomenon. Nature 2013, 504, 71–78.
Santos, I.R.; Eyre, B.D.; Huettel, M. The driving forces of pore water and groundwater flow in
permeable coastal sediments: A review. Estuar. Coast. Shelf Sci. 2012, 98, 1–15.
Santos, I.R.; Burnett, W.C.; Dittmar, T.; Suryaputra, I.; Chanton, J. Tidal pumping drives
nutrient and dissolved organic matter dynamics in a Gulf of Mexico subterranean estuary.
Geochim. Cosmochim. Acta 2009, 73, 1325–1339.
Peterson, R.N.; Burnett, W.C.; Taniguchi, M.; Chen, J.Y.; Santos, I.R.; Ishitobi, T. Radon and
radium isotope assessment of submarine groundwater discharge in the Yellow River delta,
China. J. Geophys. Res. Oceans 2008, 113, 1978–2012.
Moore, W.S. Fifteen years’ experience in measuring 224Ra and 223Ra by delayed-coincidence
counting. Mar. Chem. 2008, 109, 188–197.
Taniguchi, M.; Iwakawa, H. Submarine groundwater discharge in Osaka Bay, Japan. Limnology
2004, 5, 25–32.
Burnett, W.C.; Dulaiova, H. Estimating the dynamics of groundwater input into the coastal zone
via continuous radon-222 measurements. J. Environ. Radioact. 2003, 69, 21–35.
Taniguchi, M. Tidal effects on submarine groundwater discharge into the ocean. Geophys.
Res. Lett. 2002, 29, doi:10.1029/2002GL014987.
Corbett, D.R.; Dillon, K.; Burnett, W.C.; Chanton, J.P. Estimating the groundwater contribution
into Florida Bay via natural tracers 222Rn and CH4. Limnol. Oceanogr. 2000, 45, 1546–1557.
Corbett, D.R.; Burnett, W.C.; Cable, P.H.; Clark, S.B. Radon tracing of groundwater input into
Par Pond, Savannah River Site. J. Hydrol. 1997, 203, 209–227.
Lee, D.R. A device for measuring seepage flux in lakes and estuaries. Limnol. Oceanogr. 1977,
22, 140–147.
Schmidt, A.; Schlueter, M.; Melles, M.; Schubert, M. Continuous and discrete on-site detection
of radon-222 in ground- and surface waters by means of an extraction module. Appl. Radiat.
Isotopes 2008, 66, 1939–1944.
Schubert, M.; Brüggemann, L.; Schirmer, M.; Knoeller, K. Radon monitoring in wells as tool for
the estimation of the groundwater flow rate. Water Resourc. Res. 2011, 47, doi:10.1029/
2010WR009572.
161
17. Schmidt, A.; Gibson, J.J.; Santos, I.R.; Schubert, M.; Tattrie, K.; Weiss, H. The relevance of
groundwater discharge to the overall water budget of two typical Boreal lakes in Alberta/Canada
estimated from a radon mass balance. Hydrol. Earth Syst. Sci. 2010, 14, 79–89.
18. Schubert, M.; Schmidt, A.; Lopez, A.; Balcázar, M.; Paschkel, A. In-situ determination of radon
in surface water bodies by means of a hydrophobic membrane tubing. Radiat. Measur. 2008,
43, 111–120.
19. Schubert, M.; Bürkin, W.; Peña, P.; Lopez, A.; Balcázar, M. On-site determination of the radon
concentration in water samples: Methodical background and results from laboratory studies and
a field-scale test. Radiat. Measur. 2006, 41, 492–497.
20. Schubert, M.; Paschke, A.; Bednorz, D.; Bürkin, W.; Stieglitz, T. The kinetics of the water/air
phase transition of radon and its implication on detection of radon-in-water concentrations:
Practical assessment of different on-site radon extraction methods. Environ. Sci. Technol. 2012,
46, 8945–8951.
21. Charette, M.A.; Buesseler, K.O.; Andrews, J.E. Utility of radium isotopes for evaluating the
input and transport of groundwater-derived nitrogen to a Cape Cod estuary. Limnol. Oceanogr.
2001, 46, 465–470.
22. Charette, M.A.; Moore, W.S.; Burnett, W.C. Uranium- and Thorium-Series Nuclides as Tracers
of Submarine Groundwater Discharge. In Radioactivity in the Environment; Krishnaswami, S.,
Cochran, J.K., Eds.; Elsevier: Oxford, UK, 2008.
23. Moore, W.S. The subterranean estuary: A reaction zone of ground water and sea water.
Mar. Chem. 1999, 65, 111–125.
24. Craig, H. Isotopic variations in meteoric waters. Science 1961, 133, 1702–1703.
25. Epstein, S.; Mayeda, T. Variation of 18O content of waters from natural sources. Geochem.
Cosmochem. Acta 1953, 4, 213–224.
26. Schmidt, A.; Santos, I.R.; Burnett, W.; Niencheski, F.; Knöller, K. Groundwater sources in a
permeable coastal barrier: Evidence from stable isotopes. J. Hydrol. 2011, 406, 66–72.
27. Zektser, I.S.; Dzhamalov, R.G.; Everett, L.G. Submarine Groundwater; CRC Press: Boca
Raton, FL, USA, 2007.
28. Gilli, E. Etudes des Sources Karstiques Sous-Marines et Littorales des Alpes Maritimes Entre
Menton et Nice. In Ministre de l’ Environnement, Direction Régional de l’Environnement;
DREAL PACA: Marseille, France, 1997.
29. Mallast, U.; Siebert, C.; Gloaguen, R.; Friesen, J.; Rödiger, T.; Geyer, S.; Merz, R. How to
identify groundwater-caused thermal anomalies in lakes based on multi-temporal satellite data
in semi-arid regions. Hydrol. Earth Syst. Sci. 2013, 10, 4901–4949.
30. Gilli, E. Détection de sources sous-marines et précision de l’impluvium par mesure des
variations de salinité. L’exemple de la source de Cabbé-Massolins (Roquebrune-Cap-Martin,
France). Comptes Rendus de l’Académie des Sciences–Series IIA–Earth Planet. Sci. 1999, 329,
109–116.
31. Schubert, M.; Paschke, A.; Lieberman, E.; Burnett, W.C. Air-water partitioning of 222Rn and its
dependence on water salinity. Environ. Sci. Technol. 2012, 46, 3905–3911.
162
32. Moore, W.S.; Arnold, R. Measurement of 223Ra and 224Ra in coastal waters using a delayed
coincidence counter. J. Geophys. Res. 1996, 101, 1321–1329.
33. Scholten, J.; Pham, M.K.; Blinova, O.; Charette, M.A.; Dulaiova, H.; Eriksson, M. Preparation
of Mn-fiber standards for the efficiency calibration of the delayed coincidence counting system
(RaDeCC). Mar. Chem. 2010, 121, 206–214.
34. Gehre, M.; Geilmann, H.; Richter, J.; Werner, R.A.; Brand, W.A. Continuous flow 2H/1H and
18
O/16O analysis of water samples with dual inlet precision. Rapid Commun. Mass Spectrom.
2004, 18, 2650–2660.
35. Barsi, J.A.; Barker, J.L.; Schott, J.R. An Atmospheric Correction Parameter Calculator for a
Single Thermal Band Earth-Sensing Instrument. In Proceedings of IGARSS03, Centre de
Congres Pierre Baudis, Toulouse, France, 21–25 July 2003.
36. Johnson, A.G.; Glenn, C.R.; Burnett, W.C.; Peterson, R.N.; Lucey, P.G. Aerial infrared imaging
reveals large nutrient-rich groundwater inputs to the ocean. Geophys. Res. Lett. 2008,
35, L15606.
37. Tsabaris, C.; Scholten, J.; Karageorgis, A.P.; Comanducci, J.F.; Georgopoulos, D.;
Wee Kwong, L.L.; Patiris, D.L.; Papathanassiou, E. Underwater in situ measurements of
radionuclides in selected submarine groundwater springs, Mediterranean Sea. Radiat.
Protect. Dosim. 2010, 142, 273–281.
38. Ferrarin, C.; Rapaglia, J.; Zaggia, L.; Umgiesser, G.; Zuppi, G.M. Coincident application of a
mass balance of radium and a hydrodynamic model for the seasonal quantification of
groundwater flux into the Venice Lagoon, Italy. Mar. Chem. 2008, 112, 179–188.
39. Garcia-Solsona, E.; Garcia-Orellana, J.; Masque, P.; Rodellas, V.; Mejias, M.; Ballesteros, B.;
Dominguez, J.A. Groundwater and nutrient discharge through karstic coastal springs (Castello,
Spain). Biogeosciences 2010, 7, 2625–2638.
40. Rapaglia, J.; Koukoulas, S.; Zaggia, L.; Lichter, M.; Manfé, G.; Vafeidis, A.T. Quantification
of submarine groundwater discharge and optimal radium sampling distribution in the Lesina
Lagoon, Italy. J. Mar. Syst. 2012, 91, 11–19.
41. Gonneea, M.E.; Morris, P.J.; Dulaiova, H.; Charette, M.A. New perspectives on radium
behavior within a subterranean estuary. Mar. Chem. 2008, 109, 250–267.
42. IAEA/WMO. Global Network of Isotopes in Precipitation. The GNIP Database 2014. Available
online: http://www.iaea.org/water (accessed on 26 February 2014).
Section 3: Investigations Using Natural Tracers
in Combination with Applied Tracers
164
Use of Natural and Applied Tracers to Guide Targeted
Remediation Efforts in an Acid Mine Drainage System,
Colorado Rockies, USA
Rory Cowie, Mark W. Williams, Mike Wireman and Robert L. Runkel
Abstract: Stream water quality in areas of the western United States continues to be degraded by
acid mine drainage (AMD), a legacy of hard-rock mining. The Rico-Argentine Mine in southwestern
Colorado consists of complex multiple-level mine workings connected to a drainage tunnel
discharging AMD to passive treatment ponds that discharge to the Dolores River. The mine workings
are excavated into the hillslope on either side of a tributary stream with workings passing directly
under the stream channel. There is a need to define hydrologic connections between surface water,
groundwater, and mine workings to understand the source of both water and contaminants in the
drainage tunnel discharge. Source identification will allow targeted remediation strategies to be
developed. To identify hydrologic connections we employed a combination of natural and applied
tracers including isotopes, ionic tracers, and fluorescent dyes. Stable water isotopes (į18O/įD) show
a well-mixed hydrological system, while tritium levels in mine waters indicate a fast flow-through
system with mean residence times of years not decades or longer. Addition of multiple independent
tracers indicated that water is traveling through mine workings with minimal obstructions. The
results from a simultaneous salt and dye tracer application demonstrated that both tracer types can
be successfully used in acidic mine water conditions.
Reprinted from Water. Cite as: Cowie, R.; Williams, M.W.; Wireman, M.; Runkel, R.L. Use of
Natural and Applied Tracers to Guide Targeted Remediation Efforts in an Acid Mine Drainage
System, Colorado Rockies, USA. Water 2014, 6, 745-777.
1. Introduction
Acidic, metal-rich drainage from abandoned hard-rock mines can produce both acute and chronic
environmental problems [1]. The legacy of past hard-rock mining in the United States includes more
than 200,000 abandoned or inactive mines [2] with thousands of abandoned mines located near
headwater regions of the Rocky Mountains of Colorado [3]. Watersheds in mineralized regions often
receive drainage from a complex distribution of mine systems [4], which are gravity driven and often
discharge at low points adjacent to surface waters. The combination of low pH and high
concentrations of metals associated with the acid mine drainage (AMD) can then create severe
toxicological effects on local and downstream aquatic ecosystems [5]. Traditionally, an end of the
pipe (e.g., at the mine discharge point) treatment strategy has been employed to handle AMD prior
to mixing with local surface waters. However, this strategy is very expensive and treatment must
occur in perpetuity, which does not represent a permanent solution to the problem. A more recent
approach to controlling AMD involves developing targeted remediation strategies that address the
feasibility of actually reducing or shutting off the AMD at its source [1]. Targeted remediation can
be thought of as a source and pathway control measure and may refer to the source of the acid
165
producing minerals themselves, the source of the water that mobilizes contaminants, or both. In most
mine settings it is impractical to isolate or remove the mineralized rock itself. Therefore, remediation
efforts may be most successful when the source of water producing the AMD can be targeted and
separated, isolated, or removed from the area of a mine most prone to AMD production.
A major challenge in using a targeted remediation approach to control AMD is that many
hard-rock mine settings are located in mountainous areas of high mineralization that are commonly
associated with fracturing and structurally deformed rocks. The result is high secondary porosity,
which combined with the diverse spatial scales of man-made mining excavations, creates a highly
complex hydrogeologic setting where water flow paths are particularly difficult to quantify [6]. Steep
slopes and large amounts of snowfall add to the complexity of understanding the hydrology of these
sites. In addition, many of the hard-rock mining sites in the Western U.S. have been abandoned for
many decades leading to varying degrees of degradation to the originally engineered designs for mine
drainage control. Information on the mine sites is often incomplete because mine maps may be
unavailable, incomplete, or inaccurate. Therefore, the hydrologic connectivity of abandoned mines
is often unknown or poorly understood. The complexities of flow paths in flooded mines are often
comparable to flow paths found in karstic aquifers [7], where flow may be concentrated in subsurface
conduits, making a Darcy’s law approach inapplicable for evaluation of subsurface flow regimes in
flooded mines [8]. A reasonable approach to understanding the hydrology of these systems involves
applying ground-water tracing techniques to abandoned mine sites that generate empirical data while
measuring properties in-situ to minimize assumptions about hydrogeological conditions [9].
In recent years, surface water and groundwater tracing techniques have been used in a variety of
complex hydrogeologic settings to aid in characterizing groundwater flow systems [10–12]. Tracers
have been used in various combinations of natural tracers, injected tracers, and chemical
perturbations to identify and quantify transport processes in mountain streams impacted by
AMD [13–15]. Fluorescent dyes are often used as an applied tracer [16–18], with the use of
fluorescent dye to trace groundwater dating back to at least 1877 when sodium fluorescein (uranine)
was used to evaluate the connection between the Danube River and the Aach spring [19]. Fluorescent
dyes are commonly chosen as applied tracers for groundwater studies in areas with low clay content,
and recent studies have found dye tracers to work well in both karst and fractured crystalline rock
settings [8]. However, the use of such dyes is problematic in AMD waters, because below a pH of 6
the sorptivity of uranine increases and its fluorescence intensity diminishes [20]; Smart and
Laidlaw [16] demonstrated that the fluorescence of uranine can be reduced by as much as 50% below
a pH of 5. An additional difficulty in dye tracer application in groundwater and mine systems is
accurately quantifying the mixing reservoir. The reservoir represents all waters (mine pools and
inflows of surface or groundwater) that the tracer could mix with between the injection point and
the sampling point and will influence the mass of tracer applied in order to produce appropriate dye
concentrations in collected samples [9]. If the mixing reservoir is overestimated then resulting dye
concentrations may become toxic or exceed the dynamic range of the instruments, whereas an
underestimated reservoir will result in low dye concentrations, possibly below analytical detection.
Therefore, a multiple tracer approach is often recommended for complex hydrologic settings with
166
limited access points and unknown flow-through times, especially when field-work time and
logistical support are limited [9,21,22].
The Rico-Argentine Mine Site near Rico, Colorado (USA) provides an opportunity for the use of
natural and applied tracers to understand the hydrological connectivity of a perturbed system where
AMD is produced. The mine consists of multiple levels of underground workings that are interconnected
by a series of tunnels that pass directly underneath a tributary creek, resulting in potential pathways
for hydrologic connections between the mine and surface waters. Additionally, the mine complex is
connected to a series of long drainage tunnels, which transport AMD from the mine to a discharge
point adjacent to the Dolores River, a relatively pristine headwater ecosystem. At present, the AMD
passively flows through a series of degraded water treatment ponds before entering the river.
The objective of this paper is to use multiple natural and applied tracers to quantitatively and
qualitatively address the hydrologic connections between local inputs from precipitation, surface
waters, and groundwater to interior mine workings, and resultant mine discharges. As recommended
by Wolkersdorfer [9], the paper first highlights the use of synoptic and time series analyses of
naturally occurring isotopic and geochemical tracers to develop a conceptual understanding of the
hydrogeology of the mine system. Secondly, applied tracers (salts and fluorescent dyes) were
introduced, either simultaneously or at separate discrete locations, to provide a comparative analysis
of tracer approaches in an AMD setting. To expand on previous research by Naurath et al. [23], this
study aims to further investigate the effectiveness of uranine as a tracer in acid mine waters by
performing a dual tracer application with uranine and lithium salt.
The results of this study will help determine the feasibility of reducing the volume of water and/or
the load of contaminants that discharge from the Rico-Argentine mine. Reducing the flow of water into
and through the mine workings, reducing the mobilization of contaminants within the mine, and/or
isolating high-concentration contaminant source water for limited smaller-scale treatment, may
create alternative targeted remediation strategies for managing the AMD discharge that are not
currently available.
2. Methods
2.1. Site Description
The historic Rico-Argentine mines are located in the San Juan Mountains of southwestern
Colorado, USA (Figure 1a) and are situated along Silver Creek (Figure 1b), about 2.5 kilometers
(km) northeast of the Town of Rico (Figure 1c). The mines were built to access sulfide replacement
deposits associated with hydrothermal mineralization of faults in the Pennsylvania-age Hermosa
formation. The mines were active from the 1860’s until the 1970’s with silver, zinc, and lead as the
primary mining products. The mines consist of extensive underground mine workings in the ridge
southeast of Silver Creek that were historically accessed by a number of entrances at and above the
elevation of Silver Creek. There are also at least 7 levels (30 m of vertical separation between each)
below the surface elevation of Silver Creek with historical maps indicating several levels of the mine
tunnels pass directly under Silver Creek and connect to mine workings on the northwest of Silver
Creek. The proximity of the mine workings adjacent to, and below, Silver Creek was hypothesized
167
as supporting a possible hydrologic pathway (in either direction) between surface waters in the creek
and the subsurface mine workings. The mine workings on the northwest side of Silver Creek are
connected to the SE crosscut, which connects to the St. Louis Tunnel (Figure 1c). The crosscut and
the St. Louis Tunnel extend approximately 2500 m from the mine workings to the portal adjacent to
the Dolores River. The SE crosscut and the St. Louis Tunnel connect to the Rico-Argentine mine at
the 500 Level (name is a reference to being about 500 ft (152 m) below Silver Creek) and were built
primarily to de-water or drain the upper mine workings using gravity flow. Water with elevated levels
of metals including zinc and cadmium currently discharge from the St. Louis Tunnel and flow
through a series of passive treatment ponds, eventually entering the Dolores River from the pond
system outfall (Figure 1b).
For safety reasons, mine access at the time of study was limited to just a few locations in
the vicinity of Silver Creek. The first accessible tunnel on the southeast side of the creek was
the Argentine Tunnel, located a few hundred feet upslope of the Blaine Tunnel (Figure 1b).
The Argentine Tunnel was accessible for 183 m before becoming unsafe to access, at which point
samples were collected from a stagnant mine pool. The mine pool was believed to have drained to
lower level mine workings, as the water was not flowing out the tunnel to the surface. The second
access point was the Blaine Tunnel, with a portal just 12 m from Silver Creek on the Southeast side.
At the time of the study, water was flowing into the back of the Blaine tunnel, presumable as drainage
from above workings (i.e., those accessed by the Argentine Tunnel), resulting in about 1 m of water
on the tunnel floor behind a cofferdam located 244 m back from the tunnel portal. Water levels
behind the cofferdam were relatively stable (fluctuated < 5 cm over a 6 week period) suggesting that
inflows were proportionate to water movement out of the tunnel towards the lower level workings
that went under Silver Creek. Samples were collected in the Blaine tunnel from both the inflow and
the pooled water behind the cofferdam.
The third access point was a short tunnel on the northwest side of Silver Creek, directly across
Silver Creek from the Blaine portal, which accessed the vertical 517 Shaft approximately 50 m from
the portal along Silver Creek. Historical mine maps indicated that the 517 Shaft extended vertically
to at least the 500 level and that several of the tunnels under Silver Creek intersected the shaft at
multiple levels (Figure 1d). The water level in the 517 Shaft was approximately 146 m below the
elevation of the shaft collar, which corresponds to the elevation of the SE Crosscut and St. Louis
Tunnel. Water samples were collected from the shaft using a bailer and mechanical pulley system
located at the collar. The final access point to sample mine water was the St. Louis Tunnel portal.
The portal had collapsed preventing underground access so samples were collected at first emergence
from the collapse.
168
Figure 1. (a) Map of State of Colorado with location of Town of Rico located in the San
Juan Mountains of southwestern Colorado; (b) Site map of Rico-Argentine mine
complex and Silver Creek. Silver Creek sampling location numbers refer to the distance
downstream (m) from the tracer injection point; (c) Overview map of entire Rico-Argentine
mine complex in relation to the Town of Rico and the Dolores River. Yellow box shows
location of mine complex and represents extent of figure 1b. Location and extent of the
subsurface workings are overlain on top of aerial imagery to demonstrate how mine
workings near Silver Creek are connected to the St. Louis Portal; (d) Vertical view of
mine workings in relation to Silver Creek and access points. The 200 and 300 levels pass
directly under Silver Creek.
