The Gulf War

I. DEFENDING THE SHORE

The boat captain—an elderly man in coveralls who had spent most of his life on the water—was travelling along one of the main passes through the Louisiana marshland. He had steered his vessel, a rectangular flat-bottomed outboard, from a dock in St. Bernard Parish, navigating among the roseau cane and black mangroves. The sky was a pale-blue dome suspended over motionless white clouds. It was a June morning, and under a hot sun the bayou was permeated with quiet unease. Other vessels on the water were working to contain oil from the Deepwater Horizon well site, which, on April 20th, had erupted forty miles offshore––an explosion that left eleven men dead and threatened to render the coast an ecological wreck for years. The captain directed his boat past shrimp trawlers going out to sea, presumably to skim the oil in waters beyond the coastline. Three- and four-man teams in other small boats were laying down brightly colored boom: floating barriers designed to impede oil slicks. The boom looked like giant yellow Magic Markers, forming long chains across the labyrinthine bayous.

“I can see open water—just after that booming boat,” the captain called out. He looked out past a barge with piles of equipment stacked upon it. “No way to get around the boom,” he said. (Fearing that he might lose his contract with BP, which operated the Deepwater Horizon, he insisted that he not be named.) Next to him, Marty Cramer, a thin, middle-aged Californian with weathered skin, squinted and nodded. Cramer, a specialist in emergency response for oil spills, had flown to Louisiana to help hunt for the deepwater crude. He was working with a company called Polaris Applied Sciences, which had been hired by BP to map the oil as it came ashore. He was joined by Carol Oz, a scientist from the California Department of Fish and Game, and Vincent Chatelain, a volunteer from the Louisiana Department of Environmental Quality. The team was one of six assessing marshes and beaches, using a method called Shoreline Clean-Up Assessment Technique, or SCAT.

Cramer had been surveying in Louisiana for almost three weeks. By then, heavily oiled marshes had become the staple subject of television crews and photographers. Their images, typically in tight focus, showed suffocating swirls of shimmery crude and sickly pelicans. The scenes were riveting and heartbreaking, but they fundamentally misrepresented the situation. For many responders, the amount of oil in the marshland was a relief. Cramer had worked on the Exxon Valdez spill, twenty-two years ago, and the difference between the oil in Alaska’s Prince William Sound and in the Gulf of Mexico was stark. In Alaska, a liquid façade of thick, waxy North Slope crude had coated nearly two hundred miles of rocky coast. The oil at the Gulf Coast did not remotely compare; by early summer, the crude gushing from the BP well had caused only twenty-five miles of “heavy oiling”—a SCAT term for coverage of more than fifty per cent—on the entire sixteen-hundred-mile Gulf coastline. One had to travel, sometimes an hour or more, to see the oil—one had to hunt for it.

Cramer studied the boom. “Well, maybe they can open it up for us,” he said.

“I don’t know how the hell we’re going to get out,” the captain said.

But by half past ten he had found a way through, and headed toward Deadman Island, a patch of tall grass in the water. Cramer recorded the location, and studied the grass with binoculars.

“Any oil-soaked pelicans out there?” Chatelain asked.

“Not that I can see,” Cramer said.

Oz peered at the island and reported that she could see laughing gulls and brown pelicans. “The birds seem to be nesting,” she said.

Deadman Island was surrounded by a band of yellow boom, but Cramer spotted some grass that looked oiled. “Go as close as you can,” he said, and the captain steered the boat toward some boom that was partially submerged.

“Can you jump it?” Cramer asked.

The captain glared at him, swung the boat around, and rammed its hull over the yellow barrier.

“I did not do this,” he said.

As the boat headed for the strip of brownish vegetation, Cramer looked at the grass with his binoculars. “I don’t think that’s oil,” he said. When the boat hit the shore, Oz leaned over and pulled several strands of the grass out of the water. Their stems were brown and lumpy, but the discoloration had nothing to do with pollution. “Yeah,” she said. “That’s natural growth.”

It has become conventional wisdom that the BP-funded response to the spill was a chaotic and mismanaged affair, driven by corporate avarice, lacking in urgency, and at times willfully negligent of the problem’s scope—the idea being that any organization that had caused such a catastrophe, and that was so clearly unprepared for it, could not in good faith clean up the scene of the disaster. The evidence for this is much like the imagery of heavy oiling: vivid and convincing upon first consideration, but also fragmentary, anecdotal. At the peak of the cleanup effort, forty-seven thousand people were fighting the oil, a community equivalent in size to Annapolis, or the workforce of G.M.—as one federal scientist called it, “a company built in the middle of the night.” In just half a year, the response expended nearly sixty million man-hours, roughly nine times what it took to build the Empire State Building. After the well ruptured, BP accepted help from competing oil companies, and hired the world’s leading oil-pollution specialists to run key operations. The logistical demands on the effort, which spanned the entire Gulf coast—a region of varied geography and political culture—were immense. President Obama was not exaggerating when he announced in June, “This is the largest response to an environmental disaster of this kind in the history of our country.”

BP hired the designer of SCAT, Ed Owens, a British geologist, to implement the surveys. Owens, a founder of Polaris Applied Sciences, is a broad-shouldered man in his mid-sixties who lives outside Seattle. Like many people in the world of spill response, he speaks about spills as if each were a military deployment. “I was in Desert Storm,” he told me, referring to the American military involvement in Kuwait, when Saddam Hussein released as much as nine million barrels of crude into the Persian Gulf—a defensive maneuver that caused the world’s largest oil spill. “We were in the middle of a war, and I am flying along the shoreline, videotaping, with a battery of Patriot missiles locked onto my helicopter.” He had come up with SCAT in 1989, after an oil barge collided with a tug off Washington State and released fifty-five hundred barrels of fuel, contaminating ninety-five miles of shoreline. Owens devised standard terminology for the various levels of pollution, and created surveys that allowed government responders and oil companies to trust the same data. Before that, Cramer told me, “people would look and say, ‘There’s a bunch of oil,’ but there wasn’t a real systematic process.”

In Louisiana, members of the SCAT teams regarded themselves as intelligence officers for the cleanup. Nearly every evening, they returned to a BP training facility outside Houma, forty-five miles southwest of New Orleans, one of the central hubs in the response. They crammed into a room where Owens listened to their findings. The information was analyzed by a night shift of technicians, and plotted on satellite maps so that the oil’s movements could be tracked.

On the boat, Cramer told the captain, “Oil was reported at 29.48, 89.17,” and they set off. The coördinates marked Gardner Island, a strip of marshland bisected by two rock jetties—part of a vast network of channels created by the oil industry to service rigs. Nearby in the water, stained boom was adrift and twisting in on itself like a befouled eel. The captain pulled near the shore, and the team jumped onto a beach covered in broken shells. A penumbra of trash—a cut-up tire, shards of plastic, water bottles—had formed along the waterline. But Cramer and his team were looking for oil, and at Gardner Island they finally found some: a narrow row of droplets on the beach.

“This is very light, less than one-per-cent coverage,” Cramer said, and recorded the finding. Then he walked a bit farther. “We’ve got a dead bird here,” he called out. “It looks like it is oily.” Everyone huddled around the bird. It was a pelican, its carcass stained with muck. Federal wildlife officials would have to pick up the remains, establish the cause of death, and catalogue the information. Looking at the carcass, Cramer said, “They get oil on their feathers, can’t maintain their insulation, and die of hypothermia. This is considered evidence now.”

While Oz called the U.S. Fish and Wildlife Service on her cell phone, Cramer waded through a small inlet and continued surveying. He was walking slowly across the shells under the hot sun, his head turned toward the ground, but there was nothing else to record.

II. FIRE AND WATER

The old saying has it that oil and water don’t mix, but every day the world’s oceans absorb colossal amounts of oil. When hydrocarbons flow into the sea—whether from spills, or leaky ships, or natural seeps—experts call them “petroleum input.” The world’s total petroleum input is thought to be about three hundred and eighty million gallons per year—a quantity similar to the catastrophic Gulf War spill—with a fifth of it happening in American waters. Much of the input off the United States comes from natural seeps. Some of the largest of those are in the Gulf of Mexico, which is thought to absorb more than fifty million gallons of oil annually.

Approximately twenty thousand oil spills are reported in America every year. Most of them are small and do not attract much attention; only a tiny fraction cost more than a million dollars to clean up. An economy based on oil must be prepared to deal with large amounts of pollution, and over many decades this country has evolved a way to respond to spills. “There is no plan,” one politician took to saying as the response progressed last summer. But there was a plan. Its origins dated back to the first major industrial oil spill at sea: the collision of a tanker called the Torrey Canyon against Pollard Rock, off the coast of England, in 1967.

When the Torrey Canyon ran aground, its broken hull released thirty-seven million gallons of Kuwaiti crude into the water. Oil poured forth in heavy slabs: one drifted toward France; another coated two hundred miles of shoreline in western Cornwall. Twenty-five thousand birds died, and local communities and the British government fought to contain the mess. People on beaches tried in vain to soak up the oil with straw, or they used bulldozers and pumps to recover the oozing petroleum. From the other side of the Channel, the French government dumped three thousand tons of chalk containing stearic acid into the oil, hoping it would sink or disperse. Eventually, the Royal Navy bombed the tanker with a mixture of napalm, sodium chlorate, and aviation fuel, in an effort to incinerate the oil. This, too, was largely ineffective.

The American government watched the incident with alarm, and the following year Congress created the first National Contingency Plan—a blueprint for dealing with a similar catastrophe. A few years later, the Coast Guard set up three oil-spill strike teams in different parts of the country. But when the Exxon Valdez ran aground, in Prince William Sound, in 1989, this evolving system of spill response was put to a tremendous test, and in many ways it failed. Though the Exxon Valdez spill is only the world’s fifty-seventh largest, it was ecologically devastating. The rocky, remote Alaska shoreline was difficult to clean, and the subarctic weather made it impossible to work in winter. On a number of occasions, the response’s methods, such as the use of high-powered jets to blast crude off rocks and beaches, did more damage to the environment than the oil did—but public outrage often demanded action, even if scientists advised against it. Eleven thousand people gathered in Prince William Sound to assist in the effort, and they fell into arguments over basic decisions. Vice-Admiral Clyde Robbins, who led the federal spill response, struggled to get Exxon and government authorities to set aside their mutual distrust and collaborate. “It made it difficult to move ahead on anything,” he told me. “I didn’t really have authority.”

The problems that the Coast Guard faced in Alaska were not entirely about the oil. They were also about emergency response and public perception. “All oil spills are emotional events,” Ann Hayward-Walker, a responder who had worked on the Exxon Valdez incident, told me one evening in Houma. It is possible to fight a forest fire and not be distracted by how the calamity was caused, and whether the cause taints the integrity of the people who deal with it. But oil spills are saturated in blame and political confusion—and opportunity. There is a sense that they are not accidents but accidents waiting to happen, and thus acts of greed. As a result, oil-soaked birds and fish come to symbolize a reviled industry’s heedless behavior. Every year, as many as four hundred thousand birds are killed in America by electricity-generating wind turbines, but they do not make the cover of Time. Incremental ecological damage, even if it is severe, does not easily cause outrage.

Mistrust of Exxon, along with ambiguity over who was in charge, brought about the Oil Pollution Act, in 1990, which gave the Coast Guard full control in a spill. Rather than create an organizational structure from scratch, the Coast Guard borrowed one from forestry. In 1970, firestorms in California ravaged more than half a million acres, and at one point more than nineteen thousand people from five hundred agencies were trying to put out the flames. Equipment and manpower were marshalled in haphazard ways, and amid the confusion the fires grew worse. The Forest Service, along with other agencies, spent four years working out a solution to such problems, called the Incident Command System, or I.C.S. At its most basic level, I.C.S. divides labor into four groups: Command, Planning, Operations, and Logistics. Each team can grow rapidly as more people arrive. If the system is working well, a responder from New York can walk into an incident in Texas, be assigned to Logistics, and know what to do. “There is an elegance about it,” Ed Owens told me. The 9/11 Commission advised that I.C.S. be used in all domestic catastrophes, and in 2003 President George W. Bush made the recommendation law.

Shortly after the Deepwater Horizon exploded, the Coast Guard and BP, following the National Contingency Plan and I.C.S., came together in a swiftly expanding structure known as the Unified Command. Joining with BP was politically volatile. The company had a reputation for aggressive drilling and poor safety, and had been responsible for some of the greatest oil-industry disasters in recent history. In 2005, an explosion at the company’s Texas City Refinery killed fifteen people and injured a hundred and seventy; BP pleaded guilty to criminal charges. The following year, one of its pipelines ruptured in Prudhoe Bay, Alaska, causing the largest spill in the North Slope. The company’s management of the Deepwater Horizon appeared to have been similarly reckless, and its response plan for the rig clearly inadequate. Like the rest of the industry, BP barely invested in technology or manpower devoted to oil pollution. In recent cost-cutting, its emergency-response division had been reduced from thirty-two people to eight, with only one employee, David Fritz, dedicated to oil-spill response full time. “I didn’t have programs,” Fritz told me. “The program was me.”

A joint headquarters called the Unified Area Command was set up in Robert, Louisiana, and it oversaw five Incident Command Posts, covering Texas, Louisiana, Alabama, Mississippi, and Florida. The command posts in turn oversaw dozens of staging areas by the shore. At each echelon in the hierarchy, the government was legally in charge, but in practice federal officials collaborated with BP—a distinction that was often confusing to outsiders. At the Unified Area Command, Rear Admiral Mary Landry, the top federal officer, had legal authority over BP, but she told me that on a day-to-day basis she rarely needed to exercise it. As the response outgrew what BP was obligated by law to support, the company nonetheless gave the Coast Guard nearly everything it asked for, and experience dictated that collaboration usually led to the best outcome. BP was paying for the response, and could mobilize resources and expertise in a way that the government could not. “They could have said, ‘We’re not doing this,’ ” Landry told me. “ ‘We’re done. We’ll see you in the courts.’ ”

The Obama Administration quickly learned that talking about this relationship publicly was difficult. On April 29th, Janet Napolitano, the Secretary of Homeland Security, held a White House press briefing to explain the escalating national response. A Coast Guard officer, Rear Admiral Sally Brice-O’Hara, standing alongside her, was asked if the federal government intended to exclude BP from the cleanup and simply send the company a bill.

“We are certainly not at that point now,” she said. “And I don’t imagine, given the professionalism of our partner, BP, and—maybe ‘partner’ was—let me back up—”

“They are not our partner,” Napolitano interrupted. “They are not our partner.”

“Bad choice of words,” Brice-O’Hara hastily explained.

Soon after, Napolitano announced that Admiral Thad Allen, who was then the commandant of the Coast Guard, would be taking charge of the response on a national level. “We could never get past the perception—not the law, the perception—that somehow BP was making decisions independently, and that those decisions weren’t in the best interest of the response,” Allen told me. “We’re not talking about reality. Reality had no place in portions of this response. The perception that there wasn’t government oversight—clear down to where boom was going in the water—resulted in us redirecting resources to reinforce the fact, whether it was needed or not, that there was government oversight.”

III. HOUMA

The Houma Incident Command Post was situated amid wild grass near an interstate, in a large BP corporate-training center covered with white tile and glass. Across the highway was a gas station with a mini-mart, the only other structure in the area. Responders called the training center Houma, for the nearby city of the same name, as if the facility were a metropolis all to itself—which, in certain respects, it was. On the night of the blowout, the building was largely empty. But it was designed for emergency response, with generators and satellite communications. In the first days of the spill, local authorities parked trailers in front of it, and a BP manager scrambled to acquire everything from hotel rooms, for the expected influx of people, to office supplies and sophisticated printers that could generate maps of the oil. “We went to a local engineering firm and borrowed a plotter from them, until we could buy three or four more,” the manager told me. The community grew rapidly.

From Houma, responders fought the oil along the Louisiana shore and out near the source of the spill. At the operation’s peak, more than a thousand responders worked there, many of them nearly around the clock. Temporary offices were built alongside the facility. The parking lot was expanded until it reached beyond the highway, and vehicles had to be hired to shuttle people from their cars. One vehicle, a bus designed for bachelor parties, was equipped for pole dancing, and people sat awkwardly in its dimmed lights as they were driven to work. Men and women in blue Coast Guard uniforms came. Engineers and geologists in jeans and faded button-downs came. There were scientists, lawyers, members of the National Guard, a sheriff who sat out front to make sure there would be no trouble, and salesmen—some of them so eager that they made their pitches in the parking lot.

For most of April and May, the commander at Houma was a Coast Guard captain named Edwin Stanton, a Louisianan whose ancestors can be traced back to the first French expedition to the Gulf, more than three hundred years ago. “I have hunted and fished in the coastal waters, and have more than a passing familiarity with the lay of the land down here,” he told me. Stanton, a stout man with a shaved head and a welcoming, rounded countenance, has worked on oil spills for thirty years. He believes in a big show of force; for a spill in Puerto Rico, he once spent a hundred million dollars in three months. Stanton told me that he was following the Kime Directive, named for Admiral John Kime, a Coast Guard commandant who helped revamp the service’s approach to spills after the Exxon Valdez ran aground. Kime believed that overwhelming resources should be brought to a crisis—excess matériel could always be withdrawn—and Stanton tried to implement this philosophy at Houma. At the same time, he often spoke about the oil in a blunt and pragmatic manner, which some locals found upsetting. “Marshes before beaches!” he told responders: the marsh ecosystem was far more vulnerable, even if the beaches were cherished vacation spots.

When Stanton arrived, the response was taking shape; everyone worked in one room, and people wore color-coded vests designating their I.C.S. roles—red for Operations, blue for Planning, yellow for Logistics. The responders clustered in tribes around folding tables. Booming and skimming operations had begun, and airplanes were spraying chemical dispersants onto the oil to help it dissolve into the sea. At first, BP responders at Houma were making crucial decisions by videoconference with the company’s executives in Houston. Stanton insisted that this was crippling Houma’s effectiveness, and BP gave its employees greater autonomy.

Stanton had known many of the responders for years, and, shortly after he settled into his command, Ed Levine, a thin, soft-spoken man with glasses and a neatly trimmed beard, came to see him. Levine is an oceanographer who for twenty-three years has worked on spills for the National Oceanic and Atmospheric Administration, from an old Coast Guard building at the tip of Manhattan. NOAA has been cleaning up hazardous pollution with the Coast Guard since the nineteen-seventies, and Levine had come to serve as Houma’s chief federal scientist. “Our mantra was: What got spilled? Where is it going? What’s in its path?” he told me.

While the Deepwater Horizon was still afloat, there was no real oil spill to speak of. Crude and natural gas rushed up into the rig, and were incinerated in fires that had been consuming the structure. When the rig sank to the ocean floor, it created clouds of debris, making it difficult to tell how much oil was being released. “It took probably thirty-six hours to get good imagery, because so much sediment and silt was raised when the thing crashed,” Admiral Allen told me. After the sediment had cleared, days of bad weather further complicated underwater surveys of the wellhead area. Around this time, BP estimated that a thousand barrels per day were flowing out of the well—a provisional number that, NOAA announced, would be verified when the weather allowed. But the estimate, obviously inadequate, drew heavy criticism.

Before it was clear that the spill could not be readily contained, BP was reluctant to allow outside specialists to see the oil flow. “At first, a Webcam showing the wellhead was restricted viewing,” Levine told me. “Those images were very closely held, but I finally convinced BP to allow me to make a ten-second video, which I sent to Seattle,” where NOAA’s headquarters for oil-pollution response is situated. “The people in Seattle said, ‘We can’t guess how much is coming out of there.’ ”

Only under congressional pressure did BP publicly release high-definition video of the wellhead, and this secrecy reinforced the impression that the company was trying to obscure the problem. “The pipe wasn’t coming from a place above ground where you could see it, and I think the public felt that something was being hidden,” Levine said. “Even I felt it a little bit when BP wouldn’t let us show people what we were seeing.” Anderson Cooper, whose CNN news show, “AC360,” became a populist forum for Gulf locals to vent their frustrations, said one night on the air, “I mean, how do you know how to put out a fire unless you know how big the fire is?”