169
2.2. Study Design
The study was designed to understand the hydrologic connections between surface water,
groundwater, and mine water associated with the Rico-Argentine mine. The tracer study was
performed in October 2011, when surface water flows were at low-flow conditions, to try and
minimize event based influences (rain and snowmelt) on the prevailing hydrologic conditions. Much
of the southwestern USA, including the San Juan Mountains, can be characterized by bimodal
precipitation patterns, i.e., spring snowmelt and summer monsoon rains [24]. Previous mine drainage
studies in the region have demonstrated that consistent low-flow conditions are most attainable in
late summer (August–October) after monsoon rains have subsided and prior to freezing conditions
during the winter months [15,25,26]. To quantitatively and qualitatively assess individual
components of the hydrologic system, a series of independent environmental and applied tracers
were applied and/or sampled at discrete locations within the study area. All tracers used in the study
are summarized in Table 1.
To broadly understand how inputs to the system (precipitation) contribute to the surface and mine
waters, the naturally occurring stable (į18O/įD) and radioactive (3H) water isotopes were measured
in precipitation and at all other sampling locations. The stable isotopes provide information on source
water contributions whereas the radioactive isotopes provide insight on mean residence time of
waters in the mine system. The second objective of the study was to determine if Silver Creek is
losing water to the mine workings. Silver Creek flow was measured using several different
measurement techniques to evaluate the presence, magnitude, and timing of stream flows on Silver
Creek around the area of the Rico-Argentine mine. Additionally, the mine waters (517 Shaft and St.
Louis Tunnel) and mine tailings seep were analyzed for detection of tracers used in Silver Creek to
look for direct hydrologic connections between the creek and the mine waters. The third objective
was to determine the hydrologic connection between the Blaine tunnel mine water and mine water
in the 517 Shaft. The approach was a slug injection (Table 1) into the Blaine tunnel mine pool and
sampling for the breakthrough of that tracer in the 517 Shaft. The surface water in Silver Creek was
concurrently monitored for Blaine tunnel tracer to check for movement of Blaine mine water towards
surface flows. Finally, the study addressed the hydrologic connection between the mine water in
the 517 Shaft and mine water discharging from the St. Louis Tunnel. The approach involved
simultaneous slug additions of two independent tracers (Lithium salt and uranine dye; Table 1) in
the 517 Shaft, which were analyzed for in samples continuously collected at the St. Louis Tunnel
portal. A dual tracer approach was chosen at this location to have unique data sets to compare tracer
functionality in an AMD setting.
170
Table 1. Summary of natural and applied tracers including the amount of tracer used and the application method. Silver Creek locations are
marked in Figure 1b. Detection limits for stable isotopes reported as twice the precision. (TU, tritium units)
Detection
Limit
Type of Application; Location; Date
Sampling Location; Date
Objective
Oxygen (18O)
Deuterium (2H)
Amount of
tracer
applied
NA
NA
0.056 (‰)
0.296 (‰)
Natural, Local Precipitation, 2011
Natural, Local Precipitation, 2011
All
All
Tritium (3H)
NA
0.3 (TU)
Natural, Local Precipitation, 2011
All
Sulfate (SO4í)
NA
0.0022 (mg/L)
Source water identification
Source water identification
Source water identification;
water apparent age
Hydrologic connectivity;
mine water ļ surface water
Sodium
Bromide (NaBr)
75 kg NaBr
0.0031 (mg/L)
9 kg LiOH
0.0108 (mg/L)
6 kg NaF
0.0057 (mg/L)
Tracer
Lithium
(LiOH)
Sodium
Fluoride
(NaF)
Sodium
Chloride (NaCl)
1.4 Kg
NaCl per
slug
0.0017 (mg/L)
Rhodamine WT
(liquid)
1.2 L
0.006 (ppb)
Uranine
(Sodium
Fluorescein)
13.63 kg
0.002 (ppb)
Naturally occurring in mine impacted
waters
Constant injection (from 189 L
reservoir); Silver Creek (0);
5 October to 7 October 2011
Slug injection (227 L); 517 Shaft;
4 October 2011
Silver Creek (0–1131 m)
Identify changes in streamflow
St. Louis Tunnel portal;
4 October to 16 November 2011
Hydrologic connectivity;
mine ĺ drainage tunnel
Slug injection (30 L); Blaine Tunnel;
5 October 2011
517 Shaft;
5 October to 6 November 2011
Hydrologic connectivity;
mine ļ mine
Silver Creek
(106, 321, 493, 636);
3 October to 7 October 2011
Estimate streamflow above
and below mine workings
Silver Creek (443), Tailings
Seep, 517 Shaft;
4 October to 11 November 2011
Hydrologic connectivity;
Silver Creek ļ mine, tailings
seep
St. Louis Tunnel Portal;
4 October to 11 November 2011
Hydrologic connectivity;
mine ĺ drainage tunnel
Slug injections (8 L);
Silver Creek (0,198,220, 384,
493,530);
3 October to 7 October 2011
Constant injection (from 189 L
reservoir);
Silver Creek (0);
5 October to 7 October 2011
Slug injection (227 L);
517 Shaft; 4 October 2011
Silver Creek (0–1131 m)
171
2.3. Natural Tracers and Synoptic Sampling
Representative water samples were collected from along the study reach on Silver Creek, from
the St. Louis Tunnel portal, and from the Dolores River above and below the mine treatment ponds.
Prior to any disturbance from tracer applications, additional samples were collected from all mine
locations described above and from a seep below the mine tailings pile along Silver Creek to establish
background conditions (Figure 1b). Local snow and rain precipitation were analyzed for isotopes to
characterize sources and timing of inputs to local surface water, groundwater, and mine water. The
snow was a depth integrated sample collected adjacent to the mine complex in April 2011 to represent
maximum snow accumulation prior to the study while rain samples were collected as weekly
composites from five rain gauges located in the San Juan Mountains (within 50 km radius). Surface
water and mine water samples were collected at several different times between April and October
2011 to examine both short-term (i.e., daily) and seasonal variations at different locations. All
samples were analyzed for water isotopes including oxygen-18 (18O), deuterium (2H or D), and
tritium (3H). Water samples were also analyzed for total and dissolved metals and solute chemistry
to identify and distinguish the many contributions to metal load in Silver Creek and the St. Louis tunnel.
Water samples for D and 18O analysis were collected unfiltered in clean 25-mL borosilicate bottles
with no-headspace lids to avoid any evaporation or fractionation. The D and 18O analyses were
performed at the Kiowa Environmental Chemistry Laboratory in Boulder, Colorado, using an Picarro
L1102-i Isotopic Liquid Wavelength- Scanned Cavity Ring Down Spectroscopy (WS-CRDS), a
time-based measurement using near-infrared laser to quantify spectral features of molecules in a gas
ratio of the sample to the Vienna Standard Mean Ocean Water (V-SMOW) standard, as shown
for 18O:
ͳͺ
ͳͺை
ቁ
െ ቀ ைቁ
ͳை ௌ
ͳை ௩௦௪
ߜ ଵ଼ ܱ ൌ
ൈ ͳͲͲͲ‰
ͳͺை
ቁ
ቀ
ͳை ௩௦௪
ቀ
(1)
The 1íı precision of δ18O was ± 0.028 ‰ and of D was ± 0.148 ‰.
Water samples to be analyzed for tritium were collected as grab samples in high-density
polyethylene (HDPE) bottles and stored at 4°C until analysis. The tritium samples were analyzed at
the USGS Tritium Laboratory in Menlo Park, California, by electrolytic enrichment and Liquid
Scintillation Counting. Distilled sample water was reduced electrolytically in electrolysis cells to
10 mL from an initial 200 mL in a cooling bath. The detection limit is reported as twice the precision.
Tritium results reported in tritium units (TU) where 1 TU = 1 tritium atom per 1018 hydrogen atoms.
Water samples collected during synoptic sampling from along the study reach on Silver Creek,
within the mine workings, and at the St. Louis Tunnel portal were analyzed for total and dissolved
metals, select anions, and hardness. Onsite processing included filtration and measurement of pH.
Filtration was completed using 0.45 ȝm capsule filters. Aliquots for cation analysis were acidified to
pH < 2.5 with ultrapure nitric acid. Total recoverable and dissolved cation concentrations were
determined from unfiltered and filtered samples, respectively (dissolved is used herein as an
operational definition that refers to the concentration of the water after filtration; some colloidal
material may pass through the 0.45 ȝm filter). Samples were analyzed at the U.S. EPA Region 8
172
Laboratory operated by the Environmental Sampling Assistance Team (ESAT) contractor for total
and dissolved metals analysis using EPA 200 Series methods, for anions using EPA Method 300,
and hardness was calculated using EPA method 2340B.
During synoptic sampling a subset of each unfiltered sample was also sent for chemical analysis
to the Kiowa Environmental Chemistry Laboratory in Boulder, Colorado, where samples were
immediately filtered through pre-combusted glass fiber filters with a nominal pore size of 0.7 ȝm
and stored in the dark at 4 °C prior to analyses. Samples were analyzed for chloride (Clí), nitrate
(NO3í), sulfate (SO42í), bromide (Brí), and fluoride (Fí), using a Metrohm 761 Compact Ion
Chromatograph. The within run precision (%RSD) was < 1.1% for all solutes. Lithium (Li+) was
analyzed on a Perkin Elmer AAnalyst 200 Atomic Absorption Spectrometer with a within run
precision of 1.06%.
2.4. Applied Tracers
The usefulness of applied tracer test results is highly dependent on proper test design (particularly
determination of injection and sampling locations), the nature of the tracer, the ability to detect the
tracer at low concentrations, and correct interpretation of recovery data [9,18,27,28]. Results from
initial synoptic sampling (in May and June 2011), along with background information on the mine
complex design and local geology, were used to develop a basic understanding of the hydrogeologic
setting allowing for design of the applied tracer study which was performed in October 2011.
Additionally, due to the hypothesis that the entire system under investigation was hydrologically
connected the study was designed to ensure that all applied tracers were distinctly and independently
measureable from each other in the event that mixing of tracers occurred. A previous dye tracer study
was conducted by Davis [29] at a separate mine site (the Atlantic Cable mine) in Rico, Colorado,
at the confluence of Silver Creek and the Dolores River, approximately 2.5 km downstream from the
Rico-Argentine mine complex (Figure 1). The results of the Davis [29] study indicated that water
chemistry (metals analysis) alone was inconclusive with respect to identifying and separating mine
waters from local surface waters. The study was also unsuccessful at identifying surface water and
mine water interactions using Sulpho-Rhodamine B dye tracer (no dye was observed at any of the
sampling points around the mine site) and suggested that Rhodamine WT (RWT) may be a more
effective tracer for future mine water studies in the area. These results further support the use of a
multiple, natural and applied tracer approach to better understand surface water and mine water
interactions in the current study of the Rico-Argentine Mine.
2.4.1. Tracer Selection and Application
In the state of Colorado, the use of surface water or groundwater tracers at permitted mining sites
must be approved by the state mining regulatory agency, which is the Division of Reclamation, Mining,
and Safety (DRMS) [30]. The study was performed under the authority of the U.S. EPA Superfund
Technical Assessment and Response Team and the DRMS, whom both approved the experimental
design and provided technical assistance and oversight throughout the course of the study. All applied
tracers were chosen based on suggestions from literature reviews and based on analytical capabilities
173
and on the ability to safely use at concentrations that are distinguishable above background signals.
Tracer amounts were determined following guidelines presented in Wolkersdorfer [9]. For all applied
tracers the minimum detection limit (Table 1) was considered along with measured background
concentrations and estimated mixing reservoir size (anticipated dilution) to ensure the tracer
concentrations would be clearly identifiable at the sampling point.
There were six tracers used in the study (Table 1) including 4 salts and 2 fluorescent dyes. Tracers
applied to surface waters (Silver Creek) included sodium bromide (NaBr) and sodium chloride
(NaCl), which were applied to quantify streamflow and are described in further detail in Section 2.5.
Additionally, the fluorescent dye Rhodamine WT (RWT; CI Acid Red 388; CAS 37299-86-8;
Keystone Analine Corporation # 703-010-27, Chicago, IL, USA) was applied to Silver Creek
concurrently with the continuous injection of NaBr to provide a dye that could be sampled for
concurrently with the NaBr at several mine sampling points. The RWT was used because it is one of
the most commonly recommended dye tracers for surface water application [18] and was
recommended for future use by a previous study performed at a nearby mine interacting with Silver
Creek [29]. Although RWT has been reported to have some genotoxic properties [16,18,31], the dye
was found to exhibit ecological toxicity at concentrations greater than 1 milligram per liter (mg/L;
1000 parts per billion (ppb)) and human toxicity at concentrations greater than 100 mg/L
(100,000 ppb) [8]. Therefore, the RWT was applied to the stream at low concentrations (<30 ppb)
that were detectable but created minimal visible disturbance and remained well below concentrations
of concern. The continuous injection ran from 5 October to 8 October with two rounds of synoptic
sampling for bromide occurring on 5 and 7 October. Samples were also collected from the mine
Tailings seep and 517 Shaft and analyzed for bromide and RWT to test for presence of Silver Creek
water in the in mine workings or associated tailings pile.
The remaining 3 tracers were applied as slugs directly into mine waters with sodium fluoride
(NaF) applied to the Blaine tunnel mine pool whereas lithium (as LiOH) and uranine (Uranine; C.I.
Acid Yellow 73; CAS 6417-85-2; Keystone Analine Corporation 801-073042 Chicago, IL, USA)
were applied concurrently to the mine pool at the bottom of the 517 shaft. The ability to easily sample
and accurately quantify discharge at the St. Louis Tunnel portal motivated the application of multiple
tracers at this location. By calculating the mass recovery of independent tracers the effectiveness of
each tracer could be addressed. The lithium tracer was chosen because of low background signal in
the mine water and because several of the other most common salt tracers (NaCl, NaBr, and NaF)
had been used in the study. Uranine was chosen as the second tracer because it has long been used
as a subsurface tracer due to low sorptive properties [16,32–35] and for being safe in terms of human
or ecological toxicity [8,18,31]. Uranine also readily undergoes photo-degradation [16], but has
shown long-term stability when not affected by sunlight [30], making it most suited for subsurface
tracing. Uranine has only negative functional groups, and sorbs least onto negatively-charged media
and most onto positively charged surfaces [35]. Mine waters often contain metal cations, especially
iron (Fe), which can precipitate as hydroxides at higher pH values [36]. Uranine may therefore
undergo adsorption onto iron hydroxides and be removed from the sample through filtration [36].
Conversely, if the collected mine samples are not filtered then fluorescence intensity may be
174
overestimated due to scattering effects caused by increased sample turbidity. Results from the dual
tracer application will therefore enable this potential complication to be addressed.
On 4 October at 12:00, a 189-L slug containing 9 kg of LiOH and 13.63 kg of uranine was injected
into the 517 Shaft. The size of the mine pool (mixing reservoir) at the bottom of the 517 Shaft was
unknown due to inaccessibility to the flooded lower levels of the mine. Therefore, the necessary mass
of tracer was estimated using the flow rate at the St. Louis Tunnel portal (assumed mine pool
discharge point), the distance between injection and sampling location, and by assuming a reasonable
estimated mine water flow velocity. Additionally, background concentrations of lithium were low
(0.025 mg/L) and background fluorescence in the wavelengths used to detect uranine were below
detection (<0.002 ppb) at the St. Louis Tunnel portal confirming minimal background interference
with tracer detection. The estimated velocity (and mass) was verified based on the authors’
experience with a similar tracer test at another mine site drainage tunnel with similar distance
between injection and sampling points [37]. Additionally, the mine outflow at the St. Louis Tunnel
portal flowed directly into a series of passive settling ponds and not into natural waters, which
eliminated concerns for high concentrations of tracers (from over estimating tracer mass) rapidly
entering local surface waters in the event of incomplete mixing of the tracer slug. The slug was mixed
using acidic water (pH 2.7) from the Blaine Tunnel to simulate the water found at the bottom of the
517 Shaft (pH 2.5). The slug was mixed in two new, clean 115-L plastic tubs located adjacent to the
top of the shaft and then siphoned from the tubs into the shaft using 137 m of new, clean garden hose.
To avoid tracer contamination all materials used to transport and mix the tracers (including
applicators clothing) were left at the injection site. The slug injection took approximately 45 minutes
to complete and was then chased by approximately 189,000 L of water from Silver Creek delivered
by a high capacity water pump over a 30-min period.
On 5 October at 14:15, a slug containing 6 kg of NaF was added to the Blaine Tunnel mine pool
behind the cofferdam. The tracer mass was estimated based on the size of the mixing reservoir at
injection and sampling points (Blaine tunnel mine pool and 517 Shaft mine pool), and information
on background levels of fluoride at the injection point. The tracer mass was sufficient to increase Fí
concentrations to approximately three times background (from 50.2 to 143.1 mg/L) in the Blaine
Tunnel. The tracer recovery point (517 Shaft) had Fí background of only 1.95 mg/L. The slug was
made by dissolving NaF powder into 30 L of Blaine Tunnel water using two 20-L buckets and a
stirring rod made the slug. The NaF solution was slowly poured in the mine pool about 8 m behind
the cofferdam. The mine pool had an estimated minimum volume of 42,500 L, but the inflow/outflow
rates were not quantifiable. After the slug injection, a 7.6 cm diaphragm water pump was used to move
water from the mine pool over a visible tunnel collapse (pile of rock and debris) and toward a drainage
stope that extended to the lower mine levels that intersect with the 517 Shaft. The pump was able to
move between 22,700 and 36,300 liters of water past the initial visible collapse in the first 2 h after
injection. However, during that period no reduction in water level was observed in the mine pool and
pumping was then terminated due to freezing of equipment and unsteady pump rates, so it was not
possible to calculate the rate at which the entire mine pool (containing the fluoride slug) was moved
past the initial blockage.
175
The exact pathway of the water (and tracer) leaving the Blaine Tunnel pool is unknown, but based
on existing maps it was expected that water flows down a complex series of interconnected inclines,
raises, winzes, stopes, tunnels, and shafts before reaching the 500 level via the 517 Shaft. Given the
myriad of potential pathways, all in varying states of structural integrity, it was difficult to determine
the actual distance the water and tracers would travel. Using all available information is was estimated
that the shortest possible pathway from the Blaine Tunnel mine pool to the bottom of the 517 Shaft
was 200 m.
2.4.2. Tracer Sampling and Analysis
Automated samplers were placed at 3 locations; SC-493, the mine Tailings seep, and the St. Louis
Tunnel portal (Figure 1). The samplers in Silver Creek and at the mine Tailings seep collected
samples at 1-h intervals for the duration of the constant injection of tracers in Silver Creek and then
at 4-h intervals for 3 weeks following tracer application. At the St. Louis Tunnel portal water samples
were collected at 1-h intervals for the first 40 h, starting 2 h before the 517 Shaft tracer injections. A
Cyclops 7 field fluorometer from Turner Designs, Inc. (Sunnyvale, CA, USA), confirmed that the
peak of the uranine slug occurred within the first 36 h, so sample collection was reduced to 4-h
intervals on 7 October and continued for 6 weeks.
At the 517 Shaft, water samples were collected at a depth of 7.6 m below the water surface, which
was 137 m below the collar of the shaft. Samples were collected manually using a stainless steel
bailer lowered to the same depth for each sample using a calibrated cable reel and motorized pulley
system. Samples were collected at 1-h intervals for the first 36 h after tracer injections were made in
the Blaine Tunnel and then daily for 2 weeks. The last sample was collected in the 517 Shaft 690 h
after tracer was injected in the Blaine Tunnel.
To provide real time confirmation of tracer emergence, the Cyclops-7 field fluorometer
from Turner Designs Inc. was used to make in-situ field measurements for the presence of the
two fluorescent dyes, RWT and uranine. The field fluorometer was calibrated for both dyes using
a 4-point calibration with standards of 0, 1, 100, and 400 ppb solutions. The standards were prepared
in the laboratory by diluting the purchased dye concentrates using water collected directly from the
mine site at locations where the dyes were to be applied or measured. The RWT was mixed using
water from Silver Creek (pH 8.2) and the uranine was mixed using water from the St. Louis Tunnel
portal (pH 7.4). The manufacturer stated that the instrument’s dynamic range for detection of uranine
dye was 0 to 500 ppb so calibration from 0 to 400 ppb was considered sufficient for the intended use
of the field fluorometer. The field fluorometer was used to qualitatively provide a rapid assessment
of samples for presence/absence of the dye of interest to constrain the number of field samples that
would be sent to the laboratory for analysis. At the Kiowa Laboratory, a Fluoromax 2 (F2)
spectrophotometer was used to analyze water samples for the concentrations of the two fluorescent
dyes. The presence of each dye was analyzed using a single excitation and a single emission value
and a record of the spectrum for a 100 ppb standard for both dyes was created to determine the
appropriate excitation/emission values to use for analyzing the dye tracers. The RWT dye was run
with 550/580 nm excitation/emission and the uranine was run with 492/512 nm excitation/emission.
The excitation and emission values for both dyes were within the expected variability [19] of values
176
previously reported for fluorescence analysis of these dyes [8,18,23]. For RWT, a 5-point calibration
curve was developed using standards between 1 and 400 ppb concentration. For uranine, a 12-point
calibration curve was developed with concentrations ranging from 1 to 1000 ppb. Dye concentrations
were plotted against emissions values to generate calibration curves. Interestingly, for the uranine,
F2 emissions peaked with the 500 ppb standard and then became inversely related to concentrations
over 500 ppb. As reported in Käss [19] the intensity of the fluorescence likely decreased as a
consequence of its individual light absorption and due to retrograde dissociation. To compensate for
the decrease in emissions for high concentration samples, the standard power calibration curve (dye
concentration in ppb (C) = 10í13 × E2.1558; where E = measured emission (nm); R2 = 0.99, N = 6) was
used to calculate dye concentrations in samples with concentrations < 500 ppb and a second linear
equation (dye concentration in ppb (C) = í10í4 × E + 2948; where E = measured emission (nm);
R2 = 0.9, N = 6) was used to estimate dye concentrations in samples with concentrations > 500 ppb.