But, for the responders, a larger number would have made no immediate difference. Early in the spill, an internal Coast Guard communication noted that the blown-out well might release as much as a hundred and ten thousand barrels of oil into the Gulf per day. “The very first discussion we had was: Here is that worst-case scenario, and obviously that’s like an Exxon Valdez every couple of days,” Stanton told me. “But then the estimate came back from BP that it was a thousand barrels per day, and, you know, a thousand barrels per day is bad enough. I was going to be ordering in everything I could get. I ordered in every single federal response asset east and west of the Rocky Mountains.” Stanton brought in boom from Norway, and hired skimmers from Holland that could work at night and in rough weather. Levine told me, “We said, ‘We need everything,’ and it’s not like we’re going to ask for ten times everything. Everything is everything.”

Near the wellhead, crude was gathering in large, slick pools of red and brown fluid. One morning, Levine approached Stanton. “How come we’re not burning this?” he asked.

“You’re right!” Stanton said, and mockingly hit himself in the head. He had used burning to combat smaller oil spills in the marshes of the Mississippi Delta. But Levine had something different in mind: burning the oil offshore, in rough and open water, before it hit land. This had never been done before on a large scale. Stanton told BP’s top commander at Houma about the idea, and the next day the world’s leading expert in controlled burns, Alan Allen, came to the facility. Allen began his career on a spill off the coast of Santa Barbara, in 1969—the largest offshore blowout in American waters prior to the Deepwater Horizon—and several years later he had invented the technique. Special boom that could withstand temperatures hotter than two thousand degrees would be required. Neré Mabile, a BP engineer, told me, “In a matter of forty-eight hours, we got fire boom here. We got the right people together, and we went out and did the first burn—I think a hundred barrels.”

On April 28th, Levine left the command post to examine the test burn with David Fritz, the BP oil-spill adviser. It was the first time the two men had flown over the oil, and as they approached the wellhead they saw ribbons of crude extending to the horizon. Levine had once worked on a spill of about a thousand barrels in the Delaware River, and, as he looked down on the oil from the airplane window, he roughly calculated that the daily volume emanating from the BP well was ten times that amount.

“There’s a whole lot more than a thousand barrels coming out of there,” he told Fritz. “The estimates are very low.”

“Very low,” Fritz said.

They flew above a plane that was spraying dispersants, and over skimming vessels, and shrimp boats that were dragging boom in U-shaped arcs, gathering oil to be burned. “I do remember thinking how small those skimmers all looked,” Fritz recalled. “They had to go so slow, because that’s just the way the physics of skimming is. You’ve got this ocean full of oil and these tiny little skimmers that don’t appear to be moving, and it looked so fruitless.”

On the flight back to Houma, Fritz told Levine, “How humbling—to think that we are going to make any real dent in it.” They did not speak much for the remainder of the flight. Once they were back on land, Levine called Stanton. “This is going to be monumental,” he told him. “With all of the skimmers in the world out there, you might as well be using thimbles.”

In the first week of the spill, the vast ribbons of crude by the wellhead were not yet making landfall, but they were nonetheless daunting; even the oil’s chemistry was still a mystery. Although it was generally thought that the Deepwater Horizon had been drilling “light, sweet” crude, BP did not quickly provide the response with a specimen. “The first samples of oil came out of a skimmer,” Levine told me—thick blobs with the consistency of cold cream. Without any other evidence to assess, NOAA sent samples to Edward Overton, a professor emeritus of environmental sciences at Louisiana State University. He determined that they contained unusually high amounts of asphaltenes, heavy chemicals found in pavement. “My level of apprehension went from moderate to the red zone,” Overton told me. “It’s not going to be easy to degrade. It’s not going to be easy to burn. It’s not going to be easy to disperse.”

In the press, oil spills are typically judged by the amount of oil released, but volume can be a misleading standard. Wind patterns, ocean hydrodynamics, the chemistry of the oil, the temperature of the water—all these factors are significant. Within NOAA, the stories of two tankers have come to exemplify the way that chemistry and the environment can affect the outcome of a spill: the case of the Amoco Cadiz, and that of the Braer. In the spring of 1978, the Amoco Cadiz, carrying sixty million gallons of light crude, ran aground off the coast of France. The oil coated two hundred miles of coastline, in some places nearly two feet thick. People used boom, skimmers, and even “honey wagons,” vacuums designed to suck up liquid manure. “But this thing just overwhelmed anything that humans could do,” John Robinson, the founder of NOAA’s oil-pollution-response program, said. In contrast, the Braer broke apart off the coast of Scotland in 1993, and released twenty-five million gallons of Norwegian light crude—a spill more than twice the size of the Exxon Valdez—but the oil naturally dispersed very rapidly. “A couple of hundred metres from the ship, you took a sample and it didn’t look like oil; it looked like coffee,” a NOAA scientist told me. “If you stuck your hand in the water, and pulled it out, it didn’t feel slick or greasy.” Not long after the spill, barely any visible trace of it remained.

For Levine, the question of whether the BP oil would act like the crude from the Amoco Cadiz or like that from the Braer was essential to the shape of the cleanup. About a week after the first samples arrived, BP provided better specimens: the oil was exceptionally light. (The early samples were most likely residue from oil that had been burned.) Moreover, forty per cent of what was spewing out of the undersea wellhead was not oil but methane. “Once you started to get close to the site, you would find that you were in a very large patch—ten miles across—of thick oil that was perhaps a millimetre or two thick,” a responder told me. “It had a consistency to it not very different from water. When we first started, the image that we had in mind was the Exxon Valdez. Once we actually got there, you began to realize just how nonpersistent the oil was—only a small fraction persisted beyond twenty miles from the source.”

IV. THE GENERAL

While operations took shape at Houma and at other command posts across the Gulf, officials in Louisiana began to mobilize their own response, mostly in parallel to the federal effort. On April 28th, the day Levine and Fritz flew over the oil, the official flow rate was increased to five thousand barrels, a very rough estimate made by NOAA scientists. (That estimate, it turned out, was also too low.) Governor Bobby Jindal, of Louisiana, declared a state of emergency, as did a number of local officials. The most prominent of them was Billy Nungesser, the president of Plaquemines Parish—in many respects, the region most vulnerable to oiling. To fight the oil, Nungesser had put together an initiative called the Inland Waterways Strike Force. He started a command center, and had an artist draw up a crest: a jumping marlin, two boats, a helicopter, and some boom, and beneath them three Latin words: protego nostrum facundia—roughly, “protect our resources.”

Like no other official during the spill, Nungesser embodied the rage, anxiety, and frustration that swept through South Louisiana. Physically, he resembles the atom bomb that was detonated over Nagasaki: his body, shaped like two parabolic cones stuck together at each base, is expansive at the midsection and then roundly tapers at either end. Ever since the spill began, he had adopted a Pattonesque attitude toward his parish—a demeanor that was either mocked by his constituents as political theatrics in a well-worn Louisiana style or admired for its vehement authenticity. One morning, he held a breakfast meeting at Lil’ G’s Kajun Restaurant, in Belle Chasse, and local oystermen got in his way to shake his hand. “That’s our general,” one of them said. “He’s got a set of nuts. Jindal is a scholar, but Billy—we are all his warriors.” When BP’s C.E.O., Tony Hayward, expressed doubt that oil was entering the marshland, Nungesser responded, “Let’s take him out there and dunk him in the water, and when he comes out black with oil, let’s ask him the same question.”

Plaquemines Parish is a place of diminishing geological fortunes. Unlike the land beneath Houston, or Boston, or Chicago—tectonic formations that developed over millions of years—its land is only a few thousand years old, and natural forces, among them hurricanes and rising seas and coastal erosion, regularly put it in peril. After hurricanes Katrina, Gustav, and Ike devastated the parish, it was easy to regard the BP oil spill as just another catastrophe bearing down on the state. And, as one Louisianan told me, “Kicking the hell out of the federal government is good sport down here.”

Shortly after the Deepwater Horizon sank, Nungesser’s director for coastal management, P. J. Hahn, went into the water near the wellhead with a team of divers. “From horizon to horizon, all you could see was oil, and we actually dove through it,” Hahn told me. “One of the divers got sick immediately. It coats you. My eyes got stuck, and I had to consciously open them to blink.” The oil, Hahn recalled, took the form of liquid clouds, dynamic and irregular, in some instances spreading downward thirty feet. He told Nungesser what he had seen, and Nungesser told me, “It scared the hell out of me.”

As the response grew, a Coast Guard captain took Nungesser and several other politicians on a helicopter flight over the Louisiana coastline. There were darkened spots on the water, which the politicians feared were oil, but they were only shadows created by cloud cover. “We are going to fight this offshore,” the captain said, over the din of the engines; Nungesser, he recalled, looked upon his threatened parish and grew quiet.

Later, Nungesser turned the trip into a parable—one that may have drifted from the truth but that nonetheless expressed a real difference between the Coast Guard and the Gulf communities in how each regarded the oil. “He said, ‘Don’t worry, Mr. President, we’re going to help you clean up,’ ” Nungesser said. “And I turned to him and said, ‘Land this chopper. There is no cleanup in South Louisiana. You either prevent it from getting in—or we’re dead.’ ”

On May 2nd, President Obama flew to Louisiana, and in Venice, a small dockside community in Plaquemines Parish, he met with Admiral Allen, Stanton, and several local politicians. By then, the most sophisticated aspects of Houma’s work—the controlled burns and the application of dispersants—were under way, but these measures occurred forty miles offshore, largely out of public view. What was visible was a national organization that was still figuring out how to structure itself. Boom that Houma had ordered quickly accumulated at large staging areas, and more than two hundred thousand feet of it was deployed by responders in small boats. Still, politicians across the Gulf were scrambling for boom of their own, and some complained to the media that they weren’t getting enough of it. Allen told me that “the amount of boom sent to various states was almost a political litmus test on federal support.” Inside the Unified Command, this phenomenon became known as the Boom Wars.

When the President arrived in Venice, Nungesser was among the people he met with. “He introduced himself,” Nungesser told me. “He said, ‘What’s going on? What do we got to get done?’ ”

Nungesser explained that he wanted BP to pay for three jack-up boats—barges that can be cranked up on tall pylons and suspended in the air—so that members of his strike force could clean up oil and deploy boom around the clock. “The President said to Allen, ‘Commander, what’s wrong with the plan?’ ” Nungesser told me. Allen looked to a local Coast Guard officer for an explanation. “We deploy from Venice every day and we don’t think it’s necessary,” the officer said, adding that the boats were expensive, and would have no real operational impact. But Nungesser reiterated his case, and Obama told Allen, “I want to get Billy’s jack-up boats out there.” The boats were hired the next day, and Obama called Nungesser to insure that it had happened.

In Venice, Stanton spoke about what might happen once the oil made landfall. “It’s going to be very ugly,” he predicted.

“You’re a pretty gloomy guy,” Obama said.

“Yes, sir,” he said. “They call me Captain Eeyore.”

Stanton explained some of the measures that Houma intended to take: dispersants, skimming, and the use of controlled burns in the marsh—an idea that Jindal dismissed as crazy. By then, the Governor had sidelined the state agency responsible for oil spills and begun to manage Louisiana’s response with his closest staffers. Frustrated with the Coast Guard’s allocation of resources, and convinced that the pace at Houma was “lackadaisical,” he teamed up with the parishes and created his own booming plan. Like Nungesser, Jindal was determined not to let oil enter the Mississippi Delta. His plan for Louisiana called for five million feet of containment boom, and an aide told Stanton that the state wanted it within a week. “The number was just astonishing to me,” Stanton recalled. “When I got back from that meeting, I remember very clearly telling BP, ‘You have got to provide this boom.’ I said, ‘You need to begin ramping up your shoreline strategy. Whatever you think you need, I need you to multiply it times three.’ It got to the point where we had exhausted every single boom source in the continental United States.”

Eventually, three hundred thousand feet of boom were being made for the response every week, and more was shipped in from fourteen countries. “We had to work at these end-of-the-road communities where there is not enough dirt to hold millions of feet of boom,” Jim Black, a BP manager who served as Houma’s operations chief, told me. “They didn’t have infrastructure—hotels in the range of a few hundred rooms. And yet it is also their community. We are the interlopers responding to our emergency in their back yards.”

Still, complaints about the distribution of boom persisted. Garret Graves, Louisiana’s head of coastal restoration, told me that Jindal deployed National Guardsmen to track boom from one staging area to another. “We never could figure out where it went. We don’t know if it went to the Bermuda Triangle, or what,” he said. “There was this mystery five miles of high-seas boom floating around. A month later, we found it was still in a staging area.”

In late May, the state decided to demonstrate some of these issues to Stanton by flying him over areas that needed boom, and staging areas where it sat, apparently unused. “We said, ‘Let’s get him out here,’ ” Graves told me. “He came back, and blood vessels were popping out of his neck and head. He was yelling on the phone.” Stanton by then had grown frustrated with the Unified Command, an officer told me; he wanted it to think bigger. Shortly after he arrived at Houma, he had asked for six thousand National Guardsmen, to be deployed throughout the coast, but was denied. (The guardsmen had just returned from overseas, and Allen had agreed not to deploy them unless it was absolutely necessary. Eventually, eleven hundred soldiers assisted the response in Louisiana.)

Following his flight with the Governor, Stanton held a press conference at a staging area. “I’m gonna go right back to tell BP to hire more boats, hire more people, get more boom, and put it out,” he said, and conceded, “We did have a problem of getting boom down here to begin with. It’s my job to direct this response in Louisiana, absolutely.”

A reporter asked why more boom had not already been brought in.

Stanton looked exhausted. By then, the response had deployed more equipment than had been used for the Exxon Valdez. “Well, the ‘why’—is that really important?” he said.

“Yes, sir,” the reporter said.

“All right,” Stanton said. “Well, I guess I am just slow and dumb.”

The next day, Stanton was relieved as Incident Commander. “The training for Coast Guard public affairs has to start including political acumen,” a public-affairs officer assigned to the response from Washington told me. But many responders were sympathetic. “Ed Stanton is a very practical man,” a Coast Guard officer told me. “But how do you deal with a very impractical community?”

Stanton was replaced by a friend and protégé, Roger Laferriere, the captain of the Port of Los Angeles. Laferriere, who is forty-eight years old, and has brown eyes and close-cropped hair, is one of the Coast Guard’s leading experts on oil-pollution response. When he arrived in Louisiana, he went straight to Houma. “I said to Ed Levine, ‘I need to go out there and look at the source,’ ” he told me. “It was surreal. The oil that was out there went for miles and miles.”

Laferriere spoke urgently about the oil, sometimes referring to it as a sentient enemy, which he called Grendel. He regarded spill response as a kind of military campaign, with hostile popular sentiment a constant concern. “When you win any war, you need to integrate the locals,” he told me. “So one of the things that I set out to do is figure out where that disconnect is between us and the locals, and see how we could mend it.” He brought to Houma a staff of officers to manage operations while he met with politicians and community members, and carried a “hot folder” containing urgent requests from people he met. He took Ed Owens and Ed Levine to town-hall meetings, hoping that they could speak convincingly about the environmental science. “We were conducting a major educational campaign in the middle of an operation,” he told me.

“I think Billy Nungesser really wanted to see an army of folks out there cleaning up the oil as soon as possible,” he added. “Our No. 1 goal isn’t necessarily rapid response. It is doing what is in the best interests of the environment. We would say, ‘O.K., Billy, when this oil comes to shore, we need to take a look at it.’ What’s there? Are there birds there? Are there fish? And sometimes, believe it or not, leaving it alone is the best option.” After the Amoco Cadiz spill, some marshes were so damaged by the cleanup that, a decade later, as much as forty per cent of the vegetation had not recovered. Still, Laferriere said, “The people of Louisiana are extremely passionate about their marsh, and they viewed the oil almost as nuclear waste: ‘We want it picked up and removed now!’ ”

Laferriere learned that the local strike teams were laying hundreds of thousands of feet of boom, in addition to the work being done from Houma. In some cases, they were attempting to lay down boom in double layers; in others, their boom trapped oil in the marsh, rather than keeping it out. Levine told me, “You could throw out marshmallows and it would be as effective.” So Laferriere campaigned to explain that there was a science to boom deployment—if done wrong, it could cause more harm than good, by crushing and killing grass, hastening erosion, or trampling oil into root systems. He also emphasized the need to hold boom in reserve. “I kept on telling people, ‘Remember your history in World War II? Remember the Maginot Line in France?’ ” he told me. “That failed miserably. The oil was not going to hit the coastline in a single massive wave. It comes as large ribbons, what we called ‘streamers,’ and, if we can get them before they come ashore, that’s what we wanted to do.”

V. WALLS OF SAND

By the end of May, the oil was surfacing above the wellhead in countless pieces, which moved unpredictably through the Gulf. “We were dealing with hundreds of thousands of patches of oil,” Admiral Allen told me. “We didn’t have a large, monolithic spill.” One long patch slid around Plaquemines Parish, but then suddenly retracted without really hitting the shore. Responders at Houma called it the Devil’s Tail. “The oil was very dynamic,” Laferriere recalled. “When you flew out to the source, what you saw was plumes—it was almost like a ballet of oil plumes coming up to the surface.” But, outside Houma, it was hard to convey this. Graphic imagery, produced by NOAA or BP and circulated on the news, gave an impression of a single sheath of crude covering the sea. “Computer graphics really gave a false image of the problem,” Ed Owens told me. “You may have oil on the water’s surface that’s ten-per-cent coverage in a square mile, yet on a map it’s all colored in.” Oil likes to spread—a drop of it can expand until it is one molecule thick. Sheen is thinner than hair, and easily evaporates. The maps gave the impression that an armada of oceangoing skimmers could suck up the oil in one place. This was never possible.

In May, the oil finally made its way to the shore, and some of it made landfall at Pass a Loutre, a large opening in the marsh, in a protected wildlife area at the tip of Plaquemines Parish. The responders at Houma informed the Governor’s office, and a couple of days later Jindal flew over the site. For weeks, the Louisiana Department of Wildlife and Fisheries had been monitoring the pass, and fifty-five thousand feet of boom had been deployed in the area. The oil—black and viscous, and therefore anomalous for this spill—gathered into a patch of shoreline a mile or so in length. To both sides were miles of unblemished marshland, and the water extending into the Gulf, except for some sheen, was clear. Nearby, there was a long slice of marsh that had been carved out to serve as a canal for the oil industry. This kind of channelling, conducted for decades throughout South Louisiana with the state’s encouragement, in all likelihood has caused more severe and lasting damage to the marsh ecology than any oil spill.

On May 26th, Nungesser, Jindal, and James Carville, the political strategist (a Louisiana native), went to Pass a Loutre to draw media attention to the landfall. By then, oil had been at the pass for about three weeks. The SCAT teams were focussing their surveys on other heavily oiled areas, and on nearby rookeries, some of which were oiled while birds were nesting. Anderson Cooper joined the politicians, and his footage of the trip gave the impression of an ecological apocalypse. “In Pass a Loutre, the marshes are covered in crude,” he said. “The sight is sickening. So is the smell. There’s little sound, no sign of life.” Photographs of the trip, framed around the oil, showed the Governor trying to scoop up crude with a net. Nungesser spoke as if he were in a wasteland. “Usually, you would pull in here, the fish, the minnows—a trout would be running behind a minnow,” he said. “Action all over. It’s dead. Look, there’s not a bug. There’s not a squirrel. There’s not a fly. There’s nothing. And there is no one cleaning it up.”