The inflection point of 500 ppb was chosen from visual interpretation of emission/concentration plot.
Additionally, when the standard curve was applied to all F2 emission results there was a false double
peak in calculated uranine concentrations in the St. Louis Tunnel samples. Fortunately, the sample
with the lowest uranine concentration (between the 2 false peaks) had the highest measured
concentration of lithium (the second concurrently applied tracer) confirming a simultaneous arrival
of the peak concentration of both tracers. The two false peaks were also at concentrations just below
500 ppb, further confirming that at concentrations over 500 ppb the F2 was unable to accurately
quantify the amount of fluorescing compounds in the sample.
The fluorescence intensity of dyes is known to have a variable response to changes in
pH [16,18,20,21,23,29,34,35]. To test for the effects of acidification on fluorescence, high
concentration (400 ppb) standards of both uranine and RWT were acidified to a pH of 2.5 for 24 h
and then filtered. The result was a 77% reduction in uranine emissions and an 11% reduction in RWT
emissions. The reduction in uranine fluorescence agrees with previous research documenting
considerable decrease in fluorescence intensity (up to 90% at pH 3) due to reversible ion exchange
reactions caused by the acidic conditions [16,18,21]. However, as reported by Käss [18], the pH
effects were reversible; when the standards were re-neutralized to the slightly alkaline conditions of
the original standards (pH 8.2 and 7.4 for RWT and uranine, respectively) there was full recovery of
the dye fluorescence emissions for both uranine and RWT. Although the uranine dye reaches
maximum fluorescence under alkaline conditions (pH 9) [16,18,23], the dye standards (for calibration
and pH response measurements) were made with the actual mine water at the sampling location, so
the above mentioned results should be representative of the range of pH the dye will be exposed to
in the study.
2.5. Measuring Streamflow and Mine Discharge
There are many challenges to measuring streamflow in mountains including bed surface
roughness, difficult to constrain channel dimensions, and steep stream gradients with varying degrees
of hyporheic zone storage [38,39]. Furthermore, Marchand et al. [40] suggest that conventional
current meter measurements of discharge do not properly account for hydraulic conditions commonly
found in high-gradient, shallow streams of mountain catchments. Therefore a number of streamflow
177
measurement techniques were used to determine the magnitude and timing of flows in Silver Creek,
and to evaluate potential flow losses from Silver Creek to mine workings underlying and on either
side of the creek. The measurements would also identify any major inflows to Silver Creek in the
vicinity of the mine. Quantification of streamflow was performed during both high flow (spring
snowmelt) and low flow (fall base flow) to understand stream dynamics across the full range of the
hydrograph. For this paper, only the low flow discharge will be discussed in detail, as applied tracer
studies were performed during base flow.
Silver Creek flow was measured at two locations: an existing rectangular concrete weir structure
with attached gauge house located upstream of the Blackhawk fault and mine workings (SC-68;
Figure 1b), and at the downstream side of a road culvert just below the mine workings (SC-493;
Figure 1b). Stage-discharge relationships (rating curves) were developed by correlating data from
pressure transducers (stage) with discharge measurements made using an electromagnetic flow
meter. Pressure transducers recorded stage every 15 min from June to August and October to
November and discharge measurements were collected periodically to capture the full range of the
stream hydrograph. Pressure transducers were removed in August to prevent vandalism and were
re-installed in the same locations prior to the start of tracer tests in October.
Slug additions of sodium chloride (NaCl) were made to develop point estimates of streamflow at
select locations above and below the mine workings along Silver Creek. Transport of the slug to a
downstream observation point was monitored by a portable sonde that recorded specific conductance
at 4-s intervals. Several grab samples were collected from the stream during slug tests and analyzed
for Clí concentrations, which were matched with corresponding observations of specific conductance
to develop a chloride-specific conductance relationship. The relationship was then used to convert
the observed increases in specific conductance to chloride concentrations with the resultant profiles
integrated to provide estimates of streamflow (Q, L/s) (1) following Kilpatrick and Cobb [41]:
Q=
C
V
(2)
where C is the mass of Cl added (in kg) (C = 0.59 × mass of NaCl) and V is the time integral of the
Cl concentration (above background) at the monitoring site. The units of V are (mg/L) per second.
The third approach for quantifying streamflow in Silver Creek consisted of a continuous sodium
bromide (NaBr) injection to provide streamflow estimates via the tracer dilution method [41] and
document any potential areas of flow loss. Once the bromide tracer concentration reached a plateau
synoptic sampling provided a spatial snapshot of bromide concentration and was used to determine
a flow regime. In gaining streams, dilution of bromide with respect to distance is indicative of
increased streamflow. Losing or constant flow streams, in contrast, will exhibit steady bromide
concentrations with distance, as water leaving the stream does not affect the in-stream concentration.
Discharge was also measured at the St. Louis tunnel portal. Flow was directed through a 9-inch
(22.86 cm) Parshall flume and stage heights were measured at 3 mm resolution at 15-min intervals
with an ultrasonic automated water level detector that was installed on 12 May 2011. Flow estimates
(L/s) were then determined using the flume’s standardized stage-discharge relationship. Due to
limited access to the interior portions of the mine, flow could not be directly measured along tracer
flow paths in the mine workings.
178
3. Results and Discussion
3.1. Silver Creek: Streamflow and Hydrologic Connectivity to Mine
The multiple techniques used to quantify Silver Creek streamflow across the study site found that
flow varied between 19.8 and 56.6 L/s during the October tracer study. Unfortunately, the high
gradient mountain stream combined with unsteady streamflow (caused by two small precipitation
events) produced unavoidable challenges in accurately quantifying streamflow. During the study
period the discharge records from stage-discharge relationships at SC-68 (above the mine) and
SC-493 (below the mine) produced “cross-over” where the largest instantaneous flow was at the
upstream location during some periods of the study and at the downstream location at other times.
Therefore, the two locations of measured discharge alone were inconclusive on determining if Silver
Creek was only gaining or only losing water as it moved past the mine complex, and suggest that
both inflows and outflows may be occurring within the study reach. The slug addition techniques
also produced varying results ranging from 7% to 29% loss across the study reach during the
sampling event with loss rates being smaller after precipitation events. The precipitation events likely
masked some of the streamflow losses by adding small but influential surface water inflows along
the study reach. The constantly injected RWT and Brí tracers were analyzed by two separate synoptic
sampling events, both following precipitation events, and the Brí profiles on both sweeps suggested
a gaining stream at the upstream end of the study reach (above the mine workings) and the potential
for flow loss downstream of the mine workings between SC-542 and SC-759. To summarize, Silver
Creek likely has some net flow loss along the study reach but high gradient stream channel and
the precipitation events made it difficult to accurately quantify streamflow variations during the
study period.
Concurrent with the continuous injection of Brí and RWT, the 517 Shaft and the mine Tailing
seep were sampled for the presence of the injected tracers to identify if water leaving Silver Creek
(as identified by discharge calculations) was interacting with the mine system or Tailings pile on the
northern side of the creek. The Br and RWT tracers injected into Silver Creek were not detected at
elevated concentrations in the mine workings or discharges from the Tailings seep and St. Louis
Tunnel portal. At the seep Brí concentrations remained below detection (<0.003 mg/L) for the
duration of the sampling. There was no detectable increase in RWT concentrations, which remained
below detection (0.006 ppb) throughout the sampling period. In the 517 Shaft, background
concentrations of Br (0.046, ı = 0.037 mg/L) were detected both before and during the tracer study,
but concentrations remained at background levels with no evidence of an increase or breakthrough
slug emerging at the 517 Shaft. There was no detection of RWT before or during the tracer study.
The results fail to confirm a hydrologic connection from Silver Creek to the Tailings seep or the
mine workings directly connected with the 517 Shaft. However, results prevent positive confirmation
or rejection of minor flows from Silver Creek through the mine to the 517 Shaft during low-flow
conditions. One logistical constraint in the tracer design was that the RWT concentrations in Silver
Creek were maintained at very low concentrations (range 16 to 28 ppb) to avoid any potential
negative (toxicological or visual) impacts. Additionally, the in-stream Brí concentrations were kept
at a measureable but conservative concentration (4.73, ı = 0.7 mg/L) to ensure that there were no
179
potentially toxic effects to aquatic life. The low concentrations of stream tracers, combined with the
large volumes of mine waters, made it unlikely for the tracers to be detected, especially if additional
dilutions from ground water were considered. A second challenge to identifying stream tracers in the
mine, was that the length of the sampling window. Tracer recovery sampling in the mine was limited
by personnel availability and seasonal access, which only continued for approximately 30 days after
the tracer injections. Therefore tracers may not have arrived at the sampling locations during the
sampling window. Third, it is possible that some Silver Creek water (containing the applied tracers)
reached the subsurface mine workings but underwent dilution from mixing with other groundwater
or mine water prior to arrival at the sampling point, causing dilution to background concentrations.
The use of low concentrations of tracers in Silver Creek to minimize adverse environmental impacts
was likely the ultimate limiting factor preventing positive confirmation or rejection of hydrologic
connection between Silver Creek and the sampled mine water locations.
3.2. Natural Tracers
3.2.1. Water Quality
Water quality varied widely across all sampling sites, with the highest concentration of
contaminants generally associated with the mine pools and lowest concentrations in Silver Creek and
the St. Louis Tunnel portal. The pH of samples ranged from a high of 8.7 (±0.19 (1SD), N = 17) in
Silver Creek to a low of 2.38 (± 0.06 (1SD), N = 5) in the Blaine Tunnel mine pool (Figure 2).
Dissolved aluminum and zinc concentrations were highest in the mine workings to the southeast of Silver
Creek and decreased prior to exiting the St. Louis Tunnel portal. Aluminum ranged from a high
concentration of 654 mg/L in the Argentine Tunnel to less than 0.1 mg/L at the St. Louis Tunnel
portal (Figure 3). Similarly, zinc concentrations were highest in the Argentine Tunnel pool at
24,600 mg/L and decreased to 3.8 mg/L at the St. Louis Tunnel portal. Solute loading analysis was
not performed as flow rates from the Argentine Tunnel, Blaine Tunnel, and 517 Shaft mine sites
were not measured during the study.
During the tracer experiment water quality along Silver Creek was synoptically sampled at the
same locations as were used for the stream tracer study (Figure 1b). Sulfate (SO42í), generally
considered a conservative tracer in AMD settings [27], was measured at all sampling points and showed
a measureable increase from under 10 mg/L above SC-198 to over 50 mg/L at SC-1131 the furthest
downstream sampling point (Figure 4). Dissolved zinc concentrations were measured at nine of
the sampling points and increased from under 0.005 mg/L upstream of the mines to 570 mg/L at
SC-1131. Both sulfate and zinc concentrations had the largest increases in the 200 m upstream of
the mine entrances (at approximately 400 m) and then remained relatively constant until increasing
again beyond 800 m.
180
Figure 2. Box plot of pH in the mine workings along with that of the St. Louis Tunnel
portal and tailings seep discharges and of surface waters in Silver Creek.
Figure 3. Dissolved zinc and total dissolved aluminum concentrations in mine workings.
The mine workings are listed from left to right in the direction of hydrologic flow towards
the discharge point at the St. Louis Tunnel portal.
181
Figure 4. Sulfate and dissolved zinc concentrations in Silver Creek from synoptic
sampling during base flow conditions on 7 October 2011. Stream locations are in m
downstream from the location of the stream tracer injections with the stream crossing the
mine workings at approximately 400 m. There are known inflows to Silver Creek from
seepage below the mine tailings entering between SC-895 and SC-1131 in Figure 1.
Silver Creek chemistry suggests that mining activities impact Silver Creek and the sources of
AMD products (sulfate and zinc) are occurring in similar locations. The sulfate and mining-associated
metal concentrations begin to increase in Silver Creek in the vicinity of the Rico-Argentine mines
and indicate that even though there is a potential net loss of flow in the reach, there may be small
inputs of mine water or other inflows that influence the stream chemistry. The first considerable
increases in stream solutes occur between 200 and 300 m, which is upstream of the primary mine
entrances (Figure 1b), but within the area of historical mining operations. Due to mining activities,
there are considerable amounts of mine debris (mill tailings and mine waste rock) scattered along the
banks of the stream channel starting at approximately 100 m (Figure 1b). Given the variable
discharge (gains and losses) measured along the study reach it is likely that the Silver Creek waters
are undergoing hyporheic zone mixing in areas where solute laden mine wastes are present. As a
result, contaminated sediments in the hyporheic zone may be representing a long-lasting supply of
contaminants to hyporheic pore water [42]. A second source of solutes is slow percolation of highly
concentrated AMD water in the Blaine Tunnel mine pool into the subsurface and eventual movement
towards the stream channel, which would likely occur at or just downstream of the Blaine Tunnel
portal between 350 and 400 m. The relatively consistent solute concentrations between 400 and
700 m is likely because previous mine cleanup work at this site included capping the mine tailing
pile (Figure 1) and reinforcing the bank between the tailings pile and Silver Creek to reduce erosion
of the tailings pile into Silver Creek. As a result, the amount of surface water and hyporheic zone
mixing may be decreasing along this segment creating less opportunity for mining-associated sulfate
and zinc to enter Silver Creek. The increase in AMD contaminants to Silver Creek at locations below
182
the mine entrances (below SC-826) are likely from diffuse inflows from the tailings pile seep which
are visually present between SC-895 and SC-1131 in Figure 1.
The results in Figures 3 and 4 therefore suggest that, when excluding the contributions from the
large mine tailings pile, the sulfide deposits sourcing the AMD may be most prevalent in the mine
workings and shallow subsurface on the southeastern side of Silver Creek between 200 and 400 m.
As suggested by Nordstrom [6] this shallow acidic groundwater can then be mobilized by infiltration
of meteoric waters, resulting in movement towards surface waters (or further into the mine complex).
To elaborate, the high metals concentrations found in the mine waters to the southeast of Silver Creek
may be partially explained by the mines artificial lowering of the local water table. The mine drainage
tunnels were built to help dewater the mine workings but as a consequence the mine workings also
exposed large amounts of sulfide minerals to air, which increased oxidation reactions. Therefore, the
combined exposure to air and water (from seasonal meteoric infiltration and shallow groundwater
flow) in the mine workings adjacent to, and above, Silver Creek has created an ideal situation for
AMD production that is impacting the water quality in Silver Creek and at the St. Louis Tunnel.
3.2.2. Isotopes
Stable isotope results from all samples including precipitation, surface water, and mine water
were plotted as the įD–į18O relationship (Figure 5A). The į18O value for snow of approximately
í20‰ and the mean rain value of í5‰ are characteristic values for the Colorado Rocky Mountains
for winter and summer precipitation [43]. Empirical results have shown that įD/į18O values in
precipitation co-vary and are generally described by the relationship [44]:
ߜ ܦൌ ͺߜ ଵ଼
(3)
which is defined as the Global Meteoric Water Line (GMWL). The Local Meteoric Water line
(LMWL) has a similar slope and y-intercept:
(4)
ߜ ܦൌ Ǥͺߜ ଵ଼
183
Figure 5. (A) Plot of į18O vs. įD for precipitation, surface waters, and mine waters. The
St. Louis Tunnel, 517 shaft, Blaine tunnel, Tailings seep, and Silver Creek are the
arithmetic mean from all samples collected. The Local Meteoric Water Line (LMWL) is
plotted in black; (B) Seasonal variations of į18O vs. įD for surface waters and mine
waters in 2011. The LMWL is plotted with a second water line (WL1; red dash),
representing samples from the Tailings seep and the Blaine Tunnel and a third water line
(WL2; blue dot) representing samples from Silver Creek and the St. Louis Tunnel. The
symbol shapes correspond to similar sites in (A) and (B), while colors represent similar
dates for different sites in (B).
-20
Snow
Rain
St. Louis Tunnel
ɷD ‰
517 Sha
-90
Tailings Seep
Silver Creek June
Silver Creek Oct
Blaine Tunnel
LMWL
-160
-25
-20
-15
-10
-5
0
ɷ 18O ‰
A
ɷD ‰
-102
-107
-112
-15.5
B
Silver Creek (June)
Silver Creek (Oct)
St. Louis Tunnel (May)
St. Louis Tunnel (June)
St. Louis Tunnel (Aug)
St. Louis Tunnel (Oct)
Blaine Tunnel (Aug)
Blaine Tunnel (Oct)
Tailings Seep (June)
Tailings seep (Aug)
Tailings seep (Oct)
517 Sha (Aug)
517 Sha (Oct)
LMWL
Linear (WL-1)
Linear (WL-2)
-15.2
-14.9
-14.6
-14.3
-14.0
ɷ 18O ‰
The similar values in slope between the LMWL (7.8) and the GWML (8.0) suggest an absence of
complex kinetic fractionation processes affecting the įD–į18O relationship of precipitation (inputs)
in the local hydrologic system [45]. All surface water and mine water samples fell on a mixing line
between the snow and rain inputs, suggesting that they were a mixture of the two precipitation types
with į18O values between í14‰ and í15‰.
To better observe the variations in surface waters and subsurface waters (relative to the LMWL)
the įD–į18O relationship was examined at a finer resolution (Figure 5B). The isotopic concentration
of Silver Creek had the greatest seasonal variation with waters more depleted in įD and į18O in June
184
as a result of snow melt and more enriched in įD and į18O in October from summer rains and/or
groundwater contributions. All of the samples collected in or near the mine (Tailings seep, Blaine
Tunnel, 517 Shaft) had an isotopic composition similar to Silver Creek in October, indicating that
stream water during baseflow was of similar origin to the water moving through both the upper mine
and the seep. Additionally, there was little variation in isotopic signature of the Tailings seep, which
did not follow variations observed in Silver Creek, indicating that water exiting the Tailings seep
was coming from a well-mixed source distinct from Silver Creek waters. However, the samples
from the Tailings seep and the Blaine Tunnel all plot below the LMWL with a decreased slope
and y-intercept:
ߜ ܦൌ ͵Ǥͳߜ ଵ଼
(5)
ߜ ܦൌ ǤͶߜ ଵ଼
(6)
This linear relationship is represented by WL-1 in figure 5b and the decreased slope suggests that
an evaporation trend occurred at these locations since recharge [45,46]. Conversely, samples from
the St. Louis Tunnel and Silver Creek express the following įD–į18O relationship:
The linear relationship is represented by WL-2 in figure 5B, and the similar slope between WL-2
(7.4) and LMWL (7.8) again suggests an absence of fractionation processes affecting those water
sources since meteoric recharge. The two samples from the 517 Shaft plot close to, and slightly
below, the LWML indicative of a mixture of predominantly unfractionated waters (since recharge)
with a lesser amount of fractionated waters (i.e., from the Blaine Tunnel). These results therefore
suggest that the waters with evidence of evaporation (Tailings seep and Blaine Tunnel) are distinct
from local precipitation (LMWL), surface water (Silver Creek), and the dominant source of water
exiting the mine system (St. Louis Tunnel).
Figure 5B also indicates that the water discharging from the St. Louis Tunnel had similar isotopic
concentration in May, June, and October, indicating the water was also predominantly from a wellmixed groundwater source with limited seasonal variation. The St. Louis Tunnel samples were also
consistently more depleted than the samples from the 517 Shaft and Blaine Tunnel, indicating that
water flowing downgradient from the 517 Shaft was mixing with water that has a distinctly different
isotopic signature (more depleted) before emerging at the portal.
The stable isotope analysis provides initial insight into the hydrologic connectivity of the system,
but due to the limited range (magnitude) of isotopic variability across all sites, is insufficient as a
stand-alone tool for identifying the sources and relative contributions of water contributing to the AMD.
Tritium values from 21 rain and snow samples collected in the San Juan Mountain region between
January 2010 and October of 2011 had a mean value of 6.2 TU with the 25th and 75th percentiles of
samples being 4.6 and 6.5 TU respectively (Figure 6). The narrow range of tritium in incoming
precipitation provided a strong estimate of the current meteoric inputs of tritium to the local
hydrologic system under investigation. For comparison the annual mean tritium concentration in
precipitation in Colorado was only 7.5 TU in 1999 [37], suggesting that tritium levels in meteoric
waters have only decreased by about 1 TU in the decade preceding the current study. The tritium
data provide insight as to the mean residence time of “old” versus “new” groundwater in the study.
Tritium (t1/2 = 12.43 years) is naturally present only in minute quantities but was produced in large
185
quantities during thermonuclear weapons testing from 1952 to 1963 [47]. As a consequence, without
consideration for complex mixing scenarios, water with a tritium value <1 TU may be considered
“old” water recharged prior to 1952 while higher values would represent “new” water being wholly
or partially recharged since 1952. Additionally, if a tritium concentration is greater than the current
meteoric recharge it can be considered to have some portion of water that precipitated with a “bomb
spike” signal in the decades following weapons testing. No samples had <1 TU while all mine water
and surface water tritium concentrations were near or within the range of recent precipitation
suggesting that those waters were derived predominantly from recent (within a few years) meteoric
recharge. The Argentine Tunnel sample, along with the Silver Creek and the Dolores River samples,
had slightly elevated tritium concentrations relative to the current (2010–2011) meteoric inputs,
suggesting that the water in those locations may have had slightly longer mean residence times than
the waters emerging at the Tailings seep and moving through the mine to the St. Louis Tunnel. The
tritium results provide only a qualitative estimate of water residence times, but suggest that the mean
residence times of the mine waters are not distinctly different from residence times of the local and
regional surface waters.
Figure 6. Tritium (3H) values from recent (2010–2011) regional precipitation (blue),
surface waters (green and pink), and mine waters (orange). The blue precipitation bar
represents the mean value of all samples while the red error bar represents the 25th and
75th percentiles.