Across the Gulf, the response had documented fewer than seven hundred animal deaths. But the Governor’s trip to Pass a Loutre quickly became an emblem of the spill, and images of it were repeated on CNN throughout the summer. The trip occurred two days before the President’s second visit to the Gulf, and appeared calculated to illustrate that the Unified Command was not doing enough. Inside Houma, some responders couldn’t understand the outrage over the oil at Pass a Loutre. One of them acknowledged, “We were probably a little too slow to respond to that area.” But weeks before the spill a Chevron pipeline had ruptured nearby, releasing at least seven hundred and fifty thousand gallons of crude into the marsh, and local politicians had never spoken about it.

By the time Nungesser went to Pass a Loutre, he had become fixed on a plan: the construction of a network of artificial sand dunes, or berms, to block the oil. Shortly before the BP well blew out, his office had been in touch with a Dutch company, Van Oord, that was seeking a contract to dredge sediment from the coastal seabed and pile it on barrier islands that were rapidly eroding. After the spill, the plan was reformulated as an oil-response measure. The proposal envisaged a sand barrier that would span a hundred and twenty miles, virtually encircling the Mississippi Delta. It promised to circumvent the need for a cleanup while, theoretically, restoring the coastal ecology. “We’re doing this to stop the oil,” Nungesser told an aide.

At the time, the Unified Command had been considering many nonstandard ideas; a special unit sifted through more than a hundred and twenty thousand of them, which arrived by letter, e-mail, and fax. Some, like detonating a nuclear bomb on the seabed, were quickly dismissed. Others—Kevin Costner’s centrifuge technology, for instance, or the A Whale, a Taiwanese tanker that was redesigned as a massive oil skimmer—received publicity but were impractical for this spill. The berm project was not part of this process, because the Governor proposed it directly to the federal government. The Interior Department reacted to the plan with skepticism, and NOAA’s scientists were concerned that the proposal might even “lead to increased environmental injury.” They wondered if the berms would alter currents, or harm the animals that they were intended to protect. Federal officials were apprehensive about oil mixing inextricably into the clouds of sediment that the dredging would cause.

After two weeks of deliberation—an unprecedentedly short period for such a project—the Army Corps of Engineers decided to grant approval for forty miles of berms, divided into six discrete segments. One federal scientist told me that everyone agreed that the six berms would at least do no ecological harm. The Army Corps estimated that, with the available dredges, it would take between fourteen and twenty-one months to construct the longest of them, and that the shortest berm would take at least three months. The over-all cost could be more than a billion dollars. Nungesser and Jindal expected BP to fund the project, whatever the amount. Allen approved the building of one berm, as a prototype, but refused to compel BP to pay for all six. Two other admirals called Nungesser to explain that they did not think the berms would be effective, and that forcing BP to pay for them would create rifts within the response which could distract from meaningful work. “You have given me excuse after excuse after excuse,” Nungesser responded. “I’m going to hang up now, because you are not giving me anything useful to save my parish.”

On May 28th, Obama travelled to Louisiana for a second visit, this time to Grand Isle, to meet with an assembly of politicians from the Gulf states. Nungesser, who had met with the President a few weeks before, was not invited. Undeterred, he chartered a boat and made his way past local officials who were patrolling the bayous. When he arrived, an official turned him away. “I said, ‘You call the White House and see if I can be put on the list, while I call Anderson Cooper,’ ” Nungesser recalled. “She went into the guardhouse, came back in five minutes, and said, ‘The White House said to let y’all in. Have a good day.’ ”

That week, Nungesser had announced that he might commandeer a flotilla of boats and construction equipment and risk arrest to begin work on the rest of the berms. He told ABC News, “There was no response early. Now, whether that’s BP, the Coast Guard—I don’t think the President knew what was going on.” Inside the Coast Guard station, he ran into Admiral Landry. She told him, “Could you tone it down? You are kicking every Coast Guard officer who is working out there in the stomach.” Members of the Coast Guard had worked themselves to exhaustion; she pointed out that there were responders who had been struck by lightning, and others who had suffered heatstroke.

At the conference table, Nungesser took a chair next to Obama’s designated seat, and Jindal sat next to him. Nungesser told me that when the President entered he said, “Billy, I thought we were on the same team.”

“Mr. President, we are,” Nungesser said. “Thad Allen is the one not doing his job.”

Obama sat down, and Nungesser erupted. “Mr. President, without these berms, we don’t have a chance,” he said. A Coast Guard official who was there told me that Jindal kept saying, “Make BP pay for the berms.”

“Well, a lot of people don’t think they are necessary,” Obama said, and explained that the government was legally obligated to determine that the berms were viable before it could compel BP to pay for them. “We need validation that it is appropriate, and won’t cause more damage,” he said.

Nungesser had brought with him a binder filled with facts and figures. “All the documents are in there—and that’s not something I dreamed up overnight,” he said.

Obama turned to Allen, and said, “Well, what’s the problem?”

Allen explained that scientists working for the Army Corps had estimated that berms could take more than a year to build, making them useless against the oil. Nungesser interrupted him. “This is bullshit!” he yelled. “That’s not true!” (Under the table, Jindal tugged Nungesser’s shirt.) Nungesser insisted that if the President mobilized every dredge in the world the project could be completed within two months. Obama said, “If the engineers and scientists believe it is the right thing to do, then we’ll do it. If they don’t, then we’re not. We are going to make our decisions based on facts, not hope.” He concluded, “I tell you what—Thad, I want to get a panel together, and get a roundtable discussion, and then I want to be briefed on the answer, and I want it done in the next few days.”

As the meeting ended, Obama told the politicians, “I have given you full access to the White House, and I have been pretty forthright. I don’t want to hear on the news that the federal government is incompetent, and then you get on the phone and tell me that everything is O.K.” A Coast Guard official who was there told me, “They all kind of shook their heads, ‘Yes, sir,’ ‘Yes, sir.’ And I was, like, Wow, the President just told them to knock it off.” Nungesser told me that Obama addressed him specifically about his appearances on CNN. “If you can’t get it done through the chain of command, you pick up the phone and call the White House,” the President told him. “If you can’t get me on the phone, then you can go call Anderson Cooper.” The two men shook hands.

The next evening, Nungesser joined Cooper on the air and told him what the President had said.

“He didn’t actually say that,” Cooper said.

“Yes, sir,” Nungesser said. “Yes, he did—”

“Wait,” Cooper said. “He said call—he said—that’s funny!”

But Nungesser said, “I truly believe that he cares.” He added, “We’re behind the cleanup now.”

Nungesser did not keep himself away from CNN for long, and soon his criticism returned in force. By June, he was telling people, “Thad Allen is a fucking idiot, to put it mildly.” But Allen was quietly bringing greater command and control to the massive effort. He had convened a group of specialists to measure the oil coming out of the wellhead. (They estimated that the rate was as high as sixty-two thousand barrels per day.) He tripled the number of Coast Guard officers deployed throughout the Gulf—a decision that put a strain on the service. He reorganized Vessels of Opportunity, a program that hired local boats to help with the response, which had grown to more than ten thousand craft—“everything from a canoe to a professionally licensed offshore vessel,” the officer who oversaw it told me. The response enlisted the National Geospatial-Intelligence Agency to track boom, using technology that the military employs for battle-damage assessments. Furthermore, Allen told the President that he wanted to take control over the Gulf airspace. “We had eight near-midair collisions,” Allen told me. “There were planes flying out for logistics, spotter planes, planes spraying dispersants. There were Air National Guard planes flying. NASA was flying planes.” He adopted the system used during the Haitian earthquake; the Air Force lent him the personnel, and began coördinating the hundreds of daily flights from Tyndall Air Force Base, in Florida.

On June 1st, Allen held the roundtable discussion on the question of the berms. “We were put in a pretty bad spot,” he told me. “We were having a meeting based on a proposal that was made to the President, regarding a project that was approaching four hundred million dollars, and we had to get this thing resolved in four or five days—the case that built up in the media was so extraordinary that it required some resolution. That’s not the best way to run operations.” At first, only scientists and technical experts had a chance to speak, and throughout the discussion Nungesser fidgeted. During a break, he recalled, he confronted Allen. “I said, ‘Admiral Allen, this is bullshit. This is a dog-and-pony show. I am calling the President.’ ” Then he stormed out and spoke with TV reporters in the hallway. “I’m sick to my stomach,” he said. “We ought to have some of the dead animals in there.”

When Nungesser came back, Allen said, “We’re going to let the parish presidents and the Governor speak five minutes each.” The local officials clapped, and when Jindal spoke Nungesser wept. The next day, Allen informed Jindal that BP had agreed to spend three hundred and sixty million dollars on the project. When an official at the Interior Department suspended the dredging for one berm until it could be conducted with fewer environmental risks, Nungesser threatened to mail her an oiled pelican carcass.

The fight over the berms obscured real improvements. Many of the logistical problems that had hampered the earliest phase of the response were being resolved. BP began to acquire matériel on a vast scale. Every week, responders were using roughly three million hazmat suits, far more than the world’s supply, so Logistics teams scoured the globe for stockpiles and alternatives to make up for the shortfall. When the manufacturer of the dispersants, Nalco, could not make the chemical fast enough, BP helped with its supply chain. The company spent money liberally. “Supply and demand kind of goes out the window,” one BP responder told me. “We understood and tolerated and even smiled when we paid ten, twenty, thirty per cent above market.” By midsummer, more than eight hundred skimmers had been mobilized by the shore and around the wellhead. “This was the largest offshore-skimming operation in the history of spill response,” Laferriere told me. The burning operations were removing millions of gallons of oil. About as much oil was incinerated in the Gulf as had spilled from the Exxon Valdez. Louisiana officials, meanwhile, opened barriers restraining the Mississippi River; the water’s outward force appears to have greatly prevented the oil from penetrating too deeply into certain bayous. By June, most of the thick crude had been removed from Pass a Loutre.

Still, the threat of the oil lingered. In late June, NOAA predicted gusts of wind surging from sea toward land, prompting dread that masses of crude would assault the Louisiana shoreline. Laferriere, frustrated by allegations that Houma was not acting with enough urgency, and wanting to keep the morale of his officers buoyed, sent an e-mail to all of his responders. “Now we will be fully tested,” he told them. “Despite our best efforts, oil may go beyond our defenses. We must not be discouraged, but fall back to those areas of impact, and attack the oil, quickly and aggressively, and ensure the oil is removed from these shores and marshes with uncommon quickness. We are not unlike the three hundred warriors that stood at the Hot Gates of the Battle of Thermopylae, under King Leonidas, who faced tens of thousands of Persians. They were outnumbered, but they didn’t give up. They fought till the end, as we will. They fought for a way of life, and we are fighting for a way of life.”

VI. THE CHEMICALS

Even as Laferriere tried to motivate his responders for an all-out assault upon the coastline, he recognized that the principal fight against the oil was offshore, to be conducted with a weapon—dispersants—that many people thought was more harmful than the spill itself. “How do you view the various technologies and their ability to fight oil?” he said. “There are really two components to that. One is: How much oil do they take out of the environment? How much oil can be skimmed or burned or dispersed? Then, there is another factor that is equally important: What is the ‘encounter rate’ of the technology? Remember, the oil on the water is about a millimetre thick. Its area is huge. So if you can only go about a knot, which is the average skimming capacity, and less than a knot when you are burning, it is not possible, physically, even with all the vessels in the world, to keep up with the spreading of the oil.” Skimming is virtually impossible with waves of more than six feet, and burning is difficult with three-foot waves. Dispersants, on the other hand, make use of chaotic seas; the ocean’s natural energy helps the oil break up and dissolve. “A dispersant plane can go a hundred and forty knots,” Laferriere continued. “It can cover a huge area. There was a huge slick off the Chandeleur Islands. It got away from our skimmers; it got away from the burning vessels. What’s the option? We could disperse it, or let it hit the islands.”

It was commonly reported that BP was making decisions—often in secret—about where to spray dispersants. Rolling Stone claimed that “BP decided to wage chemical warfare in the Gulf.” In fact, the Unified Command made the decisions jointly. “It’s a team effort,” Laferriere told me. “But I am also the police officer right there.” Scientists like Ed Levine helped plan the sorties, and Coast Guard officers monitored them. Charlie Huber, an independent contractor who managed the spraying, told me, “BP was not coming down and saying, ‘Spray here, and spray there.’ ”

There were echoes of the Boom Wars in the debate over dispersants. Alarmist criticism easily earned credibility. At a hearing about the spill, Representative Jerrold Nadler, from New York, argued that lives were being endangered. “There is no scientific evidence that dispersants can be effective in oil spills of this magnitude,” he claimed, inaccurately. “But these chemicals make it harder to track how much oil is there, and where it’s going, and thus to determine liability.” He said that the dispersants reminded him of Agent Orange, and expressed concern that responders were conducting “an uncontrolled experiment that could result in thousands and thousands of people getting sick or dying as a result of the cleanup”—he emphasized, jabbing a pointed finger—“not of the original disaster.”

Scientists and activists spoke of the dispersants as if they had been concocted in a weapons lab. “Goes right through skin,” Susan Shaw, a marine toxicologist, asserted on CNN, adding that the chemicals had caused one shrimper to start “bleeding from the rectum.” Some locals wondered if the chemicals had spawned a disease. “It is commonly being called the Blue Flu, because the alleged symptoms include blue lips and skin, and it’s scaring the hell out of people,” one blogger wrote. A television reporter in Alabama did a segment in which a chemist speculated, without giving any evidence, that dispersants in seawater by a marina may have caused a sample to unexpectedly combust. A common rumor held that BP had been spraying dispersants at night to sink the oil—even though there were no nighttime sorties. “Why don’t we stop spreading the dispersants, let it come to the surface, and let’s fight it where we can see it?” Nungesser told Anderson Cooper one night. “Let’s get every ship from around the country, around the world—put ’em out there.”

For the people who live in the Gulf, the fear of dispersants often blended with rage and frustration of a broader type: concerns about unpaid bills, restrictions on fishing, the demise of a way of life. Late in the spill, at a town-hall meeting on Grand Isle, a BP representative named Jason French sat behind a folding table at the front of a room full of locals. During the meeting, a thickset seafood supplier named Dean Blanchard stood and spoke about his hardships. “Y’all didn’t give me enough money to pay my bills,” he said, adding, “How do you expect a man to live on less than ten per cent of what he was projected to make?” After speaking for several minutes, Blanchard sat down. “Y’all not trying to pick up that oil!” he yelled. “Come on, man. It’s a joke.”

French attempted to allay everyone’s concerns. “Our task forces are looking for oil each day,” he said. “I understand the anger, and I understand the frustration.” He tapped the table nervously with his finger. “But someone’s gonna have to explain to me why BP doesn’t want to clean up this oil. This was not—”

“Because it’s more cost-effective for y’all to come at night and sink the son of a bitch!” Blanchard shouted. “Y’all send the planes, and y’all fucking sink it!” He sat back down. “What do you think, we’re stupid?” The room went silent, and his voice had grown hoarse. “Y’all are putting the oil on the bottom of our fishing grounds!” he screamed. “Y’all are messing me up for the rest of my life. I ain’t gonna live long enough to buy any more shrimp!”

The fear of using chemicals to clean up large marine oil spills is nearly as old as the spills themselves. When the Torrey Canyon ran aground, in 1967, the response team dumped an industrial detergent called BP1002 into the slick. British Petroleum, as the company was known at the time, had designed the chemical to degrease and clean its equipment, and the hope was that it would cause the spilled oil to break up into manageable amounts. At the time, no one knew how dangerous the detergent was, or even how to apply it; the chemical was dribbled from hoses and dumped in barrels from cliff tops. Seven hundred thousand gallons of BP1002 were used during the spill, and scientists later concluded that this had done more harm than good.

Shortly after the Torrey Canyon incident, chemists began to search for less toxic and more effective alternatives, and, in 1968, a scientist named Gerard Canevari, with Esso, created the original formula for Corexit, the dispersant that was used in the BP spill. Early experiments showed that half a gallon of it could break up fifty gallons of oil into tiny droplets; toxicity tests that the company had commissioned were encouraging. Canevari, who is now retired, spent most of his professional life refining Corexit’s various formulas. He told me that he based some of the dispersant’s ingredients on food-grade additives that had F.D.A. approval. “You have ingested some of the components of it,” he said. The formula used most widely in the Gulf, Corexit 9500, has seven main ingredients, each of which can be found in products that Americans handle regularly, including Klondike bars, Lubriderm, and ibuprofen. Sorbitan monooleate, for instance, is in juices and shampoos, and hydrotreated light-petroleum distillates can be found in air fresheners.

At one point during the spill, a blogger wrote on the San Francisco Chronicle Web site, “I was disturbed to get another anonymous tip that Corexit 9500 also has dihydrogen monoxide, but I can’t confirm this because Nalco will not reveal if dihydrogen monoxide is in fact a secret ingredient in Corexit 9500.” The blogger explained that the chemical was “really bad and nasty stuff,” used in explosives and poisonous compounds. “It mutates DNA, denatures proteins, disrupts cell membranes, and chemically alters critical neurotransmitters.” Dihydrogen monoxide—better known by its chemical symbol, H2O—is plain water.

Dispersants don’t remove oil from the sea, but they are designed to help nature do so. “We often lose faith when considering the power of Mother Nature to remove oil or pollutants from the environment,” Laferriere told me. Ocean-borne bacteria consume floating hydrocarbon molecules that are common in certain parts of the sea. When the oil is dispersed, its over-all surface area is increased, and this allows more bacteria to consume it at once, speeding up the process. (Imagine a cake the size of a house, and a hundred thousand people trying to wolf it down at once; then imagine that cake cut into slices and passed around to the same crowd.) After the microorganisms eat the oil, it is essentially purged from the ecosystem. It’s difficult to predict whether Corexit will advance this process in every circumstance. In some experiments, microorganisms appear to eat only portions of the oil, leaving the rest behind. In others, they consume all of it, or none of it. In the Gulf of Mexico, where there are so many natural oil seeps, and so many forms of bacteria that feed on them, there was every reason to think that the chemicals would work as intended.

The decision to use dispersants was made on April 21st, while the rig was still aflame. Right away, a specially equipped C-130 from Arizona was sent to Houma, and a smaller King Air BE-90 aircraft in Mississippi was put at the ready. The use of dispersants is standard practice in the U.S. The E.P.A. maintains a list of them that are approved for spraying in oil spills, and Corexit had already been used at least seven times in the Gulf. In 1979, a rig called Ixtoc-I blew out in Bahia de Campeche, in Mexico, causing a massive spill, and two and a half million gallons of Corexit were deployed to fight it—considerably more than was used in the BP spill.

On April 22nd, David Fritz told me, “a member of the strike team walked in at around two in the morning and asked what our dispersant plan was. We were, like, ‘Here are the approval forms,’ and he was, like, ‘No, I want to see your plans.’ So we cobbled something together.” That day, the first sortie released about eighteen hundred gallons of dispersant into the water. The forms made it clear that the chemicals could be used off the coast of Louisiana if they were sprayed more than three miles from the shoreline, but did not say how much could be used. In a matter of days, with no regulatory limit, the sorties were spraying as much as forty thousand gallons a day over the Gulf.

BP had an obvious interest in maintaining the spraying, and the responders at Houma believed that fears about the chemicals should be addressed by patient explanation. But the E.P.A.’s leadership believed that the response would be fundamentally unsuccessful if it did not accommodate itself more directly to public anxieties. The rifts threatened to cripple the Unified Command. “Just as I could not have BP walk out, I could not have the E.P.A. walk out,” Admiral Landry, the federal on-scene coördinator for the early part of the response, told me. These tensions were not expressed publicly, but they shaped the way the response added 1.8 million gallons of Corexit to the sea.