3.3. Applied Tracers
3.3.1. Blaine Tunnel to 517 Shaft
The first hydrologic connection to be investigated with applied tracers was the movement of water
from the Blaine Tunnel to the 517 Shaft and the interior mine workings. There was no practical way
186
to quantify the volume of water entering the 517 Shaft from the Blaine Tunnel area, thus
quantification of tracer mass recovery was not possible. Additionally, it was estimated that only
between 50% and 85% of the mine pool water (containing the fluoride tracer slug) was successfully
pumped over the first debris blockage towards the drainage stope exiting the Blaine Tunnel making
it unclear what percentage of the initial tracer was moved out of the Blaine Tunnel mine pool via
pumping. However, positive detection of the tracer provided a qualitative assessment of hydrologic
connectivity between the Blaine Tunnel and the 517 Shaft area of the mine. Figure 7 displays a time
series plot of fluoride and chloride concentrations in the 517 Shaft starting on 5 October at 14:00
when the fluoride slug was added to the Blaine Tunnel behind the cofferdam. Chloride concentrations
remained relatively steady (2.11, ı = 0.08 mg/L) over the first 48 h, indicating that the changes in
fluoride concentrations were occurring independently from any changes in the mine water chemistry
caused by previous hydrological alterations such as the water chase applied to the 517 Shaft during
the uranine/lithium tracer application. Interestingly, the chloride concentrations did decrease to
a mean of 1.78 (ı = 0.05) mg/L for approximately 96 h (approximately hour 48 to 144 in Figure 7),
which corresponds to the period when the highest fluoride concentrations were measured in the
517 Shaft. Given that background chloride concentration was lower in the Blaine Tunnel (1.54 mg/L)
than in the 517 Shaft (2.14 mg/L) it is reasonable to suggest that the Blaine Tunnel water (which was
partially moved out of the Blaine Tunnel by pumping) arrived at the 517 Shaft at an increased rate
relative to background flow rates creating increases and decreases in fluoride and chloride
concentrations respectively.
Figure 7. Fluoride (Fí) concentrations measured in the 517 Shaft. Background chloride
(Clí) concentrations are plotted over the same time period. Time (x-axis) plotted as log scale.
The initial fluoride concentration in the 517 Shaft (1.95 mg/L) indicated that there was a
measureable background presence of fluoride in the system. However, the fluoride concentrations
began to show a steady increase above background concentrations beginning 10 h after injection, and
more than doubled to reach a peak concentration of 4.94 (mg/L) 68 h after the slug injection.
187
The fluoride concentration then began to fall sharply approximately 8 days after the tracer injection.
Given an approximate minimum distance from the Blaine Tunnel cofferdam to the base of the
517 Shaft of 200 m, the velocity of the advection front (maximal velocity) was 0.34 m/min while the
average velocity of the peak fluoride concentration was just 0.05 m/min. When compared to results
reported by Wolkersdorfer [9], the average velocity is considerably slower than the range of 0.3 to
1.6 m/min reported as 95% confidence interval of 42 tracer tests. From a qualitative standpoint the
results suggest that the hydrologic flow paths between the Blaine Tunnel and 517 Shaft may have
major obstructions and/or the water is moving slower than has been most commonly observed in
previously studied mine water environments. Additionally, there is diffusion in the system because
there was not a distinct breakthrough curve. One anticipated possibility is one or several obstructions
at points along the tunnel/stopes that causes the water to interact with rock and sediment. The
collapsed portion of the adit over which water was pumped immediately after tracer injection is one
likely obstruction. As mentioned in the methods, the tracer recovery at the 517 Shaft was achieved
by sampling from a mine pool of unknown volume and flow-through rates. Additionally, the precise
volume and rate of water moved out of the Blaine Tunnel mine pool (injection point) was not
quantified due to pumping inefficiencies and unsafe access to drainage points. The flow path of the
tracer between the Blaine Tunnel and the 517 Shaft was also likely diverse (multiple interconnected
tunnels, stopes, and inclines present) making quantification of tracer travel distance difficult.
Therefore the tracer breakthrough rates should be considered in a qualitative manner and no
quantitative transport model (e.g., [9,18,27]) was used to evaluate solute transport and dispersion.
During the Blaine Tunnel tracer injection, water samples were also collected from SC-493
(about 120 m downstream from the Blaine Tunnel adit, Figure 1b) to determine if a detectable
amount of water was flowing from the Blaine Tunnel, through the subsurface, to Silver Creek.
The fluoride concentrations at SC-443 after the Blaine tunnel injection remained relatively steady
(0.104, ı = 0.018 mg/L) with intermittent but not sustained increases, which neither confirms nor
rejects the hypothesis that water contained in the Blaine Tunnel may be directly entering Silver Creek
in the vicinity of the portal.
The original tracer design included pairing a third fluorescent dye (PTSA; 1,3,6,8-Pyrenetetrasulfonic
Acid Tretrasodium salt; CAS # 59592-10-0) with the fluoride salt tracer applied to the Blaine tunnel.
This dye tracer has a fluorescence emission peak that is distinct from both uranine and RWT so it
could be used in a multiple tracer situation when tracer mixing is possible. Additionally, the Cyclops
7 field fluorometer from Turner Designs has a submersible sensor specifically designed for detection
of PTSA. However, knowing that there were potentially multiple blockages (rock collapses in-filled
with sediments or AMD derived precipitates) the dye tracer was not used and thus still available for
additional future studies. This is important because if a tracer is used and full recovery cannot be
confirmed then that tracer would not be desirable for future use in the same location for reasons of
cross contamination.
3.3.2. 517 Shaft to St. Louis Tunnel Portal
The arrival time for both uranine and lithium at the St. Louis Tunnel portal was at 03:00 on
5 October, 15 h after injection (Figure 8). After arrival, the concentration of both tracers increased
188
rapidly with uranine reaching a peak concentration of 2900 ppb at 08:00, and the lithium tracer
reaching a peak concentration of 0.258 mg/L at 09:00 on 5 October, representing 20 and 21 h to peak
concentration, respectively. There was no clear increase in discharge as a result of the water chase
(189,000 L in 30 min § 105 L/s) added to the 517 Shaft immediately following tracer injections. The
fast times for the advection front of the tracer movement from the 517 Shaft to the St. Louis Tunnel
portal suggest that in general the tunnel has few obstructions and water moves through it relatively
quickly as channel flow.
Figure 8 also displays field fluorometer data, which confirms the timing of tracer arrival and
subsequent rapid increase in uranine concentration. Unfortunately, the upper limit of calibration for
the field fluorometer was 400 ppb for this experiment so the instrument was unable to record the
peak concentrations of the tracer breakthrough. The field instrument was then able to pick up the
recession limb of the uranine slug. The general performance of the instrument was quantified by
determining that the instrument tracked the laboratory analyses well, though with a near linear 45%
reduction in concentration (r2 = 0.99) when sample concentrations were below 400 ppb. According
to the information provided by the manufacturer of the field fluorometer, the uranine probe was
designed to analyze samples using excitation/emission values of 485/540 nm, which is shifted from
the 492/512 nm values used for laboratory analysis. Therefore the field fluorometer may have failed
to accurately capture the uranine fluorescence peak, resulting in consistently lower measured
concentrations of uranine dye in the St. Louis Tunnel samples.
Figure 8. Time series of hourly discharge (top) and tracer break through curves (bottom)
for lithium and uranine at the St. Louis Tunnel portal. Time 0 represents 12:00 on 4 October.
189
Given an estimated tunnel distance of 2591 m between the 517 Shaft and the St. Louis Tunnel
portal, the average velocity of the advection front (maximal flow velocity) was 2.87 m/min while the
mean velocity of the lithium and uranine peaks were 2.16 m/min and 2.06 m/min, respectively.
Although time elapsed between tracer injection and passage of maximal-concentration provides a
good approximation of the mean velocity, it is recognized that the true effective flow velocity occurs
after the passage of maximal-concentration and cannot be calculated from the breakthrough curve
alone [18]. However, when a steep and narrow breakthrough curve occurs, as observed in Figure 8,
the correct time for calculating effective flow velocity is only insignificantly larger than the time of
passage of maximal-concentration and the difference can be practically neglected [18]. Therefore,
the approximated effective flow velocity was used, and when compared to results presented by
Wolkersdorfer [9] the effective flow velocity is considerably faster than the range of 0.3 to 1.6 m/min
reported as 95% confidence interval of 42 mine tracer tests. The results suggest this system differs
from those reviewed by Wolkersdorfer [9], or that the water chase applied behind tracer injection
may have artificially increased the velocity of tracers as they moved from the 517 Shaft towards the
St. Louis Tunnel. Interestingly, the discharge at the St. Louis Tunnel portal did not show a clear
increase or pulse in discharge upon tracer arrival (Figure 8). Hourly mean discharges between 01:00
and 24:00 on 5 October (the breakthrough period) fluctuated between 51 L/s and 54 L/s relative to
the mean of 52 L/s (ı = 26 L/s) over the first 100 h of tracer recovery. As a result, the artificial
influence of the water chase on tracer velocity was not directly quantifiable, making the estimated
effective flow velocity a qualitative estimate and influencing the decision not to use an established
model for solute transport simulation.
The final sample, collected 6 weeks after the tracer injection, showed that lithium concentrations
had returned to background levels (approximately 0.025 mg/L) while uranine concentrations were
still slightly elevated (§1.8 ppb) relative to background (<0.002 ppb). The results indicate that
6 weeks after the tracer injection the lithium tracer was no longer moving though the system at
concentrations greater than background, while the fluorescent dye tracer was still arriving at the
portal in detectable quantities. However, as of the last discharge measurement on 16 November,
cumulative flow and concentration data were used to calculate 74% and 109% recovery of lithium
and uranine respectively. This result conflicts the final instantaneous concentrations by suggesting
that all of the uranine was recovered while some of the lithium remained in the mine workings, either
through sorption or ending up in an immobile fluid region that was disconnected from the main
discharge conduit. Assuming the mine waters flow through a karst-like conduit system, the immobile
fluid regions could be conceptually described as resulting from vortices and eddies produced by
conduit surface irregularities [48].
Given that the St. Louis Tunnel discharge (Figure 8) showed considerable albeit irregular
fluctuations (min/max hourly discharges were 42.3 and 57.7 L/s) over the 6-week tracer recovery
period, it is reasonable to suggest that short-term variations in measured discharge may have created
uncertainty in quantifying tracer recovery. For short projects, such as this study, uncertainties in
cumulative flow can be upwards of 14% in a low gradient flow system if conditions are non-ideal [49].
The St. Louis tunnel discharge was measured directly downstream from the collapsed adit, with flow
emerging though piles of mine timber and debris before being channelized and directed through
190
a flume for quantification. There was no way to completely eliminate turbulence in the flow prior to
reaching the flume, which resulted in continuous small scale (§1–2 cm) water surface undulations
representing 5%–10% variability in total depth of flow at the flume. The variability in depth likely
created uncertainties in calculated flow and in the quantification of tracer recovery. Additionally,
there is inherent difficulty in calculating mass recovery of a fluorescent dye because the analytical
results are an indirect measurement of the dye itself. Variable dye concentrations produce different
intensities of fluorescence, which is then converted back to a mass of dye based on calibration curves
that do not have perfect fit with standards. Therefore a reasonable degree of error should be expected
when calculating mass recovery and results from this experiment are likely within that range. Given
that uranine concentrations remained slightly elevated at the end of the study period, it was not
possible to have recovered the entire tracer mass. Uranine concentrations on the final day of sampling
suggested that discharge from the St. Louis Tunnel portal was only producing approximately 8.16 g
of uranine per day, only 0.08% (per day) of the total mass of tracer added.
A second major concern with interpreting fluorescent dye results in a mine tracer study was the
effect of low pH waters on the fluorescence properties of the dyes. The results of laboratory
acidification tests (described in methods) agreed with previous work [16,18,21,23] by showing that
the uranine dye had its fluorescence (spectral signature) reduced when exposed to low pH waters.
However, results from the laboratory investigation showed that uranine fluorescence recovered when
acidic conditions were re-neutralized to the pH values observed in samples collected at the St. Louis
Tunnel portal (pH 7.34), which were likely maintained by reversible ion exchange reactions [18,23].
The study results therefore support previous recommendations from Naurath et al. [23] that uranine
can be successfully used in mine tracer studies when samples analyzed for tracer recovery are alkaline.
A final consideration for uranine recovery was the influence of mine water iron (Fe)
concentrations. The total recoverable Fe was approximately 66 mg/L in the 517 Shaft and only
4 mg/L at the St. Louis Tunnel portal during the study indicating that a significant portion of the iron
precipitated as iron hydroxides between the uranine tracer injection and recovery locations. The Fe
hydroxides can negatively influence uranine recovery rates due to dye adsorption on the hydroxide
surfaces and subsequent loss by filtration [36] or settling of precipitates into immobile fluid
regions of the conduit. All of the samples were filtered prior to fluorescence analysis to eliminate
any fluorescence intensity overestimates due to increased turbidity in the sample. However,
Naurath et al. [23] reported that good recovery can be expected in mine waters up to 100 mg/L of
Fe, so it is concluded that the Fe levels in the mine waters emerging at the St. Louis Tunnel portal
were not high enough to interfere with uranine results, as supported by the high recovery amounts
reported previously.
3.4. The St. Louis Tunnel Portal Discharge
Mean daily discharge at the St. Louis Tunnel portal from May to November 2011 ranged from 36
to 57 L/s with the peak discharge occurring on 21 September (Figure 9). Daily discharge from the
St. Louis Tunnel portal is plotted with the daily flows in Silver Creek at the downstream stream
gauge site (SC-493, Figure 1). Although the Silver Creek discharge record does not begin until
17 June, it is clear that the discharge record depicts a typical snowmelt hydrograph for Silver Creek
191
with discharge peaking in late spring (May or June) and returning to near base flow in August. For
comparison, the discharge of the Dolores River at Rico (USGS station 09165000) [50] had a similar
snowmelt hydrograph for 2011 with peak discharge occurring on 7 June, just 10 days before the
discharge record began on Silver Creek. However, at the St. Louis Tunnel portal, there is no
characteristic snowmelt pulse in the discharge. Rather, the hydrograph there has a more consistent
flow over time, with peak discharge occurring more than 3 months after peak discharge was recorded
in the local surface waters. This observation could be explained by a regional groundwater pulse
signal, which suggests that a large portion of the St. Louis Tunnel discharge is coming from or driven
by a more regional (i.e., watershed aquifer scale) groundwater reservoir.
The increase in pH and subsequent decrease in dissolved metals concentrations between the mine
workings and the St. Louis Tunnel portal also suggest dilution and perhaps titration of acidity by
non-AMD groundwater between the locations. The pH and metals concentrations support the notion
of increasing dilution as mine waters move downgradient from the Argentine Tunnel (above the
Blaine) to the Blaine Tunnel to the 517 Shaft to the St. Louis Tunnel portal, suggesting that the
workings to the south of Silver Creek may be the primary source of AMD and that mine water at the
500 level becomes diluted as it moves northeast towards the SE crosscut and the St. Louis Tunnel.
Figure 9. Mean daily discharge from the St. Louis Tunnel portal and from Silver Creek.
Silver Creek discharge was not measured between 27 August and 3 October 2011 as
pressure transducers had been removed from the gauging stations during this period.
Dilution along the drainage tunnels by considerable amounts of well-mixed non-AMD
groundwater is further supported by stable isotope results, which indicate a steady, low variability
signal at the St. Louis Tunnel portal and greater seasonal variability at the upper mine workings
(Blaine Tunnel and 517 Shaft). The potential for considerable amounts of non-AMD water entering
the system along the St. Louis Tunnel may also be explained by the regional hydrogeology.
192
The drainage tunnel extends several thousand feet into the Rico dome, which is comprised of a thick
series of interbedded sandstone, arkose, shale, limestone and dolomite of the Pennsylvania age
Hermosa formation [51]. One of the units of this formation is the Leadville Limestone, which is
locally 75 m thick and the top of the unit exhibits widespread karst and related erosional features [52].
The Leadville Limestone is also overlain by a sequence of clastic and calcareous sedimentary rocks
up to 1000 m thick [52]. The widespread limestone and sedimentary units may therefore support
a large regional aquifer. As a result, the St. Louis tunnel may likely be intercepting significant
amounts of groundwater (with low levels of mineralization relative to mine waters) moving through
the regional aquifer, leading to a dilution of the AMD water prior to emergence at the St. Louis
Tunnel portal. The delay in peak discharge from the St. Louis Tunnel relative to the local surface
waters would then be explained as being driven by the regional groundwater pulse which can lag
seasonal surface or near-surface water runoff in mountain watersheds [53].
3.5. Targeted Remediation
Targeted remediation refers to controlling the source of water (i.e., inflow of groundwater or
surface water) into underground workings or open pits and/or controlling the outflow of mine
drainage from the workings [1]. From the hydrologic characteristics identified by the multiple tracer
approach there are several options for future remediation. Results suggest that there is some degree
of hydrologic connectivity between the mine complex and Silver Creek, but it appears that even with
the creek being a no-change or losing reach across the mine complex that creek waters are not
contributing large amounts of source waters to the underlying mine workings. Therefore, it does not
seem necessary to artificially control streamflow in Silver Creek. Conversely, if the study found
significant contributions from Silver Creek, then targeted remediation options may have included
activities such as lining the stream channel with an impervious membrane or re-routing the stream
channel to minimize water exchange in the vicinity of the mine.
The most likely sources of water to the mine are therefore from localized (hillslope scale) meteoric
recharge into near surface mine workings to the southeast of Silver Creek. Results suggest that these
inputs are seasonally variable with the greatest inputs likely deriving from snowmelt, which is highly
spatially and temporally variable in mountain settings [54,55]. The variability of input, combined
with the highly variable nature of surface or near-surface mine entrance points (i.e., driven by
different rates of portal collapses) make it too expensive and difficult to control the most prevalent
source waters. Additionally, most of the degraded mine entrance points are on steep slopes with
limited access, so additional environmental impact would occur if the entrances were to be
adequately shut off.
The study results further suggest that the mine discharge at the St. Louis Tunnel portal is coming
from a combination of highly mineralized (and low pH) AMD waters and less mineralized,
well-mixed (from stable isotope results) groundwater. The tracers suggest that the highly mineralized
waters are moving in a south to north direction, or from the hillslope containing the Argentine/Blaine
Tunnels under Silver Creek towards the 517 Shaft area of the mine and along the deep drainage
tunnels to the St. Louis Tunnel portal. The remaining question is at what location does the highly
mineralized acidic mine water mix with clean, non-mine-impacted groundwater? By untangling
193
the “under-ground plumbing” in these instances, a source-control approach can be used to collect
and segregate the mineralized AMD water, before it can mix with and contaminate larger volumes
of naturally occurring groundwater flows [1].
A next step would be to perform rehabilitation of the mine workings associated with the Argentine
and Blaine Tunnels to increase safe access and provide confirmation and quantification of sources
and mixing points. It would also help to know the volume of mineralized AMD water leaving the
mine complex via the SE crosscut prior to any additional inflows occurring along the drainage
tunnels. If the contaminated water is significantly less than the volume of water exiting the St. Louis
Tunnel then active treatment of the lower volume of water may be the most economically viable
option. Currently, in-situ treatment to neutralize the acidity (§3 pH) of mine water in the 517 Shaft
is being performed. The results of this treatment are pending but if the water quality of the St. Louis
Tunnel discharge is improved then smaller scale on-site treatment of mine waters in this area may
provide the most feasible targeted remediation option.
Other targeted remediation options may include bulkheads to prevent the flow of contaminated
waters from ever reaching the St. Louis Tunnel portal. This requires an in-depth understanding of
both the structural integrity of host bedrock and regional groundwater elevations before mining and
theoretically after bulkhead installation. Installation of a bulkhead is therefore only acceptable if the
compounded waters do not back up and reemerge at unintended or uncontrollable locations.
5. Conclusions
Naturally occurring stable- and radio- isotope tracers, artificial salt and dye tracers, and synoptic
sampling of water quality parameters were utilized to understand the hydrologic connectivity of the
Rico-Argentine mine complex. Historical mining activities have led to the production of AMD at
this location and accurate hydrological characterization may enable targeted remediation to reduce
or eliminate the contaminated discharge. The natural tracer results suggest that the mine hydrology
is driven by recharge of recent meteoric waters while the mine discharge appears to be dominated by
a well-mixed groundwater signal. The synoptic sampling of water quality suggests that the most
highly contaminated mine water resides in the upper mine workings located in the hillside to the
southeast of Silver Creek and are diluted prior to discharging at the St. Louis Tunnel. Tracers applied
to Silver Creek indicated that the creek may be losing some water in the vicinity of the mine, but
were unable to confirm or reject that Silver Creek was directly contributing water to the mine. Tracers
applied directly to the mine workings indicated that mine water was moving in a south to north
direction from the workings on the Blaine Tunnel side of Silver Creek to the 517 Shaft area and
then draining to the St. Louis Tunnel portal. The concurrent application of a lithium salt and uranine
dye tracer to the mine workings produced similar results suggesting that both type of tracers can
be successfully used in acidic mine water conditions. However, it was concluded that mass recovery
of the uranine dye tracer was more difficult than mass recovery of the lithium tracer because
fluorescent dye mass is indirectly measured via fluorescence signal creating greater uncertainty in
recovery calculations.
In combination, this suite of natural and applied tracers can provide useful information on
complex hydrologic conditions that produce AMD at abandoned hardrock mine sites. The resulting
194
hydrologic characterization of the mine sites may ultimately reduce the impacts of AMD by
supporting targeted remediation efforts.
Acknowledgments
The authors of this study would like to thank three anonymous reviewers and Jeff Writer (USGS)
for greatly improving the quality of this manuscript. In addition the authors thank the following
partners: The U.S. EPA Region 8 Superfund Emergency Response Program; The USGS Toxic
Substance Hydrology Program; The Colorado Division of Reclamation, Mining, and Safety;
Jan Christner at URS Corporation; Anderson Engineering Company Inc.; The Kiowa Environmental
Laboratory of the NWT LTER program at the Institute of Arctic and Alpine Research, Boulder, CO;
The Mountain Studies Institute Silverton, CO; and numerous individuals who assisted in field
sampling and laboratory analysis for this project. The use of trade, product, or firm names in this
publication is for descriptive purposes only and does not imply endorsement by the U.S. government.
Conflicts of Interest
The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
Wireman, M.; Stover, B. Hard-rock mining and water resources. Ground Water 2011, 49, 310–316.
National Hardrock Mining Framework; U.S. EPA 833-B-97-003; U.S. EPA Office of Water:
Washington, DC, USA, 1997.
Riebsame, W. Atlas of the New West: Portrait of a Changing Region; W.W. Norton & Company:
New York, NY, USA, 1997; p. 192.
Kimball, B.A.; Runkel, R.L.; Walton-Day, K.; Bencala, K.E. Assessment of metal loads in
watersheds affected by acid mine drainage by using tracer injection and synoptic sampling:
Cement Creek, Colorado, USA. Appl. Geochem. 2002, 17, 1183–1207.
Jarvis, A.P.; Younger, P.L. Dominating chemical factors in mine water induced impoverishment
of the invertebrate fauna of two streams in the Durham coalfield, UK. Chem. Ecol. 1997, 13,
249–270.
Nordstrom, D.K. Acid rock drainage and climate change. J. Geochem. Explor. 2009, 100, 97–104.
Wolkersdorfer, C. Mine water tracing. In Mine Water Hydrogeology and Geochemistry;
Younger, P.L., Robins, N.S., Eds.; Geological Society: London, UK, 2002; pp. 47–60.
Field, M.S.; Wilhelm, R.G.; Quinlan, J.F.; Aley, T.J. An assessment of the potential adverse
properties of fluorescent tracer dyes used for groundwater tracing. Environ. Monit. Assess. 1995,
38, 75–96.
Wolkersdofer, C. Water Management at Abandoned Flooded Underground Mines—Fundamentals,
Tracer Tests, Modeling, Water Treatment; Springer: Heidelberg, Germany, 2008; p. 466.
195
10. Alvarado, J.A.C.; Purtschert, R.; Barbecot, F.; Chabault, C.; Ruedi, J.; Schneider, V.;
Aeschbach-Hertig, W.; Kipfer, R.; Loosli, H.H. Constraining the age distribution of highly
mixed groundwater using 39Ar: A multiple environmental tracer (3H/3He, 85Kr, 39Ar, and 14C)
study in the semiconfined Fontainebleau Sands Aquifer (France). Water Resour. Res. 2007,
43, 1–16.
11. Troldborg, L.; Jensen, K.H.; Engesgaard, P.; Refsgaard, J.C.; Hinsby, K. Using environmental
tracers in modeling flow in a complex shallow aquifer system. J. Hydrol. Eng. 2008, 13,
1037–1084.
12. Lavastre, V.; La Salle, C.; Michelot, J.; Giannesini, S.; Benedetti, L.; Lancelot, J.; Lavielle, B.;
Massault, M.; Thomas, B.; Gilabert, E.; et al. Establishing constraints on groundwater ages with
36Cl, 14C, 3H, and noble gases: A case study in the eastern Paris basin, France. Appl. Geochem.
2010, 25, 123–142.
13. McKnight, D.M.; Bencala, K.E. The chemistry of Iron, Aluminum, and dissolved organic
material in three acidic, metal-enriched, mountain streams, as controlled by watershed and
in-stream processes. Water Resour. Res. 1990, 26, 3087–3100.
14. Bencala, E.; McKnight, D.M.; Zellweger, G.W. Evaluation of natural tracers in an acidic and
metal-rich stream. Water Resour. Res. 1987, 23, 827–836.
15. Runkel, R.L.; Walton-Day, K.; Kimball, B.A.; Verplanck, P.L.; Nimick, D.A. Estimating
instream constituent loads using replicate synoptic sampling, Peru Creek, Colorado. J. Hydrol.
2013, 489, 26–41.
16. Smart, P.L.; Laidlaw, I.M.S. An evaluation of some fluorescent dyes used for water tracing.
Water Resour. Res. 1977, 13, 15–33.
17. Smart, P.L. Applications of Fluorescent Dyes in the planning and hydrologic appraisal of
sanitary landfills. Q. J. Eng. Geol. 1985, 18, 275–286.
18. Käß, W. Tracing Technique in Geohydrology; Balkema: Rotterdam, the Netherlands, 1998;
p. 581.
19. Knop, A. Über die hydrographischen Beziehungen zwischen der Donau und der Aachquelle im
badischen Oberlande. In Neues Jahrbuch Fur Mineralogie, Geologie und Paläontologie;
Leonhard, K.C.; Bronn, H.G.; Leonhard, G.; Geinitz, H.B., Eds.; E. Schweizerbart’sche
Verlagshandlung (E. Koch): Stuttgart, Germany, 1878, pp. 350–363.
20. Käß, W.A. Hydrologic tracing practice on underground contaminations. Environ. Geol. 1994,
23, 23–29.
21. Lyons, R.G. Identification and separation of water tracing dyes using pH response
characteristics. J. Hydrol. 1993, 152, 13–29.
22. Geyer, T.; Birk, S.; Licha, T.; Leidl, R.; Sauter, M. Multitracer test approach to characterize
reactive transport in Karst aquifers. Ground Water 2007, 45, 36–45.
23. Naurath, L.; Weidner, C.; Rüde, T.R.; Banning, A. A new approach to quantify na-fluorescein
(uranine) in acid mine waters. Mine Water Environ. 2011, 30, 231–236.
24. Bales, R.C.; Cline, D. Snow hydrology and water resources (Western United States). In Handbook
of Weather, Climate, and Water; Atmospheric Chemistry, Hydrology, and Societal Impacts;
Potter, T.D., Colman, B.R., Eds.; John While and Sons: Hoboken, NJ, USA, 2003; pp. 443–459.
196
25. Runkel, R.L.; Kimball, B.A.; Walton-Day, K.; Verplanck, P.L. A simulation-based approach
for estimating premining water quality: Red Mountain Creek, Colorado. Appl. Geochem. 2007,
22, 1899–1918.
26. Runkel, R.L.; Bencala, K.E.; Kimball, B.A.; Walton-Day, K.; Verplanck, P.L. A comparison of
pre- and post-remediation water quality, Mineral Creek, Colorado. Hydrol. Processes 2009, 23,
3319–3333.
27. Field, M.S. The QTRACER2 Program for the Tracer-Breakthrough Curve Analysis for Tracer
Tests in Karstic Aquifers and Other Hydrologic Systems; EPA/600/R-02/001; U.S.
Environmental Protection Agency—Office of Research and Development: Washington, DC,
USA, 2002; p. 179.
28. Field, M.S. Tracer-Test Planning Using the Efficient Hydrologic Tracer-Test Design (EHTD)
Program; EPA/600/R-03/034; U.S. Environmental Protection Agency—Office of Research
and Development; National Center for Environmental Assessment: Washington, DC, USA,
2003; p. 175.
29. Davis, M.W. The Use of Tracer Dyes for the Identification of a mine Flooding Problem, Rico,
Dolores County, Colorado; Colorado Geological Survey Open File Report 91-2; Colorado
Geological Survey: Lakewood, CO, USA, 1994; pp. 1–20.
30. Wolkersdorfer, C.; LeBlanc, J. Regulations, legislation, and guidelines for artificial surface
water and groundwater tracer tests in Canada. Water Qual. Res. J. Can. 2012, 47, 42–55.
31. Behrens, H.; Beims, U.; Dieter, H.; Dietze, G. Eikmann, T.; Grummt, T.; Hanisch, H.;
Henseling, H.; Käß, W.A.; Kerndorff, H.; et al. Toxicological and ecotoxicological assessment
of water tracers. Hydrogeol. J. 2001, 9, 321–325.
32. Flury, M.; Wai, N.N. Dyes as tracers for vadose zone hydrology. Rev. Geophys. 2003, 41, 1–33.
33. Davis, S.N.; Thompson, G.M.; Bentley, H.W.; Stiles, G. Ground-water tracers-A short review.
Ground Water 1980, 18, 14–23.
34. Sabatini, D.A.; Austin, T.A. Characteristics of rhodamine WT and fluorescein as adsorbing
ground-water tracers. Ground Water 1991, 29, 341–349.
35. Kasnavia, T.; Vu, D.; Sabatini, D.A. Fluorescent dye and media properties affecting sorption
and tracer selection. Ground Water 1999, 37, 376–381.
36. Aldous, P.J.; Smart, P.L. Tracing groundwater movement in abandoned coal mined aquifers
using fluorescent dyes. Ground Water 1988, 26, 172–178.
37. Davies, G. Results of Ground-Water Tracing Experiments in the Nelson-Wooster-Humphrey
Tunnel; Cambrian Ground Water Co.: Oak Ridge, TN, USA, 2002; p. 14.
38. Bencala, K.E.; McKnight, D.M.; Zellweger, G.W. Characterization of transport in an acidic and
metal-rich mountain stream bed on a lithium tracer injection and simulations of Transient
Storage. Water Resour. Res. 1990, 26, 989–1000.
39. Jarrett, R.D. Hydrologic and Hydraulic research in mountain rivers. Water Resour. Bull. 1990,
26, 419–429.
40. Marchand, J.P.; Jarrett, R.D.; Jones, L.L. Velocity Profile, Water-Surface Slope, and Bed-Material
Size for Selected Streams in Colorado; Open-File Report 1984-733; U.S. Geological Survey:
Lakewood, CO, USA, 1984; p. 82.
197
41. Kilpatrick, F.A.; Cobb, E.D. Chapter A16. Measurement of Discharge Using Tracers; USGS
Techniques of Water-Resources Investigations Reports, Book 3; U.S. Geological Survey:
Reston, VA, USA, 1985.
42. Palumbo-Roe, B.; Dearden, R. The hyporheic zone composition of a mining-impacted stream:
Evidence by multilevel sampling and DGT measurements. Appl. Geochem. 2013, 33, 330–345.
43. Liu, F.J.; Williams, M.W.; Caine, N. Source waters and flow paths in an alpine catchment,
Colorado Front Range, United States. Water Resour. Res. 2004, 40, 1–16.
44. Michel, R.L. Tritium Deposition in the Continental United States, 1953–1989; Water-Resources
Investigations Report 89-4072; U.S. Geological Survey: Reston, VA, USA, 1989.
45. Craig, H. Isotopic variations in meteoric waters. Science 1961, 133, 1702–1703.
46. Williams, M.W.; Knauf, M.; Caine, N.; Liu, F.J.; Verplanck, P. L. Geochemistry and source
waters of rock glacier outflow, Colorado Front Range. Permafr. Periglac.Process. 2006, 17,
13–33.
47. Eastoe, C.J.; Rodney, R. Isotopes as tracers of water origin in and near a regional carbonate
aquifer: The Southern Sacramento Mountains, New Mexico. Water 2014, 6, 301–323.
48. Field, M.S.; Pinsky, P.F. A two-region nonequlibrium model for solute transport in solution
conduits in karstic aquifers. J. Contam. Hydrol. 2000, 44, 329–351.
49. Brigand, F.G.; Lellouche, G.; Appelboom, T.W. Measuring flow in non-ideal conditions for
short-term projects: Uncertainties associated with the use of stage-discharge rating curves.
J. Hydrol. 2013, 503, 186–195.
50. USGS National Water Information System. Available online: http://waterdata.usgs.gov/nwis
(accessed on 1 June 2012).
51. McKnight, E.T. Geology and Ore Deposits of the Rico District, Colorado; Professional Paper
723; U.S. Geological Survey: Reston, VA, USA, 1974; p. 100.
52. Larson, P.B. Stable isotope and fluid inclusion investigations of epithermal vein and porphyry
molybdenum mineralization in the Rico Mining District, Colorado. Econ. Geol. 1987, 82,
2142–2157.
53. Hazen, J.M.; Williams, M.W.; Stover, B.; Wireman, M. Characterization of acid mine drainage
using a combination of hydrometric, chemical, and isotopic analysis, Mary Murphy mine,
Colorado. Environ. Geochem. Health 2002, 24, 1–22.
54. Williams, M.W.; Brown, A.D.; Melack, J.M. Biochemical modifications of snowmelt runoff in
an alpine basin. In Hydrological Interactions between Atmosphere, Soil and Vegetation; Kienitz,
G., Milly, P.C.D., van Genuchten, M.T., Rosbjerg, D., Shuttleworth, W.J., Eds.; International
Association of Hydrological Sciences: Wallingford, UK, 1991; IAHS-AIHS Publication 204,
pp. 457–465.
55. Caine, N. Streamflow patterns in the alpine environment of North Boulder Creek, Colorado
Front Range. Z. Geomorphol. 1996, 104, 27–42.
198
Quantifying Reaeration Rates in Alpine Streams Using
Deliberate Gas Tracer Experiments
Andrew Benson, Matthew Zane, Timothy E. Becker, Ate Visser, Stephanie H. Uriostegui,
Elizabeth DeRubeis, Jean E. Moran, Bradley K. Esser and Jordan F. Clark
Abstract: Gas exchange across the air-water interface is a critical process that maintains adequate
dissolved oxygen (DO) in the water column to support life. Oxygen reaeration rates can be accurately
measured using deliberate gas tracers, like sulfur hexafluoride (SF6) or xenon (Xe). Two continuous
release experiments were conducted in different creeks in the Sierra Nevada of California: Sagehen
Creek in September, 2009, using SF6 and Martis Creek in August, 2012, using both SF6 and Xe.
Measuring gas loss along the creek, which was approximated with the one-dimensional advectiondispersion equation, allows for the estimation of the SF6 or Xe reaeration coefficient (KSF6, KXe),
which is converted to DO reaeration (KDO or K2) using Schmidt numbers. Mean KSF6 for upper and
lower Sagehen and Martis Creeks were, respectively, 34 dayí1, 37 dayí1 and 33 dayí1, with
corresponding KDOs of 61 dayí1, 66 dayí1 and 47 dayí1. In Martis Creek, KXe was slightly higher
(21%) than KSF6, but the calculated KDO from SF6 agreed with the calculated KDO from Xe within
about 15%; this difference may be due to bubble-enhanced gas transfer. Established empirical
equations of KDO using stream characteristics did a poor job predicting KDO for both creeks.
Reprinted from Water. Cite as: Benson, A.; Zane, M.; Becker, T.E.; Visser, A.; Uriostegui, S.H.;
DeRubeis, E.; Moran, J.E.; Esser, B.K.; Clark, J.F. Quantifying Reaeration Rates in Alpine Streams
Using Deliberate Gas Tracer Experiments. Water 2014, 6, 1013-1027.
1. Introduction
Aquatic life requires adequate levels of dissolved oxygen (DO) for survival. Therefore, DO
content is a standard monitoring tool to determine the health of fresh water systems. A major
consumer of DO in aquatic systems is the microbial degradation of organic matter. A stream’s ability
to make up this oxygen deficit and return to its solubility equilibrium with the atmosphere is vital.
This is accomplished primarily through gas exchange at the air-water interface, where oxygen is
either lost or reabsorbed into the stream, due to the concentration gradient between the atmosphere
and water. If levels of demand exceed reaeration into the system, a body of water will struggle
to support life and may reach hypoxic conditions; in extreme cases, the surface water will
become anoxic.
Reaeration coefficients vary widely due to their dependence on turbulence at the air water
interface, which is poorly understood and hard to measure. Previous work has related reaeration
coefficients to stream characteristics, such as mean depth, current velocity, stream channel slope and
discharge (e.g., [1–5]). A number of equations have been postulated, with no one equation
appropriate for every channel. The empirical equations also disagree significantly within the same
channel. For instance, the equations respond differently to increasing flow: velocity-depth equations
predict decreasing reaeration rates, while energy-dissipation models (utilizing channel slope and
199
velocity) predict increasing rates (e.g., [3]). When applied to wastewater management, an
underestimation in the reaeration coefficient would result in overly restrictive regulations. With
overestimation, a stream or creek will be less resilient to wastewater discharge than predicted, risking
hypoxia and collapse of the aquatic ecosystem.
In addition to DO studies, high quality reaeration coefficients are needed for understanding most
biogeochemical cycles within streams and other surface water bodies. Many of the gases involved in
these cycles, such as CO2, N2O and CH4, are greenhouse gases. They are also needed for interpreting
222
Rn distributions that are commonly used to investigate groundwater-surface water interactions
(e.g., [6–8]). For instance, the results of the tracer experiment described here provided an estimate
of the loss rate of 222Rn from Martis Creek, which allowed for the quantification of the groundwater
influx when combined with 222Rn measurements in the stream and an estimate of the 222Rn concentration
in shallow groundwater [8].
Gas tracers can be used to accurately estimate a stream’s reaeration coefficient through point-source
injection as either a continuous release (as is the case here) or a single pulse that includes a
conservative ion, dye or a second gas tracer [9–12]. Tracer concentrations are measured at specific
intervals downstream from injection after the experimental reach has been flushed. Tracer loss from
the water to the atmosphere can be used as a proxy for gas exchange across the interface, ultimately
allowing for a calculation of reaeration rates.
Two trace gases, sulfur hexafluoride (SF6) and the noble gas, xenon (Xe), were selected for the
tracer experiments conducted in two headwater creeks in the Sierra Nevada of California. SF6 was
selected, because it has been used in numerous earlier experiments [9–11,13]. As discussed below,
it is a very strong greenhouse gas and is currently regulated in California, so during the second
experiment, a second gas tracer was tested to determine if it could be used as an alternative tracer for
these experiments. The additional objectives of this study are to compare the reaeration rates between
the two creeks and with empirical relationships, as others have done [1–5].
1.1. Study Area
The first experiment used only SF6 and was conducted in Sagehen Creek, a headwater catchment
to the Truckee River about 35 km northwest of Lake Tahoe. The second included both SF6 and Xe
and was conducted in another tributary to the Truckee River, Martis Creek, about 15 km southeast
of Sagehen near the town of Truckee, CA, USA (Figure 1a). The Sagehen experiment occurred along
an approximately 500-m reach of the creek in 2009, adjacent to the Sagehen Creek Field Station
(SCFS). The Martis experiment occurred along an approximately 1000-m reach in 2012 (prior to the
implementation of the SF6 regulations that began 1 January 2013). SCFS is part of the University of
California (UC) Natural Reserve system and is managed by UC Berkeley.
200
Figure 1. (a) The study area is located northwest of Lake Tahoe within the Sierra
Nevada. Main base map from the United States Geologic Survey (USGS) 10-m National
Elevation Dataset [14]. Experimental stream reaches of (b) Sagehen and (c) Martis Creeks.
Both creeks are on the eastern side of the Sierra Nevada at elevations above 1800 m and flow
through glacial till deposits (derived from andesite and granodiorite basement rocks) to the Truckee
River. They are shallow (mean reach depths of ca. 10 cm), meandering streams with riffle and pool
morphology and receive discharge perennially from shallow aquifers. At the time of the tracer
experiments, the mean water temperatures were 8.3 °C (Sagehen) and 15.8 °C (Martis), though
diurnal temperature fluctuations were observed. Both watersheds receive more than 80 cm of
precipitation per year, most of which falls as snow between October and April. Peak and minimum
discharge typically occur, respectively, during the month of May and September (as baseflow).
2. Materials and Methods
2.1. One-Dimensional Advection-Dispersion Equation
To obtain the reaeration coefficients for the gas tracers (SF6 and Xe), dissolved gas transport was
approximated using a one-dimensional (1D) advection-dispersion equation assuming first order
decay of a continuously released solute in a river; gas exchange is assumed to be a first order chemical
reaction. The 1D transport equation for a pollutant or tracer in a stream is written as (e.g., [11,15,16]):
∂C
∂2 C
∂C
+U
= E x 2 − KC
∂t
∂x
∂x
(1)
Where C is the concentration of the dissolved gas in excess of its solubility equilibrium value (i.e.,
C = Cob í Ceq, where Cob and Ceq are, respectively, the observed gas concentration and the solubility
equilibrium value); t is time; U is the mean stream velocity; x is the distance downstream; Ex is the
longitudinal dispersion coefficient; and K is the reaeration coefficient.
The concentration gradient (C/x) is small, because the tracer is being continuously released, and
thus, the flux due to dispersion is negligible when compared to the flux due to advection
(e.g., [15,16]). Once the stream reach has been flushed with tracer, the stream is assumed to be at a
201
steady state (i.e., C/t = 0), and samples can be collected. With these conditions, the
advection-dispersion equation can be simplified to:
U
dC
= −KC
dx
(2)
Solving this differential equation gives the relationship between concentration and sampling
distance, x, from injection:
C(x) = C0 exp ª¬(− K )x º¼
U
(3)
where C0 is the initial concentration of gas tracer at the injection point ( ݔൌ Ͳ).
Injecting the SF6 and Xe into the stream introduces a problem when calculating the appropriate
initial concentration (C0), because the injection rate is not known perfectly, especially when injecting
by bubbling, as was done during the 2009 Sagehen Creek experiment. C0 is obtained from the
downstream data. Fortunately, C0 is not needed to determine the reaeration coefficient.