VII. HURRICANE BP

The administrator of the E.P.A., Lisa Jackson, is a chemical engineer with degrees from Tulane and Princeton. She grew up in New Orleans, in the Lower Ninth Ward. Her education at Tulane was funded by a Shell Oil Company scholarship, and in the summers she interned at the company. She is a proponent of environmental justice—the notion that minorities and the poor, who suffer from pollution disproportionately, should have a say in policy. “I believe that those who are often impacted by environmental problems have the least voice, the least political power, ability, or even knowledge, to get into the process, and one of my priorities at the E.P.A. is to expand the conversation on environmentalism,” she told me. During tours of the state, she listened to Louisianans affected by the spill, and sometimes even gave them her phone number.

Jackson believed that the response could not escape local history. It is hard to hear the word “Katrina” and not think of the Superdome in a desperate state, of people stranded, seemingly ignored, while the President told the head of the Federal Emergency Management Agency, “You’re doing a heckuva job.” Katrina destroyed Jackson’s mother’s home, and Jackson drove her mother and relatives to safety. Shortly after the Deepwater Horizon caught fire, Jackson flew to the Gulf, and from her plane she wrote a blog post, which she never published, titled “Hurricane BP.” She determined that the E.P.A. would demonstrate to Louisiana that the federal government was listening. Her decisions on dispersants, she told me, would have to be “driven in many ways by public concerns,” because “the issue, whether rightly or wrongly, was one that was very much in the front of many people’s minds down there.” During one meeting with Obama, she spoke almost apologetically about her position. “Mr. President, with regard to dispersants, I am probably going to be the skunk at the picnic,” she said. “Sir, I am going to be out there asking very hard questions.” Obama told her, “That’s exactly what I expect you to do.”

At first, Jackson knew little about dispersants, so she called a former professor at Tulane who had researched them, and he introduced her to other academics. “We all reach back to people who know a lot more than we do,” she told me. When spills occur in water, the primary responsibility for the cleanup falls to the Coast Guard, and the E.P.A. plays a secondary role—for instance, monitoring air pollution, overseeing waste management, and deciding which dispersants may be used. Early in the response, she told reporters that using Corexit was permissible. “If it’s on the list and they want to use it, then they are preauthorized, if you will, to do so,” she said. But, as the spill grew, and it became clear that dispersants were being used in a manner that the regulations had not anticipated, Jackson came under intense political pressure to act aggressively. Admiral Allen told me, “We went to a congressional briefing with over a hundred members of Congress, and I saw the anger and frustration over the use of dispersants, and they were being levelled straight at her.”

Much of the consternation about dispersants emerged early in the spill, when an ExxonMobil chemical engineer who had joined the response suggested that Corexit be injected into the oil directly over the wellhead, at the ocean floor. In the late nineteen-sixties, Gerard Canevari, the inventor of Corexit, had tested the chemical’s application underwater in the Gulf. “We dispersed the oil subsea, and we never saw it again,” he told me. In 1970, he said, subsea dispersants were used for a well blowout in the Mississippi Delta, but the procedure was never mentioned in a scientific journal, and very few people knew of it.

Quickly, responders within the Unified Command embraced the idea, even though its implications had barely been studied. The undersea injections could potentially be five times as efficient as spraying, and promised to improve safety. When the oil rushed to the surface, it spread and evaporated, releasing toxic vapors. On some days, these vapors became so dangerous that firefighting tugboats were brought in to use water cannons to push oil away from vessels working to kill the well. When this was ineffective, the vessels had to be evacuated. The vapors were “an overwhelming health hazard, and an explosive hazard,” Ed Overton, the NOAA consultant, told me. Dispersing the oil undersea was likely to cause less of it to come to the surface, reducing the danger.

Jackson was skeptical at first. “BP thought this was just a dandy idea,” she told me. “They were saying, ‘We should start the injection, and let the testing catch up with it,’ but my belief was that we needed to explain to the public all we had done to assure that this application was not doing more harm than good.”

The testing had to be done a mile below the Gulf ’s surface, in punishing environmental conditions. Steve Lehmann, a NOAA scientist, told me, “I think people forget—it is kind of like applying dispersants on Mars.” The Corexit was pumped to the wellhead from a ship at the surface. The oil was erupting from the well with great force, and at first the engineers diverted the dispersants through a wand positioned beside the flow. The hope was that the Bernoulli effect—the physical process that forces an airplane wing upward during flight—would pull the dispersants into the oil. “This was highlighted at one meeting I went to as a failure, because they were projecting the dispersant outside of the plume,” an ExxonMobil chemist recalled. “But it sure looked like it was all going into the oil.” With additional work, BP suspended the wand directly into the flow.

As the Corexit entered the plume, BP officials pressed Jackson to approve the injection for continuous use. “They were waving around pictures from one of their tests: some, ostensibly, from when they weren’t injecting subsea, and they showed a huge slick, and then others from two or three or six hours later, and there was literally no oil,” Jackson said. “I never felt particularly compelled to believe those pictures.” Jackson’s primary concern was whether oil-eating microorganisms would surge in population, triggering in the water a condition of severe oxygen depletion known as hypoxia, which is potentially deadly for many creatures. Working with the E.P.A., the Unified Command commissioned vessels to gather data around the wellhead, and they discovered dispersed plumes of oil roughly a mile deep but no “large-scale impacts” on oxygen levels. The oil was being consumed, but not at a rate that appeared to damage the undersea ecosystem. On May 15th, Jackson approved the injections, believing that the spraying would diminish as a result, and announced that there would be “maybe eighty-per-cent reduction in the amount of dispersant used.”

By then, Houma was commanding a fleet of dispersant planes, and as they flew out across the Gulf they left behind them trails of tiny droplets, which one pilot described as smelling like baby shampoo. When the chemical hits fresh oil, its effect is almost instantaneous: the dark-brown crude becomes diffuse and much lighter in color; because of this, responders compare newly dispersed oil to café au lait. Within seconds, particles of oil begin to spread downward, into the water column. When oil floats on the water’s surface, wind and currents can cause it to gather into long ribbons; when it is dispersed in three dimensions, it expands like smoke from a smokestack and never coalesces again.

The responders at Houma were not involved with the undersea injections, which were managed by the Houston Incident Command Post. They rarely spoke with Jackson—Admiral Allen tried to shield them from the demands of Washington. Their attention was devoted to the daily onslaught of oil at the water’s surface. Jim Black, Houma’s operations chief, told me, “The political aspect of it—I sort of told people, ‘I’m just a soldier down here, and I’ll fight the battle with whatever tools you give me.’ ”

Jackson expected the spraying to decrease after her order on May 15th, and by coincidence this is what happened. Flying conditions that day were dangerous, so Levine took the opportunity to meet with people who had been newly hired as spotters to look for dispersible oil slicks. Not every kind of oil is suitable for dispersants; if oil is too weathered, Corexit is likely to be ineffective. “The spotters are really the ones who control where the dispersants are applied,” Levine said. “I went to meet with these folks, and I found out that they were mostly agricultural spotters: they could tell you corn from wheat, but they couldn’t tell you the good oil from the bad oil. So my two main missions were to train them and to tell them that it was illegal to be spraying marine mammals, turtles, and birds. I told them if they are observed doing it they are going to jail. I kind of overstated it.” He ordered a halt to the spraying until the spotters could be sufficiently trained.

When the spraying resumed, it suddenly raised the total daily application to seventy thousand gallons—an all-time high—and Jackson rushed to blame BP for violating an agreement. “I was pretty accusatory,” she told me. “I thought that they were being very cavalier with an issue that people really cared about. Just because here you are doing something—injecting the chemical in a way that has never been done before, and we agreed with you because it is worth this chance to keep more oil away from critical ecosystems on the shoreline—your response then shouldn’t be to just use more of it.”

The more strongly Jackson spoke against BP, the more difficult it was to acknowledge the day-to-day control that the Unified Command imposed on the dispersants. “To me, it was clear that the E.P.A. did not believe that we had the level of control that we simply had—or they didn’t trust us,” a senior Coast Guard official at Houma told me. And yet responders there were obsessed by Corexit’s impact. “I told a colleague, ‘We are on a “Star Trek” mission—we’re going in uncharted territory,’ ” Levine recalled. “Once, I woke up in the middle of the night and called my boss and said, ‘I’m not sure we’re doing the right thing.’ ” Levine also phoned Alan Mearns, a NOAA biologist in Seattle, who had worked on the Exxon Valdez response. (He was famous for monitoring for twenty years a boulder that the cleanup did not touch. The boulder—eventually called Mearns Rock—recovered as quickly as the most aggressively cleaned areas.) Levine wanted to ask Mearns’s opinion about Corexit’s toxicity and persistence.

Scientists often test the toxicity of chemicals by pouring them into a tank filled with animals and seeing how many die after ninety-six hours. There are many studies of this kind for Corexit 9500. (Similar tests had been performed on an older version, Corexit 9527, which contains a chemical called 2-butoxyethanol that can be hazardous to humans. This older version was sprayed first in the Gulf, but ultimately to a much lesser extent.) Mearns and Levine knew from such tests that larval animals were the most likely to die from Corexit. But when researchers allowed the concentration to dilute over time, as it would in the ocean, the chemical was less lethal. Only a few tests examined how the dispersants might harm marine life without causing death, so Corexit’s “sublethal” effects are poorly understood, but nearly all experiments indicated that the chemical is less toxic than oil. “You have the ocean being a very large place, and if you dilute the oil to a certain point it is not going to be a problem,” David Valentine, a professor of microbial geochemistry at the University of California, Santa Barbara, told me. “So the question is: How long is it going to take to dilute it to that point?”

The total amount of Corexit sprayed from the air was 972,880 gallons. If the entire Gulf were the Superdome filled with water, this would amount to only a tablespoon, and all the oil would equal only three beer cans. “Of course, neither was fully mixed into the Gulf,” Mearns noted, but the opportunity for dilution in so vast a body of water was nonetheless great. Planes leaving Houma were calibrated to spray a layer of Corexit no thicker than two thousandths of an inch. Mearns made some very rough estimates based on this standard. In the first metre of ocean water, before the Corexit had been diluted by wave action, the concentration would be at most about five parts per million. In the first ten metres, the average concentration would not have exceeded half a part per million. Virtually no known species, even in larval stages, suffers from acute toxicity when exposed to Corexit at half a part per million; very few do at ten times that amount.