To be useful for other applications and comparisons to earlier works, the reaeration coefficient for
the gas tracers must be converted to analogous values for other gases, such as DO, CO2, Rn and N2O.
This is done using Schmidt numbers (Sci, dimensionless ratios of the kinematic viscosity of water
and the diffusion coefficient of the gas in question). The reaeration rate of SF6 (KSF6) or Xe (KXe) is
related to other gases (Ki) by [1,9,17]:
§
·
Ki = ¨ Sci
© Sc j ¹̧
−0.5
Kj
(4)
where the Sci for each gas is defined at a given temperature usually using temperature relations
developed by Wanninkhof et al. [18] and Raymond et al. [1] and j refers to the gas tracer.
The method used to estimate the reaeration coefficient here is significantly different from most
methods reported in the literature, but similar to that used by Cook et al. [6]. This work uses a
continuous gas tracer source of SF6 and allows the concentration distribution to come to equilibrium.
A more common approach is to use instantaneous point sources of a gas tracer paired with either a
solute tracer or a second gas (e.g., [9–12]). The main advantages of using a continuous source are
that the solution to the differential equation (1) is much simpler than the solution for a point source,
only one gas tracer is needed and fewer samples need to be analyzed. Having a concentration in terms
of distance downstream makes it easy to calculate the reaeration coefficient graphically. The slope
of the line created when the log of the concentration is plotted versus distance is equal to –K/U, and
the y-intercept is C0. Most works present reaeration values normalized to a Schmidt number
of 600, which is similar to CO 2 at 20 °C in seawater or oxygen at 17.5 °C (KDO or K2) in freshwater;
we will also use this convention.
2.2. Tracer Injection
SF6 and Xe were utilized as the deliberate gas tracer in these experiments for a number of reasons.
SF6 is: (1) non-reactive; (2) non-toxic [19]; (3) man-made, so the background concentrations are
negligible, especially away from urban areas; (4) detectable at very low levels using gas
chromatography, so little gas needs to be released; and (5) relatively inexpensive. However, SF6
202
emissions are being regulated in California, because it is a strong greenhouse gas (~24,000 times
stronger than CO2 on a per molecule basis over a 100-y period [20]). Therefore, a second gas, Xe,
was released during the Martis Creek experiment to evaluate its potential as a deliberate tracer for
gas exchange. Xe, being a noble gas, is also non-reactive and non-toxic. The natural atmospheric
concentration is low (87 ppt by volume) and well established. Xe can be measured at atmospheric
abundance levels using membrane inlet mass spectrometry [21], so little gas needs to be released. While
the cost per liter is similar for the two gases, Xe is not a greenhouse gas and, therefore, not regulated.
Different methods were used to inject the gas tracers during the two experiments. During the
September 2009, Sagehen experiment, an injection system (Figure 2) was built and placed next to
the stream. Approximately a 1:10 SF6:N2 mixture was created (so that the concentration would be
within the calibration range of the analytical method) and released semi-continuously using a
switcher valve set to switch every 30 s and bubble into the stream through a diffusion stone placed
at the bottom of a small pool near mid-channel (see [13] for the details of the injection device).
Injection began the evening before the collection of the first samples and continued until after the
collection of the final set. The injection rate of the SF6 mixture was 0.34 mL/min for the first three
sets and then increased to 0.68 mL/min for the final set.
Based on 100% dissolution of the bubbles and the injection rate, the initial concentrations of SF6
should have been 4.3 × 106 pmol/L for the Sagehen Creek experiment. Because the gas was being
bubbled into the stream, most of it was lost to the atmosphere and only a small amount (less than
0.01%) of it actually dissolved into the stream. In the Martis Creek Experiment, tracer introduction
via gas permeable silicon tubing achieved a gas introduction efficiency of 100%.
Figure 2. The diagram of the sulfur hexafluoride (SF6) injectors used during the Sagehen
Creek (top) and Martis Creek (bottom) experiments.
The Martis Creek injection began at 14:15 on 14 August 2012. Injection occurred through 1/8”
inner diameter gas permeable silicon tubing fully submerged in the stream (Figure 2) following
methods similar to those of Cook et al. [6] and Clark [22]. The tubing was coiled and weighted down
with carabiners across the creek. The end of the tubing was knotted tightly, so that no large gas
bubbles could escape (and none were observed). Copper tubing was attached to the submerged silicon
203
tubing and fastened to the gas tracer tank on the creek bank. Separate setups of silicone tubing were
used for the SF6 and Xe tracers with 6 m of tubing submerged for SF6 and 11 m for Xe. The Xe
pressure in the silicone tubing was set to about 2 psi using a pressure regulator. A pressure gauge in
line with the copper tubing recorded a constant pressure of 15 psi over the course of the 3-day
injection in the N2-SF6 line. Once again, a 1:10 diluted SF6 mixture was used during this experiment.
2.3. Creek Sampling
For both SF6 experiments, eight or nine sampling locations were selected every 15–100 m
downstream of the injection apparatus (Figure 1). At each location, five samples were collected in
pre-weighed 10 mL Vacutainers™ evenly distributed across the channel 5–10 cm below the air-water
interface. The samples were collected by submerging the Vacutainers™, piercing the septa with a
needle, until 2–5 mL of stream water were collected. During the Sagehen experiment, samples were
collected during the morning, afternoon and evening of 11 September 2009. A fourth set was
collected the following morning on 12 September 2009, after the injection rate was changed.
The Xe samples were collected at the same locations as the SF6 samples during the Martis Creek
experiment. Duplicate samples were collected from the center of the stream by submersing an empty
40-mL glass vial upside down and filling it at 5–10 cm below the surface. The vial was capped under
water. At two downstream locations (Martis Creek (MC) + 1 and MC + 8), additional cross-section
samples were collected (in duplicate) at the left and right bank and in between the banks and the center.
During the Martis Creek experiment, sampling began about 1 h after the injection started. It was
interrupted by a 2-h rain event that dropped considerable water onto the field site. While this precipitation
event may have changed the creek flow, stream gauging shows that the discharge was the same on
Days 2 and 3, so if any changes occurred, they were short lived. The first full set was collected 18 h
after injection, with six complete sets collected over the next three days. Four sets were collected on
15 August 2012 (morning, afternoon with duplicates and evening), and one set each was collected
during the mornings of 16 August 2012, and 17 August 2012. Xe samples were collected on the
morning of 17 August 2012. Background sampling occurred at one location upstream of the injector
in Martis Creek. This paper will discuss only the 17 August 2012, data, when both gases
were analyzed.
2.4. Creek Measurements
To quantify channel geometry in Sagehen Creek, stream transect data were collected at
27 locations along the creek starting approximately 41 m upstream of the injection point and
continuing approximately 550 m downstream. At each transect, the stream width and the depth every
20 cm (from the north bank to the south bank) were measured. In addition, the grain size of the
streambed was descriptively characterized at each depth determination (for details, see [23]).
A United States Geological Survey (USGS) river gauging station located about 370 m downstream
from the injection point provided continuous 15-min discharge data, which showed little variation
and averaged 43 L/s during the experiment. The gauge also represents the place where the mean
channel geometry changes. In the upper reach, the average measured depth, cross-sectional area and
204
flow velocity were, respectively, 0.11 m, 0.41 m2 and 0.11 m/s, while below the gauge, they were,
respectively, 0.15 m, 0.49 m2 and 0.09 m/s.
The flow velocity in Martis Creek was determined by gauging with an FP111 Global Water™
flow probe at four stations (MC í 1, MC + 2, MC + 5, and MC + 8). Cross-sectional data were also
measured at these stations. The average depth, cross-sectional area, flow velocity and discharge were,
respectively, 0.11 m, 0.20 m2, 0.29 m/s and 57 L/s. Flow velocity varied from a low average of
0.14 m/s at MC + 2 to a high average of 0.65 m/s at MC + 5; the average velocity at a given
cross-sectional location did not vary significantly over the course of the experiment.
The channel gradient in Sagehen Creek was measured using a combination of GPS coordinates
and a 1-m digital elevation model (DEM) provided by the Sagehen Creek Field Station that was
created from aerial LiDAR flown in September 2005. The gradient was then calculated by dividing
the measured center channel length by the difference in elevation based on the DEM and was found
to be 0.0143; there was little change in slope between the upper and lower reaches. The channel slope
for Martis Creek was determined in a similar fashion, but using Google Earth rather than ArcGIS
and LiDAR elevations. It was determined to be 0.0145, very similar to Sagehen Creek, and it also
did not vary substantially along the creek.
At Martis Creek, two deep pools along the experimental reach were sampled at three depths (top,
middle and bottom) to assess vertical mixing. Little stratification was suspected, due to the stream’s
generally shallow depth. One pool was ~3 m downstream of MC + 4 with a maximum depth of
0.56 m; the other was ~1 m upstream of MC + 7 with a maximum depth of 0.66 m. Temperature and
oxygen measurements were also made using a YSI™ 556 multi-meter (Yellow Springs, OH, USA)
and averaged 16.3 °C and 15.4 °C for the two pools. This temperature difference was likely
due to the typical diurnal fluctuation found in most shallow creeks rather than an indication of
the groundwater discharge of cooler water.
2.5. Laboratory Analysis
SF6 samples were processed using the modified headspace method of Clark et al. [24], which
employs Vacutainers™. In order to minimize contamination, the containers were kept separate from
the injection equipment, during the transport to, from and while at the field sites. The laboratory
procedure was as follows: (1) each sample was weighed in order to calculate the volume of water
collected; (2) the headspace of each vial was filled with ultra-high purity N2 to atmospheric pressure
in the lab (about 1 atm); (3) vials were agitated by gently shaking in order to allow the gasses in the
headspace to become well-mixed; and (4) gases in the headspace were then fed into a gas chromatograph
system via displacement with water. SF6 standards of known concentrations (1.95 parts per billion
by volume (ppbv) and 10.0 ppbv) purchased from and certified by Scott-Marrin, Inc. Riverside, CA,
USA) were run every ~10 samples to ensure accurate instrument calibration. Ultra-high purity
nitrogen was used to flush the system of residual SF6 between runs. The analytical uncertainty of the
Vacutainer™ method is typically better than ±5%.
205
Xe samples were analyzed three days after collection on 20 August 2012, using a noble gas
membrane inlet mass spectrometer (NG-MIMS) [21]. The NG-MIMS system consists of a membrane
inlet, a dry ice water trap, a liquid nitrogen carbon-dioxide trap, two getters, a gate valve,
a turbomolecular pump and a quadrupole mass spectrometer equipped with an electron multiplier.
Vials were opened, and the sample was withdrawn from the bottom of the vial through the membrane
inlet for 5 min at 0.5 mL/min. The last two minutes of measurements were averaged to reduce
measurement noise. Dissolved Xe concentrations are determined from measurements made every
10 seconds at mass over charge ratios of 124 and 132, by comparing against air equilibrated water
(AEW) standards, with a laboratory determined uncertainty of ±8%. The linearity of the NG-MIMS
was demonstrated across a wide range of noble gas concentrations [21].
3. Results and Discussion
As expected, average SF6 and Xe concentrations at each sampling site decreased downstream
(Table 1, Figures 3 and 4). The standard deviation is used as a measure of horizontal mixing. When
the deviation of bank-to-bank concentrations at each sampling location stabilizes, the tracer is
assumed to be well-mixed across the creek. For all sets collected during the Sagehen Creek
experiment, the standard deviation of the first sample location (SC + 1) ranges from 20% to 28% of
the cross-sectional mean, while, with a few exceptions, the standard deviation for the rest of the
samples is generally less than 5% of the mean, equivalent to analytical uncertainty. The first sample
location is close to the injector (15 m downstream), and SF6 had yet to become well mixed across
the stream. By the second sampling location (44 m downstream), it had. The first samples were
collected further downstream during the Martis Creek Experiment (69 m vs. 15 m) and showed a
lower standard deviation (<15%) across the creek. The standard deviation of Xe concentrations in
the first (MC + 1) and last (MC + 8) cross-sections was 14% and 10%, respectively. The duplicate
reproducibility was 9% for the Xe samples.
In Sagehen Creek, at the location immediately upstream (SC + 6) and downstream (SC + 7) of the
USGS stream gauge (about 370 m downstream injector), the drop in concentrations of SF6 is quite
large over a very short distance. This drop indicates a higher reaeration rate, due to the waterfall
created by the gauge, as is commonly observed at spillways [25].
Furthermore, in the large pool immediately upstream of the gauge, the concentration suddenly
increased for samples from Sets 1–3. After further inspection, the increases in concentration were
coincident with a method change that required the use of different standards. It is important to note
that after the offset in standards was found, the fourth set was analyzed using the same method (high
standard) from the top to the bottom of the reach. In this paper, only the fourth set is discussed.
206
Table 1. Average tracer concentrations and standard deviations (SD) at each cross-section
during the 12 September 2009, Sagehen Creek (SC) and 17 August 2012, Martis Creek
(MC) experiments.
[SF6]
SD
[Xe]
SD
(%)
(nmol/L)
(%)
0
0.0
0.46
-
315
20
43.5
14
325
2
40.2
-
270
3
34.1
-
352
244
2
29.9
-
481
203
2
24.4
-
MC + 6
641
169
2
19.9
-
MC + 7
798
125
3
13.5
-
MC + 8
964
111
3
10.8
10
Statio
Distance
[SF6]
SD
n
(m)
(pmol/L)
(%)
SC + 1
16
118
27
MC í 1
SC + 2
44
74.9
7.8
MC + 1
69
SC + 3
114
55.9
2.8
MC + 2
139
SC + 4
180
42.0
4.3
MC + 3
257
SC + 5
294
27.1
2.8
MC + 4
SC + 6
364
23.9
5.5
MC + 5
SC + 7
372
20.5
4.9
SC + 8
445
13.5
5.6
SC + 9
507
10.5
2.4
Station
Distanc
e (m)
(pmol/L
)
The semi-log plot of the concentration for Sagehen Creek shows that the slope (íK/U) for the
lower reach is generally steeper than that of the upper reach (Figure 3). In fact, it is about 22% larger.
Detailed analysis of the channel geometry, as noted above, revealed that the mean depth of the lower
portion (0.15 m) was about 35% larger than the upper section (0.11 m).
Figure 3. Semi-log plot of SF6 concentrations with least squares fit trend lines showing
the different slopes (íK/U) for the upper and lower reaches in Sagehen Creek (red
squares from 12 September 2009) and for Martis Creek (blue circles from 17 August
2012). In the equations, “x” is the distance downstream.
In Martis Creek, relatively more tracer degassing occurred per meter between stations MC + 6
and MC + 7 than elsewhere along the creek. A few locations of riffles with white water that accelerate
gas transfer across the air-water interface were noted between these stations. A comparison of SF6
and Xe tracers show more similarities than differences (Figure 4). Both gases show reaeration, with
207
SF6 degassing at a faster rate. This is most apparent when examining the reaeration coefficient
normalized for DO (Table 2).
Reaeration rates are calculated from the semi-log plot of gas concentration versus sampling
distance (Figures 3 and 4). The slope of the exponential best-fit equals íK/U, where K is the
reaeration rate and U is mean velocity. Table 2 presents mean stream characteristics and reaeration
rates by sampling date and location.
Figure 4. Semi-log plot of SF6 (filled blue circles) and Xe (open red squares)
concentrations with trend lines for the Martis Creek 17 August 2012, sampling event.
The station closest to the injector was not used for the least squares fit and that only
center samples were used for calculating the Xe distribution. In the equations, “x” is the
distance downstream.
Table 2. Reaeration (K) values for the gas tracers (KSF6 or Xe ) and normalized to dissolved
oxygen (DO) (KDO) using temperature-dependent Schmidt numbers of 1875 and 1192 for
SF6 (Sagehen and Martis Creek), 1230 for Xe (Martis Creek) and 600 for DO (Sagehen
and Martis Creek) and Equation (4).
Discharge
Velocity U
Depth
íK/U
KSF6 or Xe
KDO
Location
Sampling Date
Tracer
Sagehen-Upper
12 September 2009
SF6
Sagehen-Lower
12 September 2009
SF6
43
0.09
0.15
0.00496
37
66
Martis Creek
17 August 2012
SF6
57
0.29
0.11
0.00133
33
47
Martis Creek
17 August 2012
Xe
57
0.29
0.11
0.00156
40
57
í1
(L/s)
(m/s)
(m)
(m )
(day )
(dayí1)
43
0.11
0.11
0.00407
34
61
í1
For comparison, KDO values for Sagehen and Martis Creeks were calculated using empirical
relationships for the experimental reaches (Table 3, Figure 5). Predictive equations show a very wide
range of values for KDO (6 to142 dayí1) with a mean of 43 ± 34 dayí1 for Sagehen Creek and (22 to
544 dayí1) with a mean of 117 ± 131 dayí1 for Martis Creek. In both creeks, the Thyssen-Jeppesen [26]
208
relationship appears to be an outlier. Removing that relationship from the data set reduces the range
and mean to 6–74 dayí1 (35 ± 20 dayí1) and 22–147 dayí1 (84 ± 48 dayí1), respectively, for Sagehen
and Martis Creeks.
Table 3. Predictive equations for reaeration (KDO) compiled by Raymond et al. [1] and
Cox [4]. See these publications for the citations to the original papers. Please note:
Raymond et al. [1] models calculate gas transfer velocities (m dayí1) rather than
reaeration coefficients, so the equations listed have been converted, and only the
equations that matched their metadata set with a R2 > 0.7 are presented.
Sagehen Creek
Reference
Predictive Equation
[27]
KDO = 3.93U0.5 Hí1.5
28
[28]
5.026UHí1.67
KDO =
KDO
(dayí1)
Martis Creek
KDO (dayí1)
58
16
59
[29]
KDO = 5.35UV0.67 Hí1.85
54
139
[30]
KDO = 5.5773U0.607 Hí1.689
47
109
[31]
KDO = 3170 S
45
46
[32]
KDO = 22,700 SU
32
95
[33]
KDO = 8784S0.93U0.734Hí0.42
74
147
[34]
KDO = 596(SU) 0.528Qí0.136
12
22
16
27
[35]
2
KDO = 23.04(1 + 0.17 F )(US
í1
)H (1.0212)
F)2.66S1.13Hí0.6
(Tí20)
142
544
[1] #1
KDO = 5037 (SU)0.89Dí0.46
34
107
[1] #2
KDO = 5937 (1í2.54 F)(SU)0.89Dí0.42
37
92
[1] #7
4725(SU)0.86Qí0.14
48
135
[36]
KDO = 23,000(1 + 0.17
0.375
KDO =
Dí0.34
Mean and Standard Deviation *
SF6 tracer
Xe tracer
43 ± 34
61, 66
†
117 ± 131
47
57
Notes: velocity (U, m/s); mean depth (H, m); stream gradient (S, m/m); Froude number (F); Temperature (T); * by removing
Thyssen-Jeppesen [36], the mean and standard deviation become 35 ± 20 and 84 ± 48 for Sagehen and Martis Creeks; † for upper
and lower Sagehen Creek.
Despite the fact that most predictive equations produced wildly differing reaeration rates,
underestimation is better than overestimation when considering wastewater discharge to streams.
Over predictions could lead to assumptions of a higher level of resiliency and less stringent and
potentially dangerous management strategies. Under-predictions, on the other hand, would lead to
overly precautious management, which, while inaccurate, would not threaten stream health.
209
Figure 5. A comparison of the field KDO (filled circles) listed in Table 2 with the
model predicted KDO (open circles) listed in Table 3, with the exception that the
Thyssen-Jeppesen [36] model values are not shown or used in the calculation of median
or quartile values.
4. Implications and Conclusions
Sagehen and Martis Creeks are very similar in terms of their morphology (both have pool and
riffle characteristics and similar channel slopes) and hydrology (with much of the late summer flow
derived from groundwater discharge and a similar mean depth at the time of the experiments).
As shown in Table 2, the discharge and current velocity in Martis Creek were higher (respectively,
57 vs. 43 L/s and 0.3 vs. 0.1 m/s). The SF6 tracer experiments revealed that the reaeration coefficients
corrected to a Schmidt number of 600 (KDO) differed substantially, with Sagehen Creek (>60 dayí1)
being larger than Martis (<50 dayí1). Interestingly, the results of the Xe tracer experiment (57 dayí1)
agree better with the SF6 result from Sagehen Creek (61–66 dayí1) than from Martis Creek (47 dayí1).
This difference between the SF6 and Xe results could be due to the bubble-enhanced gas transfer [26]
that generally removes lower solubility gases faster (SF6 is less soluble than Xe).
Most empirical formulas of reaeration rates, expressed as KDO, rely on an assessment of the current
velocity. The actual reaeration rate over a stretch of river or stream is directly observed as a change
in tracer concentration. Established empirical formulas of KDO using stream characteristics did a poor
job of predicting KDO for the two creeks. Although some agree with the field experiments in either
Sagehen or Martis Creek, no one relationship worked well for both of them.
Dissolved oxygen is one of the most important indicators of a stream’s biologic potential and
health. The United States Environmental Protection Agency’s criteria for dissolved oxygen are
designed to protect freshwater aquatic life under “worst case conditions”; the suggested DO limit
expressed as a seven-day mean minimum for cold water is 4.0 mg/L [37]. In this study, DO
concentrations in Martis Creek ranged between 5 mg/L and 9 mg/L throughout the summer.
Considering that minimum criteria estimations are conservative, DO content in this stream reach is
sometimes close to threatening levels. This exemplifies the need for accurate measurements of
210
reaeration to ensure a healthy aquatic environment. Future expansion of development in the
Tahoe and Truckee areas could also increase biological oxygen demand on the creek and lead
to noncompliance.
With mounting concerns and subsequent regulations of greenhouse gasses, the future of SF6 as a
hydrologic tracer is in doubt. Noble gasses, such as helium and Xe, are on their way to replacing SF6,
and we demonstrate that Xe is a viable replacement option for reaeration studies, as shown by
others [12]. Future SF6 use may depend on more efficient injection methods. Traditional bubbling
results in an immediate loss of most of the gas to the atmosphere, requiring more SF6 for water
tagging. The semi-permeable silicon tubing, used in the Martis Creek study, is very promising in this
regard and results in very high injection efficiency.
Acknowledgments
We would like to thank Tom Gleeson (McGill University) and Andrew Manning (USGS, Denver,
CO, USA) for their help with collecting samples and measuring stream transects. We are grateful to
Sagehen Creek Field Station; especially, station manager Jeff Brown, for technical and logistical
assistance. Gina Lee helped with some concepts and GIS. We are also grateful for the time and effort
of the two anonymous reviewers. Their comments helped to improve the paper substantially.
Financial support for this project was provided by the California Energy Commission Public
Interest Energy Research (PIER) Project PIR-08-010, the WateReuse Research Foundation project
WRF 09-11, the State of California Groundwater Ambient Monitoring & Assessment (GAMA)
Special Studies Program, and the Lawrence Livermore National Laboratory Lawrence (LLNL) Scholar
Program. Parts of this project were performed by LLNL under Contract DE-AC52-07NA27344.
Conflicts of Interest
The authors declare no conflict of interest.
References
1.
2.
3.
4.
Raymond, P.A.; Zappa, C.J.; Butman, D.; Bott, T.L.; Potter, J.; Mulholland, P.; Laursen, A.E.;
McDowell, W.H.; Newbold, D. Scaling the gas transfer velocity and hydraulic geometry in
streams and small rivers. Limnol. Oceanogr. Fluids Environ. 2012, 2, 41–53.
Moog, D.B.; Jirka, G.H. Analysis of Reaeration Equations Using Mean Multiplicative Error. In:
Air-Water Gas Transfer; Jähne, B., Monahan, E.C., Eds.; AEON Verlag & Studio: Hanau,
Germany, 1995; pp. 101–111.
St. John, J.P.; Gallagher, T.W.; Paquin, P.R. The Sensitivity of the Dissolved Oxygen Balance
to Predictive Reaeration Equations. In Gas Transfer at Water Surfaces; Brutsaert, W.,
Jirka, G.H., Eds.; D. Reidel Publishing Company: Dordrect, The Netherland, 1984; pp. 577–588.
Cox, B.A. A review of dissolved oxygen modelling techniques for lowland rivers.
Sci. Total Environ. 2003, 314, 303–334.
211
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Haider, H.; Ali, W.; Haydar, S. Evaluation of various relationships of reaeration rate coefficient
for modeling dissolved oxygen in a river with extreme flow variations in Pakistan. Hydrol.
Process. 2013, 27, 3949–3963.
Cook, P.G.; Lamontagne, S.; Berhane, D.; Clark, J.F. Quantifying groundwater discharge to
Cockburn River, southeastern Australia, using dissolved gas tracers 222Rn and SF6. Water
Resour. Res. 2006, 42, 12.
Cartwright, I.; Hofmann, H.; Sirianos, M.A.; Weaver, T.R.; Simmons, C.T. Geochemical and
222
Rn constraints on baseflow to the Murray River, Australia, and timescales for the decay of
low-salinity groundwater lenses. J. Hydrol. 2011, 405, 333–343.
DeRubeis, E.A. Radon as a Tracer of Groundwater-Surface Water Interaction in Martis Valley.
Master’s Thesis, California State University, East Bay, CA, USA, 4 October 2013.
Wanninkhof, R.; Mulholland, P.J.; Elwood, J.W. Gas exchange rates for a first-order stream
determined with deliberate and natural tracers. Water Resour. Res. 1990, 26, 1621–1630.
Clark, J.F.; Wanninkhof, R.; Schlosser, P.; Simpson, H.J. Gas-exchange rates in the tidal
Hudson River using a dual tracer technique. Tellus Ser. B Chem. Phys. Meteorol. 1994, 46,
274–285.
Hibbs, D.E.; Parkhill, K.L.; Gulliver, J.S. Sulfur hexafluoride gas tracer studies in streams.
J. Environ. Eng. Asce 1998, 124, 752–760.
Reid, S.E.; Mackinnon, P.A.; Elliot, T. Direct measurements of reaeration rates using noble gas
tracers in the River Lagan, Northern Ireland. Water Environ. J. 2007, 21, 182–191.
Clark, J.F.; Hudson, G.B.; Avisar, D. Gas transport below artificial recharge ponds: Insights
from dissolved noble gases and a dual gas (SF6 and 3He) tracer experiment. Environ. Sci.
Technol. 2005, 39, 3939–3945.
Gesch, D.B. The National Elevation Dataset. In Digital Elevation Model Technologies and
Applications: The DEM Users Manual, 2nd ed.; Maune, D., Ed.; American Society for
Photogrammetry and Remote Sensing: Bethesda, MD, USA, 2007; pp. 99–118.
Rutherford, J.C. River Mixing; Wiley: New York, NY, USA, 1994; p. 347.
Thomann, R.V.; Mueller, J.A. Principles of Surface Water Qality Modeling and Control;
Harper & Row: New York City, NY, USA, 1987; p. 644.
Jähne, B.; Heinz, G.; Dietrich, W. Measurement of the diffusion coefficients of sparingly soluble
gases in water. J. Geophys. Res. Oceans 1987, 92, 10767–10776.
Wanninkhof, R. Relationship between wind-speed and gas-exchange over the ocean. J. Geophys.
Res. Oceans 1992, 97, 7373–7382.
Lester, D.; Greenberg, L.A. The toxicity of sulfur hexafluoride. Arch. Ind. Hyg. Occup. Med.
1950, 2, 348–349.
Houghton, J.T.; Filho, L.G.M.; Bruce, J.P.; Lee, H.; Callander, B.A.; Haites, E.F.; Harris, N.;
Maskell, K. Climate Change 1994: Radiative Forcing of Climate Change and an Evaluation of the
IPCC 1992 IS92 Emission Scenarios; Cambridge University Press: Cambridge, UK, 1995; p. 339.
Visser, A.; Singleton, M.J.; Hillegonds, D.J.; Velsko, C.A.; Moran, J.E.; Esser, B.K.
A membrane inlet mass spectrometry system for noble gases at natural abundances in gas and
water samples. Rapid Commun. Mass Spectrom. 2013, 27, 2472–2482.
212
22. Clark, J.F. Defining Transport near ASR Operations Using Sulfur Hexafluoride Gas Tracer
Experiments. In Management of Aquifer Recharge for Sustainability (4th International
Symposium on Artificial Recharge of Groundwater); Dillon, P.J., Ed.; A.A. Balkema Publishers:
Exton, PA, USA, 2002; pp. 257–260.
23. Zane, M. Reaeration of Sagehen Creek near Truckee, CA. Bachelor’s Thesis, University of
California, Santa Barbara, CA, USA, 11 June 2010.
24. Clark, J.F.; Hudson, G.B.; Davisson, M.L.; Woodside, G.; Herndon, R. Geochemical imaging
of flow near an artificial recharge facility, Orange County, California. Ground Water 2004, 42,
167–174.
25. Rindels, A.J.; Gulliver, J.S. Oxygen Transfer at Spillways. In Proceedings of the Second
International Symposium on Gas Transfer at Water Surfaces, Minneapolis, MN, USA, 11–14
September 1990; American Society of Civil Engineers: New York, NY, USA, 2001; pp. 524–533.
26. Ho, D.T.; Asher, W.E.; Bliven, L.F.; Schlosser, P.; Gordan, E.L. On mechanisms of
rain-induced air-water gas exchange. J. Geophys. Res. Oceans 2000, 105, 24045–24057.
27. O’Conner, D.J.; Dobbins, W.E. Mechanisms of reaeration in natural streams. Trans. Am. Soc.
Civil Eng. 1956, 123, 641–667.
28. Churchill, M.A.; Elmore, H.L.; Buckingham, R.A. Prediction os stream reaeration rates. Int. J.
Air Water Pollut. 1962, 6, 467–504.
29. Owens, M.; Edwards, R.; Gibbs, J. Some reaeration studies in streams. Int. J. Air Water Pollut.
1964, 8, 469–486.
30. Bennett, J.P.; Rathburn, R.E. Rearation in Open-Channel Flow, USGS Professional Paper: No. 737;
United States Geological Survey: Washington, DC, USA, 1972.
31. Tsiovoglou, E.C.; Neal, L.A. Tracer measurements of reaeration: III—Prediciting the reaeration
capacity of inland streams. J. Water Pollut. Control. Fed. 1976, 48, 2669–2689.
32. Grant, R.S. Reaeration-Coefficient Measurement of 10 Small Streams in Wisconsin Using
Radioactive Tracers with a Section of the Energy Dissipation Model; U.S. Geological Survey
Water-Resources Investigations 76–96; United States Geological Survey: Madison, WI, USA, 1976.
33. Thyssen, N.; Erlandsen, M.; Jeppesen, E.; Ursin, C. Reaeration of oxygen in shallow,
macrophyte ricj streams: I—Determination of the reaeration rate coefficient. Int. Revue Ges.
Hydrobiol. 1987, 72, 405–429.
34. Melching, C.S.; Flores, H.E. Reaeration equations derived from US Geological Survey database.
J. Envirn. Eng. 1999, 125, 407–414.
35. Parkhurst, J.D.; Pomeroy, R.D. Oxygen absorption of streams—State of the art. J. Sanit. Eng.
Div. ASCE 1972, 98, 101–124.
36. Thyssen, N.; Jeppesen, E. Reaeration in small water courses. Vatten 1980, 36, 231–248.
37. US Environmental Protection Agency. Quality Criteria for Water 1986: The Gold Book; US
Environmental Protection Agency: Washington, DC, USA, 1987; p. 477.
213
Investigation of Groundwater Flow Variations near a
Recharge Pond with Repeat Deliberate Tracer Experiments
Jordan F. Clark, Sheila Morrissey, Jason Dadakis, Adam Hutchinson and Roy Herndon
Abstract: Determining hydraulic connections and travel times between recharge facilities and
production wells has become increasingly important for permitting and operating managed aquifer
recharge (MAR) sites, a water supply strategy that transfers surface water into aquifers for storage
and later extraction. This knowledge is critical for examining water quality changes and assessing
the potential for future contamination. Deliberate tracer experiments are the best method for
determining travel times and identifying preferential flow paths between recharge sites over the time
scales of weeks to a few years. This paper compares the results of two deliberate tracer experiments
at Kraemer Basin, Orange County, CA, USA. Results from the first experiment, which was
conducted in October 1998, showed that a region of highly transmissive sedimentary material
extends down gradient from the basin for more than 3 km [1]. Mean groundwater velocities were
determined to be approximately 2 km/year in this region based on the arrival time of the tracer center
of mass. A second experiment was initiated in January 2008 to determine if travel times from this
basin to monitoring and production wells changed during the past decade in response to new recharge
conditions. Results indicate that flow near Kraemer Basin was stable, and travel times to most wells
determined during both experiments agree within the experimental uncertainty.
Reprinted from Water. Cite as: Clark, J.F.; Morrissey, S.; Dadakis, J.; Hutchinson, A.; Herndon, R.
Investigation of Groundwater Flow Variations near a Recharge Pond with Repeat Deliberate Tracer
Experiments. Water 2014, 6, 1826-1839.
1. Introduction
Groundwater has been a primary source of potable and irrigation water for centuries. During the
last 50 years, the soaring demand for freshwater has placed unprecedented stresses upon many
aquifers throughout the world; many aquifers are now in overdraft. The projected growth in
population combined with uncertainties associated with a changing climate will only aggravate this
problem. A cost-effective advancement in groundwater/surface water management aimed at
augmenting local water supplies is managed aquifer recharge (MAR), the practice of artificially
recharging imported surface water, reclaimed (recycled) wastewater, or storm runoff into aquifers
for storage and later extraction [2–4].
For a MAR operation to be successful, a few challenges must be overcome. First, a source of
recharge water must be found. Second, facilities, which can rapidly transfer water into aquifers, must
be engineered. These include injection wells and spreading basins (recharge ponds), with both
requiring periodic removal of accumulated clogging material to maintain recharge rates. Third,
because the available water for the recharge operation can be of lesser quality than the local
groundwater supply, there is the potential to degrade the existing aquifer. The introduction of salts,
pathogens, disinfection by-products, and trace organic compounds such as pharmaceuticals, is a
214
concern, especially when urban runoff or reclaimed wastewater is a large component of the source
water for the operation [4]. Lesser quality sources can become a larger portion of the recharge water
at MAR operations because the availability of higher quality water such as imported water from
remote watersheds is limited and may shrink due to shifts in climate and the diversion of this water
to other uses such as maintaining riparian ecosystems (i.e., loss of California State Water Project
supplies to the delta smelt). Even in cases where the recharge water may be of a higher quality,
it may have a different geochemical character (e.g., redox state) than the ambient groundwater and
the potential exists for mobilization of intrinsic constituents, such as arsenic. For these reasons, it is
vital to understand the fate and transport of potential contaminants near MAR sites. Only from this
understanding can cost effective and appropriate regulations be developed.
Recent water quality studies near MAR operations have shown that many potential contaminants
such as dissolved organic carbon (DOC), nitrate, some pharmaceuticals, and most pathogens are
naturally removed or become inactive with time and distance in the subsurface (e.g., [4–9]). These
water quality improvements, known as soil-aquifer treatment (SAT), are considered one of the
benefits of MAR [3,4] and are generally observed soon after recharge, near the facility.
Despite these studies, water quality concerns still remain the focus of many regulations and an
obstacle for the permitting of new MAR operations that include a reuse component (i.e., reclaimed
wastewater). For instance, MAR operations in California, USA, must conform to the state Drinking
Water Program’s Groundwater Recharge Reuse Regulations, which require a subsurface retention
time prior to its extraction for a potable supply that varies based on the level of pre-recharge treatment,
in addition to a number of source water controls and how the travel times is being assessed [10].
The required retention time varies from two to six months and must be verified using a tracer study.
1.1. Travel Time Estimates
Development of field methodologies to evaluate flow near MAR facilities is critical for the
effective management of these operations where water quality assessment is warranted. The
California Drinking Water Program’s Groundwater Recharge Reuse Regulations place a greater level
of confidence on tracer data than on numerical models or Darcy flow calculations (see Table 2 in
reference [10]). Geochemical techniques provide a fundamental approach for investigating travel
times, flow paths, recharge rates, and dispersivity in groundwater. Intrinsic environmental tracers
such as Tritium/3He (T/3He) dating (e.g., [11,12]) are ideal for examining flow over spatial scales of
kilometers and temporal scales of years to decades near MAR operations [1,13,14]. Other
environmental tracer dating techniques such as chlorofluorocarbons (CFCs) or sulfur hexafluoride
(SF6) are likely to be unreliable near MAR operations because non-atmospheric sources of these
tracers may complicate the interpretation of apparent ages, especially when chlorinated reclaimed or
potable water is being recharged (e.g., [15–17]).
Because of the typical uncertainty associated with geochemical dating techniques (generally no
better than ±2 years), these methods are not well suited for examining short-term transport specified
by the Recharge Reuse Regulations in the state of California [14]. Deliberate (i.e., injected or applied)
tracer experiments using gases such as SF6 or noble gases (i.e., He or Xe isotopes) have become
acceptable methods for evaluating transport from recharge locations to wells over periods of weeks
215
to a few years [1,10,18,19]. There is a long history of using SF6 as a deliberate tracer starting with
atmospheric experiments and expanding into aqueous systems initially by the gas exchange and
oceanographic communities [20–23]. SF6 can also be used as an intrinsic or environmental tracer for
dating water in a similar fashion as CFCs [17].
The scale of deliberate tracer experiments is defined by the quantity of water that can be “tagged”
and the signal to noise ratio of the tracer being used. The three factors that often limit an experiment’s
scale are: (1) identification of a tracer that does not adversely impact potable aquifers; (2) the cost of
tracer; and (3) the ability to introduce a sufficient amount of tracer without significantly changing
the buoyancy (or density) of the tagged water. The cost of the tracer can be a particular problem
when large volumes of water (>10 5 m3 or >80 AF) need to be tagged, as is often the case near MAR
operations. As shown by the oceanographic community (e.g., [20]), SF6 can be used economically
to tag large volumes of water.
There are cost advantages of using SF6 over noble gas isotopes in terms of analysis; more SF6
samples can be analyzed over a given period of time on less expensive equipment. However, it is a
strong greenhouse gas and its emission is regulated in California. SF6 is a synthetic gas used primarily
in the electrical industry as a gas insulator and has been used as a tracer in the atmosphere and natural
waters for more than two decades. It is an ideal tracer for the following reasons: (1) SF6 is nontoxic [24]
and permission has been granted to use it as a tracer in potable aquifers in the southwestern USA
and South Australia; (2) background concentrations of SF6 in natural waters are extremely low
(<0.05 pmol/L; 1 pmol = 10í12 mole) because of its low solubility and low atmospheric mixing ratio
(ca. 8 pptv in 2014); (3) it can be measured precisely in water (±5% or better) over a concentration
range of eight orders of magnitude (0.01 fmol/L to 1 nmol/L) using a gas chromatograph equipped
with an electron capture detector [25]; and (4) laboratory scale experiments have shown that
breakthrough curves of SF6 and bromide are identical in saturated porous media which contains either
high contents of organic material or clays, demonstrating that SF6 is not retarded (has a retardation
factor of 1.0) [26,27].
SF6 differs from ionic and dye tracers in that it is a gas and is easily lost from solution across the
air—water interface to the atmosphere. It has often been used as a tracer for reaeration in surface
water [21–23,28]. Thus, it is important to monitor the SF6 concentrations in the recharge water to
determine the amount of tracer lost due to gas exchange. This is especially true if the artificial
recharge is taking place in a shallow river or spreading pond. Insufficient monitoring of the spatial
variability of the tracer concentrations in the “spiked” surface water can lead to erroneous
interpretations of travel times. Furthermore, laboratory column experiments have shown that SF6
transport is slowed (retarded) when trapped air is contained within the porous media [29,30]. While
there has been some evidence of gas tracer loss during recharge when a significant vadose zone is
present, experiments conducted at three sites in California with a long history of fairly continuous
recharge (OCWD, Orange County; Montebello Forebay, LA County; and El Rio, Ventura County)
have shown that gas loss during percolation from spreading ponds is manageable if the tracer is
introduced properly [1,13,16,18].
Fundamental questions concerning deliberate tracer experiments at MAR sites include: How
general are the results? Can the results of an experiment performed several years in the past be used
216
in the future? These questions arise because groundwater flow and travel times, e.g., to particular
wells, should reflect hydrologic conditions at the time of the experiment such as recharge rates both
at the MAR operation and surrounding area, pumping rates of nearby wells, and the regional
hydraulic gradient. All of these are likely to change over time scales of months to years and imply
that deliberate tracer results may not be similar year to year. For instance, long-term changes
in the demographics and economics of an area could lead to changes in the locations and
extraction/recharge rates from facilities. Furthermore, in areas with a Mediterranean climate such as
Southern California, it is likely that recharge and groundwater production are out of phase, with the
wet season having more recharge and less production. This should create seasonal variations in
groundwater flow that should affect travel times, especially at shorter time scales (i.e., months).
In this paper, we discuss the results of two deliberate tracer experiments conducted at a spreading
basin in Southern California that is operated by the Orange County Water District (OCWD),
ten years apart to assess if travel times changed significantly over a decade.
1.2. Field Site
A 9-km section of the Santa Ana River (SAR) and a series of spreading basins, including Kraemer
Basin and Anaheim Lake, located near Anaheim, California USA are used by the OCWD as principal
recharge locations for the Orange County groundwater basin (Figures 1 and 2). For more than 75 years,
OCWD has been actively replenishing and managing the groundwater basin that supplies about 70%
of the total water demand of approximately 2.4 million people and has been actively managing and
replenishing the local groundwater basin. Currently, it operates more than 400 hectares of surface
spreading facilities and recharges approximately 3.5 × 108 m3 (2.8 × 105 AF; 10-year average
between 2000 and 2010) per year [31].
In January 2008, OCWD completed and began operating the Groundwater Replenishment System
(GWRS), which is the world’s largest wastewater purification system for indirect potable reuse [33].
The system produces up to 2.43 × 105 m3 (197 AF) of high quality reclaimed water each day using a
three-step advanced treatment process that consists of microfiltration, reverse osmosis, and
ultraviolet light disinfection with hydrogen peroxide advanced oxidation. A variable portion of the
reclaimed water is pumped about 20 km to the MAR facilities and is recharged through Kraemer and
other permitted nearby basins to replenish the groundwater.
Detailed investigations of the groundwater flow were conducted in the late 1990s using a variety
of techniques including T/3He dating and deliberate tracer experiments at Anaheim Lake, Kraemer
Basin, and the SAR [1,34]. The January 08 SF6 experiment was required by state regulators because
of the use of Kraemer Basin for GWRS water recharge and to determine if travel times near Kraemer
Basin were similar to those determined a decade earlier during the October 1998 Xe isotope
tracer experiment.
217
Figure 1. Maps of the Orange County managed aquifer recharge (MAR) facilities.
(A) Topographic map of the field area including the location of Orange County Water
District (OCWD) recharge basins based on data from the United States Geological
Survey (USGS) 10-m National Elevation Dataset [32]; (B) Recharge occurs from the
basins and from the Santa Ana River (SAR) between the Imperial Dam and Ball Road.
The first arrival times of tracers determined in 1998 are contoured (modified from [1]).
(A)
(B)
2. Materials and Methods
The methodology of Clark et al. [1] was used during the January 2008 Kraemer Basin deliberate
tracer experiment and is outlined below. At the time of the tracer injection, the basin contained
approximately 1.4 × 10 m5 (2000 AF) of water. The volume increased during the experiment and by
the end, the pond contained about 2.1 × 10 m5 (3000 AF). The study began on 17 January 2008
(defined as day 0) when 99.8% pure SF6 gas tracer was carefully injected into Kraemer Basin over a
period of about 1 hour by bubbling the tracer through a submerged diffusion stone at a rate of
about 40 mL/min at two locations ~10 m offshore (Figure 2C). This injection technique was repeated
twice more, on day 8 and 11. During each injection, SF6 formed a rising bubble plume that only
partially dissolved in the water column (bubbles could be seen bursting at the surface); therefore, an
unknown quantity of tracer was released into the recharge water. Previously, it was estimated that
less than 5% of the bubbled gas dissolves [14,18,19,34].
Because it is vital to know the initial dissolved tracer concentration, SF6 surveys of Kraemer Basin
water were conducted on days 1, 3, 5, 8, 9, 11, 12, 14, 18, and 22. During each survey, surface
(~1 m below the pond’s surface) and bottom (~1 m above sediment) samples were collected from
five stations evenly distributed throughout the basin and marked with fixed buoys (Figure 2A,C).
Approximately 2–3 mL samples were collected in Vacutainers™ in triplicate for storage and later
analysis. Vacutainers™ are convenient storage and reaction containers that are commercially
available. Surveys were conducted until the mean SF6 concentration decreased to approximately the
detection limit. Following the tracer injection, well samples were collected for a period of one year
by personnel from OCWD in Vacutainers™ and sent to UCSB for analyses, which generally
218
occurred within two weeks of collection. The frequency of sampling was adjusted for each well based
on the results of recent sampling.
Figure 2. Maps of the Orange County MAR facilities. (A) Photograph of the basin
looking from the northern shore to the south towards the boat ramp. The inserted photo
shows one of the fixed buoys; (B) Detailed map of the January 2008 field area showing
the sampled wells and principle recharge areas. At the start of the experiment, neither
Miller nor La Jolla Basins were recharging the aquifer. For reference, the north-south
57 freeway and the east-west 91 highway have been included. The black dashed arrows
represent the northern and southern flow lines, the solid purple and dashed/dotted green
contours represent, respectively, the and November 1998 and June 2008 piezometric
surfaces; (C) Detail of Kraemer Basin at the time of the 2008 tracer injection showing
where tracer was introduced and sampled.
GWRS Discharge
(A)
(B)
(C)
219
All SF6 samples were analyzed using a headspace method similar to that described by [1]. In the
field, pre-weighed 10 mL Vacutainers™ were partially filled (2–5 mL of water). In the laboratory,
these containers were weighed (to determine the sample weight) and carefully filled with ultra-high
purity nitrogen gas (so that the final pressure was equal to about 1 atmosphere). After a brief shaking
to mix the headspace, this gas was injected through a column of Mg(ClO4)2 (to remove water vapor)
into a small sample loop of known volume (about 1.5 mL). Subsequently, the gas in the sample loop
was flushed into a gas chromatograph equipped with an electron capture detector with ultra-high
purity nitrogen carrier gas. SF6 was separated from other gases with a molecular sieve 5a column
held at room temperature. The detector response was determined by running gas standards purchased
from and certified by Scott-Marrin, Inc. (Riverside, CA, USA). The detection limit of this method is
about 0.05 pmol/L, three orders of magnitude lower than the mean pond concentration (see below).
Error on duplicate Vacutainer™ measurements was typically better than ±10% but not as good as
the ±5% reported for the syringe headspace method [25].
At each well, arrival times of tracer can be determined by evaluating breakthrough curves, which
are plots of concentration versus time. In homogenous aquifers—sand boxes, the initial and mean
arrival (center of mass or COM) times of tracer at narrow-screened monitoring wells represent,
respectively, the fastest and mean flow paths in the aquifer. In heterogeneous aquifers with
preferential flow paths tracer breakthrough curves are more complicated, often showing multiple
peaks [1,14]. Tailing is also evident on breakthrough curves and can represent the tracer reaching
the well by slower flow paths, back diffusion out of lower permeability strata, non-ideal tracer
input function, or retardation, which in the case of gas tracers can be due to trapped gas [29,30].
As discussed by Becker et al. [35], sampling biases can result in moderate and hard to quantify travel
time errors. These biases are often caused by infrequent sampling due to budget limitation.
3. Results and Discussion
Mean basin concentrations of SF6 tracer determined for each survey ranged between about
0.4 pmol/L (day 22) and 120 pmol/L (day 9); the daily infiltration rate varied between 1.7 m3/s
(day 20) and 2.2 m3/s (days 6 and 7) and averaged 1.95 ± 0.15 m3/s (68.9 ± 5.3 cfs). The basin
concentrations were the highest following the injections and decreased exponentially due to recharge
and gas loss across the air-water interface (Figure 3). The rate of loss was slightly greater during the
first week than during the second, most likely due to the progressive deepening of the basin as it was
filling with recharge water. This would increase the mean residence time of water in the basin and
decrease the gas exchange loss.
220
Figure 3. SF6 concentrations in Kraemer Basin during the injections and subsequent
monitoring periods. Injections occurred on days 0, 8, and 11; estimated basin
concentrations are plotted for these days. The dashed line represents the 14-day flow
weighted mean concentration (66 pmol/L) during the defined injection period.
The tracer injection period was defined as the first 14 days (between January 17 and 31) when
96% of the total mass injected percolated into the ground and the flow-weighted mean SF6
concentration was 66 pmol/L. It is important to note that the basin concentration decreased below
10% of the mean for 2 days (days 6 and 7) in the middle of the injection period (Figure 3). This
complexity in the basin concentration is apparent in the breakthrough at monitoring well KBS-3/1,
which displays two peaks (Figure 4B).
Although the water level contours represent different seasons and therefore different water use
times in terms of the domestic landscape irrigation cycle, they are sub-parallel indicating that the
location of recharge and pumping remained similar between experiments (Figure 2B). However, the
hydraulic gradient west-southwest of Kraemer Basin was ~50% steeper in November 1998 (~0.0064)
than June 2008 (~0.0042). Furthermore, the absolute elevation of the piezometric surface was about
5 m higher in June. These changes most likely reflect seasonality rather than long-term use patterns.
The June survey occurred near the beginning of heavy domestic landscape irrigation while the
November survey reflects water levels at the end of this irrigation period.
Groundwater samples were collected at 28 monitoring points (at 17 well sites, three of which were
multi-level: AMD-10, AMD-11, AMD-12; Figure 2B), with sufficient frequency to construct tracer
breakthrough curves (Figure 4). Tracer was detected at twelve wells down gradient (to the west) of
Kraemer Basin (Table 1). As was observed during the October 1998 Xe isotope tracer study [1,33],
detections progress systematically to the west along two flow paths that follow the local hydraulic
gradient: the northern path of KBS-3/1, AM-7, AMD-12/1, AM-48, and AM-8; and the southern path
of KBS-1/1, KB-1, AMD-11/1, AM-10, AM-9, and AM-14 (Figure 2B). The one exception is the
tracer’s first arrival was essentially the same at the deeper screened AMD-11/1 and AM-10 even
though AMD-11/1 was more distant. This implies that depth as well as distance is important when
considering travel times and that the first arrival is a difficult parameter to interpret. Interestingly,
the arrival of the COM followed the distance trend with it arriving at AM-10 prior to AMD-11/1.
221
Figure 4. Breakthrough curves at representative wells. Wells found along, respectively,
the southern (A, C, E) and northern (B, D, F) flow paths are displayed on the left and right.
The mean pond concentration was 66 pmol/L and non-detections (<0.05 pmol/L) have
been plotted as “0 pmol/L”. Please note that the scale on the y-axes (concentration) differs.
(A)
(B)
(C)
(D)
(E)
(F)
222
Table 1. Summary of travel times during the Oct-98 136Xe and Jan-08 SF6 tracer
experiments from Kraemer Basin. The Oct-98 Xe isotope data is from [33]. The distance
was measured from the basin’s shoreline directly to the well. This is the shortest distance
and does not necessarily represent the flow path length. Travel times are in weeks.
Well
Distance
(m)
Screen
Interval (m msl)*
Northern Flow Path
KBS-3/1
<100
44 to 41
AM-7
130
1 to í4
AMD-12/1
525
í36 to í42
AM-48
1250
í20 to í29
AM-8
1250
í26 to í31
Southern Flow Path
KBS-1/1
<100
4 to 1
KB1
<100
13 to 7
AM-10
1000
í3 to í8
AMD-11/1
1260
í91 to í97
AM-9
1840
í26 to í31
AM-14
2630
í32 to í38
Oct-98 136Xe
Jan-08 SF6
First
Detect
Peak
COM
First
Detect
Peak
COM
2.7
10.6
16.6
<16.6
20.9
4.7
22.9
22.9
18.7
—
5.6
25.4
30.5
25.7
37.1 †
<1.4
8.3
—
—
17.7
<1.4
15.1
—
—
23.0
†
15.3
—
—
38.3
<1.6
1.6
6.6
6.7
15.7
37.7
<1.6
3.6
20.9
29.1
26.6
37.7
†
6.0
23.8
>28*
33.8
†
1.4
2.6
15.1
—
26.4
45.8
1.4
2.3
23.0
—
39.7
—
—
3.7
28.4
—
50.6
67.9
Notes: * Depth is measured relative to mean sea level (msl). The ground surface elevation of Kraemer
Basin is about 68 m and the basin floor has an elevation of about 53 m. During the tracer injection period,
the surface of the water increased from about 57 m to 59 m; † Incomplete breakthrough, center of mass
(COM) travel time is a minimum or could not be calculated.
At relatively shallow wells (<65 m to the screen top) near the basin (KB-1, KBS-1, KBS-3), tracer
was first detected <1 to 3 weeks after the beginning of the injection period. This was not the case at
another nearby well, AMD-10, where the tracer was never detected in the five zones sampled, which
have screen tops between 90 and 285 m below ground surface. This is in agreement with the October
1998 experiment [1,33], which showed that these zones are hydraulically connected with Anaheim
Lake (Figures 1 and 2B), a recharge basin up gradient (to the east) of Kraemer Basin, once again
indicating the importance of depth to the flow field.
The influence of depth is best explained by the complicated local hydrostratigraphy. The water
recharged from the basins is not flowing through a sand box, rather it flows through conductive layer
between confining and semi-confining layers [1,33,36]. This complicated hydrostratigraphy also
helps to explain the reversal of arrival time (first vs. COM) at AM-10 and AMD-11/1 and points to
a potential problem of using a tracer with a large signal to noise ratio. There is no doubt that the first
arrival time is poorly defined: the arrival is defined by the first detection of the tracer and therefore
is defined by which tracer is being used and how good the analytical system is. Therefore the signal
to noise ratio of the tracer is vital for determining this arrival time and not the local hydrology.
It may make sense in the future to define this arrival time with C/C0, where C0 could be either the
initial concentration in the recharge water or the local peak concentration. We believe that the former
223
makes more sense because in is likely that the local peak concentration cannot capture unless
continuous sampling is employed.
The travel times of the leading edge of the tracer arrival, which is defined by the first detection,
were very similar (within three weeks) at six of the nine wells monitored during both the 2008
and 1998 experiments (Figure 5A; Table 1). The mean observed velocity of the leading edge was
about 60 m/week (3 km/year). The exceptions were the three most distant wells along the southern
flow path, AM-9, AM-10, and AM-14. During the 2008 experiment, the tracer arrival was traveling
about 50% faster and, therefore, earlier detections were observed at these wells.
Figure 5. Comparison between the arrival times and distance from the January 2008 and
October 1998 tracer experiments. (A) First arrival and (B) COM. Note the change of
scale on the x-axis. During the January 2008 experiment, COM arrivals could not be
calculated at all wells because of incomplete breakthroughs (see Table 1).
(A)
(B)
Monitoring wells AM-8, AM-48, AM-48A, and AM-49 are located along the northern flow path
approximately the same distance down gradient. However, the screen depth of these wells differs.
Tracer was not detected at the relatively shallow wells, AM-48A and AM-49 (screen depths between
20 and 30 m msl) while it was detected at the deep wells, AM-8 and AM-48 (screen depths between
í20 and í31 m msl). This implies that the tracer had migrated vertically downward beneath the water
table and was traveling through deep layers, with the shallow layers recharged by other sources, such
as the nearby La Jolla Basin (Figure 2B). The movement of the COM was also very similar during
two experiments (Figure 5B) and traveled with a velocity of about 40 m/week (2 km/year).
4. Conclusions
The results of the two deliberate tracer experiments conducted a decade apart were very similar
despite differences in the local piezometric surface. The arrival times at nine of the twelve wells
224
were nearly identical once the experimental uncertainty (i.e., frequency of sampling) is considered.
Apparently the hydraulic gradients near Kraemer Basin had not changed significantly between
experiments. This may be partly due to the fact that both experiments were conducted during similar
times of high recharge at Kraemer Basin and reduced seasonal pumping in the winter-spring study
period. As seen near other MAR facilities, vertical flow is important and must be considered when
evaluating travel time information.
Acknowledgments
We would like to thank Alex Cervantes, David Field, Brian Okey, Nira Yamachika,
Gary Yoshiba, and many others at Orange County Water District (OCWD) for hosting this research
and providing assistance over many years. Daniel Petersen assisted with the field and laboratory
work at Universoty of California Santa Barbara. Tim Becker helped draft some of the figures. The
comments and suggestions of the two anonymous reviewers improved the manuscript considerably.
Funding was provided by OCWD.
Author Contributions
All authors contributed equally to this paper.
Conflicts of Interest
The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
Clark, J.F.; Hudson, G.B.; Davisson, M.L.; Woodside, G.; Herndon, R. Geochemical imaging
of flow near an artificial recharge facility, Orange County, CA. Ground Water 2004, 42,
167–174.
Bouwer, H. Artificial recharge of groundwater: Hydrogeology and engineering. Hydrogeol. J.
2002, 10, 121–142.
Dillon, P. Future management of aquifer recharge. Hydrogeol. J. 2005, 13, 313–316.
National Research Council. Prospects for Managed Underground Storage of Recoverable Water;
National Academy Press: Washington, DC, USA, 2007.
Yates, M.V.; Yates, S.R. A comparison of geostatistical methods for estimating virus
inactivation rates in ground water. Water Res. 1987, 21, 1119–1125.
Fox, P.K.; Narayanaswamy, A.; Genz, A.; Drewes, J.E. Water quality transformations during
soil aquifer treatment at the Mesa Northwest Water Reclamation Plant, USA. Water Sci.
Technol. 2001, 43, 343–350.
Drewes, J.E.; Heberer, T.; Rauch, T.; Reddersen, K. Fate of pharmaceuticals during
groundwater recharge. Groundw. Monit. Remediat. 2002, 23, 64–72.
Hiscock, K.M.; Grischeck, T. Attenuation of groundwater pollution by bank filtration. J. Hydrol.
2002, 266, 139–144.
225
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Fox, P.K.; Makam, R. Surface area and travel time relationships in aquifer treatment systems.
Water Environ. Res. 2009, 81, 2337–2343.
Johnson, T.A. Groundwater recharge using recycled municipal wastewater in Los Angeles
County and the California Department of Public Health’s draft regulations on aquifer retention
time. Ground Water 2009, 47, 496–499.
Schlosser, P.; Stute, M.; Dörr, H.; Sonntag, C.; Münnich, K.O. Tritium/3He dating of shallow
groundwater. Earth Planet. Sci. Lett. 1988, 89, 353–362.
Solomon, D.K.; Cook, P.G. 3H and 3He. In Environmental Tracers in Subsurface Hydrology;
Cook, P.G., Herczeg, A.L., Eds.; Kluwer Academic Publishers: Boston, MA, USA, 2000;
pp. 397–424.
Massman, G.; Sültenfuss, J.; Knappe, A.; Dünnbier, U.; Taute, T.; Pekdeger, A. Investigation
of groundwater residence times during bank filtration in Berlin: A multi-tracer approach.
Hydrol. Process. 2008, 22, 788–801.
McDermott, J.A.; Avisar, D.; Johnson, T.A.; Clark, J.F. Groundwater travel times near
spreading ponds: Inferences from geochemical and physical approaches. J. Hydrol. Eng. ASCE
2008, 13, 1021–1028.
Schultz, T.R.; Randall, J.H.; Wilson, L.G.; Davis, S.N. Tracing sewage effluent recharge—
Tucson, Arizona. Ground Water 1976, 14, 463–470.
Clark, J.F.; Smethie, W.M., Jr.; Simpson, H.J. Chlorofluorocarbons in the Hudson estuary
during summer months. Water Resour. Res. 1995, 31, 2553–2560.
Busenberg, E.; Plummer, L.N. Dating young groundwater with sulfur hexafluoride: Natural and
anthropogenic sources of sulfur hexafluoride. Water Resour. Res. 2000, 36, 3011–3030.
Gamlin, J.D.; Clark, J.F.; Woodside, G.; Herndon, R. Large-scale tracing of ground water with
sulfur hexafluoride. J. Environ. Eng. ASCE 2001, 127, 171–174.
Clark, J.F.; Hudson, G.B.; Avisar, D. Gas transport below artificial recharge ponds: Insights
from dissolved noble gases and a dual gas (SF6 and 3He) tracer experiment. Environ. Sci.
Technol. 2005, 39, 3939–3945.
Ledwell, J.R.; Watson, A.J.; Law, C.S. Evidence for slow mixing across the pycnocline from an
open-ocean tracer-release experiment. Nature 1993, 364, 701–703.
Wanninkhof, R.; Ledwell, J.R.; Broecker, W.S. Gas exchange-wind speed relationship
measured with sulfur hexafluoride on a lake. Science 1985, 227, 1224–1226.
Clark, J.F.; Wanninkhof, R.; Schlosser, P.; Simpson, H.J. Gas-exchange rates in the tidal
Hudson River using a dual tracer technique. Tellus Ser. B Chem. Phys. Meteorol. 1994, 46,
274–285.
Hibbs, D.E.; Parkhill, K.L.; Gulliver, J.S. Sulfur hexafluoride gas tracer studies in streams.
J. Environ. Eng. ASCE 1998, 124, 752–760.
Lester, D.; Greenberg, L.A. The toxicity of sulfur hexafluoride. Arch. Ind. Hyd. Occup. Med.
1950, 2, 348–349.
Wanninkhof, R.; Ledwell, J.R.; Watson, A.J. Analysis of sulfur hexafluoride in seawater.
J. Geophys. Res. 1991, 96, 8733–8740.
226
26. Wilson, R.D.; Mackay, D.M. The use of sulfur hexafluoride as a conservative tracer in saturated
sandy media. Ground Water 1993, 31, 719–724.
27. Wilson, R.D.; Mackay, D.M. SF6 as a conservative tracer in saturated media with high
intragranular porosity or high organic carbon content. Ground Water 1996, 34, 241–249.
28. Wanninkhof, R.; Mulholland, P.J.; Elwood, J.W. Gas exchange rates for a first-order stream
determined with deliberate and natural tracers. Water Resour. Res. 1990, 26, 1621–1630.
29. Fry, V.A.; Istok, J.D.; Semprini, L.; O’Reilly, K.T.; Buscheck, T.E. Retardation of dissolved
oxygen due to a trapped gas phase in porous media. Ground Water 1995, 33, 391–398.
30. Vulava, V.M.; Perry, E.B.; Romanek, C.S.; Seaman, J.C. Dissolved gases as partitioning tracers
for determination of hydrogeological parameters. Environ. Sci. Technol. 2002, 36, 254–262.
31. 2009-10 Report on Groundwater Recharge in Orange County Groundwater Basin. Available
online: http://www.ocwd.com/Portals/0/Pdf/09-10AnnualRechargeReport_All.pdf (accessed on
9 June 2014).
32. Gesch, D.B. The National Elevation Dataset. In Digital Elevation Model Technologies and
Applications: The DEM Users Manual, 2nd ed.; Maune, D., Ed.; American Society for
Photogrammetry and Remote Sensing: Bethesda, MD, USA, 2007; pp. 99–118.
33. Groundwater Replenishment System. Available online: www.gwrsystem.com (accessed on
5 April 2011).
34. Davisson, M.L.; Hudson, G.B.; Clark, J.F.; Woodside, G.; Herndon, R. Final report on isotope
tracer investigations in the forebay of the Orange County groundwater basin. Lawrence
Livermore Nation Laboratory UCRL-TR-201735, 2004. Available online: https://e-reportsext.llnl.gov/pdf/303507.pdf (accessed on 14 April 2008).
35. Becker, T.E.; Clark, J.F.; Johnson, T.A. Heat 10B-enriched boric acid, and bromide as recycled
groundwater tracers for Managed Aquifer Recharge: A case study. J. Hydrol. Eng. ASCE,
submitted.
36. Thomson, A.F.B.; Carle, S.F.; Rosenberg, N.D.; Maxwell, R.M. Analysis of groundwater
migration from artificial recharge in a large urban aquifer: A simulation perspective.
Water Resour. Res. 1999, 35, 2981–2998.
MDPI AG
Klybeckstrasse 64
4057 Basel, Switzerland
Tel. +41 61 683 77 34
Fax +41 61 302 89 18
http://www.mdpi.com/
Water Editorial Office
E-mail: water@mdpi.com
http://www.mdpi.com/journal/water
MDPI • Basel • Beijing • Wuhan
ISBN 978-3-906980-92-8
www.mdpi.com