Still, there were risks. Atlantic bluefin tuna spawn in the Gulf, predominantly in April and May. Their eggs float in the upper part of the ocean, where the Corexit would be most heavily concentrated. At the same time, not using dispersants had ramifications, too. Thick, undispersed oil coated dolphins and turtles, killing them, and threatened wildlife in the delicate estuaries of the Mississippi. “We’re in nesting season, which meant that the birds in particular were not going to be easily removed from the site,” Laferriere told me. “When the media were focussing on oiled birds—that was because we couldn’t get them out of there. The first thing we try to do is to haze the birds. But mothers will not leave their nests, and I said, ‘Oh, my gosh, this is like a worst-case scenario,’ because right after you have the nesting you have the births, and then hatchlings that can’t fly. So we were in a very bad environmental situation. It is absolutely a tradeoff.”

About a month into the spill, the E.P.A. directed BP to identify a “less toxic” alternative to Corexit within twenty-four hours, or explain in detail why it could not. “How does the public know that the selection is the right one?” Jackson told me. “I believed it was a valid question.” Federal responders at Houma, who had not been consulted, questioned the ultimatum. Corexit was the most studied dispersant available; any other chemical would be inherently less well understood. Moreover, the demand was virtually impossible to comply with, because the standard tests take more than a day. “We all thought it was crazy,” one scientist at Houma said. Even as a demonstration of toughness, the directive backfired, by reinforcing the perception that BP was making decisions about dispersants on its own, and that Corexit should not have been used in the first place. On Fox News, Shepard Smith asked Jackson, “How in the world are we just figuring out this dispersant is toxic after we have put millions of gallons of it into the Gulf of Mexico?”

The following day, BP argued in a letter that only one alternative existed in sufficient quantity, and that the company was unable to acquire the proprietary formula for evaluation in so short a time. The company maintained that Corexit was the best option. Admiral Landry, who struggled to keep the various parties united, found the letter convincing. But Jackson was dissatisfied. “BP seemed to spend a lot of time saying why everything else didn’t work, but they really didn’t give me a feeling of comfort that they were asking honest questions,” she proclaimed, and vowed to have her own scientists conduct the testing—an announcement that only generated more confusion. Anderson Cooper didn’t interview Jackson about the directive. Instead, he raised the matter with a historian, Douglas Brinkley. “The E.P.A. sent out a letter to stop using dispersants, but that’s basically been ignored,” Cooper said. “Now the E.P.A. today is kind of saying, ‘Well, O.K., now we’re going to do our own research and try to find out.’ Why haven’t they been doing research for the last month?”

“Well, that’s a good question,” Brinkley said, politely.

Over a period of weeks, while the sorties continued, the E.P.A. conducted its own toxicity studies. They found that all the dispersants on the list, Corexit included, had “roughly the same impact on aquatic life.” This should have been encouraging news: the most plentiful dispersant was not the most damaging. But, as the E.P.A. began to reassess Corexit, Nungesser accused the agency of a coverup. “Now it appears the federal government is on the side of BP, with the E.P.A. jumping ship,” he told a reporter. “It’s an embarrassment for this country.”

In late May, the Unified Command made a concerted attempt to alter the public perception about dispersants. Oil had made landfall; the response was coming under furious criticism; the President had announced a moratorium on deepwater drilling in the Gulf, a decision that caused intense controversy. Some responders felt hopeless about communicating their views to the public. “You would go to your hotel at night—and the whole hotel is filled with response folks—and you congregate in the little lounge area, and you couldn’t even watch the TV,” Lieutenant Commander Chris Lee, Houma’s deputy operations chief, told me. “It felt like a battle, because if you watched the news you were losing, but in the response you were doing pretty much everything that could possibly be done.”

At the Unified Area Command, a plan was hatched to invite independent scientists to study the dispersant policy and issue a verdict on it. A professor at the University of New Hampshire, Nancy Kinner, was chosen to host a working group in Baton Rouge on May 26th. She brought in fifty experts, but on the first day of the meeting another directive came from Jackson—this time to “eliminate the surface application of dispersants,” except on rare occasions, and to cap the amount of Corexit for use undersea. “It kind of gutted the meeting,” a scientist who was at Baton Rouge told me. “They made what would appear to have been a political decision. There were E.P.A. officials there, and they—how can I put this delicately—they were basically being throttled; they were basically instructed not to take any positions at all.” (The E.P.A. disputes this.) Jeff Short, a chemist who had worked for NOAA in Alaska for decades, and was then a science director for Oceana, a conservation group, said, “We kind of rolled our eyes. That was a decision not informed by scientific advice.” Jackson told me that she was relying on her own agency’s expertise for the directive.

Like other attendees, Short had years of experience on oil spills, and was initially skeptical that the dispersants were being used prudently. But as he reviewed the data he began to change his mind. “We were all unhappy, in that we faced a situation where all of the options were bad, and we had to choose the least bad,” he recalled. “We were all uncomfortable with the choice, but we couldn’t think of a better choice.” Ultimately, every one of the scientists supported the dispersant strategy. “One of the nasty little things about dispersants is that if they are effective they are making it easier on the spiller,” Jeep Rice, a NOAA scientist, said. “But you have to wade through that and say, ‘I really don’t care what’s better or worse for the spiller. I want to do what’s best for the environment.’ ”

Responders at Houma hoped that the consensus would encourage Jackson to allow them greater control. The clause that limited the spraying of Corexit to “rare occasions” was strict: sorties had to be approved by a senior E.P.A. official, and formal estimates had to be made of the size and composition of slicks. In that time, the weather could change, or the oil could emulsify or break apart near land, making it a greater environmental threat.

A few days after the meeting in Baton Rouge, Dana Tulis, the E.P.A.’s acting director for the Office of Emergency Management, came to Houma, and several key responders tried to persuade her to urge Jackson to change her mind. Laferriere spoke of the volume of dispersants relative to all the Gulf water. Levine cited toxicity studies, and the technical problems inherent in skimming and burning the oil. Storms were forecast, and dispersants could be used in bad weather.

“Convince us why not using dispersants is better for the environment,” Levine told her.

“We’ve been getting a lot of complaints from the public,” Tulis said.

“Well, this is not an opinion poll,” he said.

As she was leaving, Laferriere caught up with her. “I said, ‘Don’t ever think for a second that this person standing in front of you would do anything to cause harm to the environment,’ ” he recalled. “ ‘I have an environmental-science degree, and I have a passion for this. I’ve been working on spills all my life.’ ” Laferriere paused. “You know, there was a nickname—that I was the Dispersant Devil—and I didn’t like that. I was a saint, hopefully, trying to protect the wildlife on the shoreline.”

The work was gruelling. Laferriere eventually had to step down, because of heart trouble. Levine told me that he and other federal responders quickly reached their statutory limit on overtime and by midsummer were working essentially as volunteers. Jim Black, Houma’s operations chief, worked frenetically for forty-one days in a row. “It was comparable to soldiers that have been to war together,” he told me. “Without being part of it, you just can’t understand.” He later died in an unrelated accident, and was posthumously honored by the Unified Command.

Jackson’s staff continued to oversee each sortie, and the spraying greatly diminished. On August 2nd, weeks after the last sorties, the E.P.A. completed another round of testing, which found that the dispersant-oil mixtures were no more toxic than the BP crude. Jackson later acknowledged, “How much toxicity was introduced? Turns out, not so much, really. Certainly less toxic than all the oil.”

VIII. THE PLUME HUNTERS

A hundred and fifty miles southwest of the wellhead, the Pisces, a NOAA research vessel, was searching for undersea plumes of oil. It was late on a September night, and in the darkness I climbed up to the bridge. A few NOAA crewmen stood by the helm, using red flashlights to see their equipment. Clouds obscured the moon, and it seemed as though the Pisces were suspended in nothingness. In the near distance, oil platforms looked like brilliant constellations of lights and flares.

By September, the BP well had been contained, and the most pressing questions for the response were: Where had the oil gone and how much harm had it done? A team of federal scientists had estimated that the total amount of oil that spewed from the well was 4.9 million barrels. Based on this number, the response estimated that seventeen per cent of the oil had been captured directly from the wellhead. The burns had eliminated five per cent of the oil; skimming had removed three per cent; and the Corexit had dispersed sixteen per cent into the sea. Altogether, the Unified Command appears to have removed and chemically dispersed two million barrels of oil—an amount equivalent to some of the largest spills in history. A comparable volume of oil seems to have naturally dissolved in the water column, or dispersed on its own, or simply evaporated. Scientists believed that only twenty-three per cent persisted, as tar balls, buried in sand, or elsewhere. Only fourteen miles of the Louisiana berms have been constructed; they absorbed barely any oil, and by stirring up silt may have caused the BP crude to sink and remain trapped in the marshland.

“The vast majority of the oil is gone,” a White House official announced on August 4th, but this statement misrepresented the Unified Command’s estimates. In fact, oil that had been dispersed was still in the water, even if it was no longer a threat to the shoreline. Scientists who were tracking hydrocarbon plumes in the ocean depths challenged the assessment, and the White House implored Admiral Allen to push back. He refused. “I told them that I was not going to take part in a public discussion,” he recalled. “But, once it was clear that this was going to be a basis for divisiveness, I had to go fix that. I figured the only way to get past this was to create a unified testing method for the Gulf.” At his direction, NOAA rapidly began to place academics in oceangoing ships, to assess the dissolved oil, and to conduct similar testing along the coastline. When I joined the Pisces, John Kessler, a chemical oceanographer from Texas A. & M., and David Valentine, the geochemist from U.C. Santa Barbara, were leading the ship’s science team. They took samples around the clock. Kessler oversaw the day shift, and Valentine worked nights.

While the crew maneuvered the Pisces in the darkness, I descended from the bridge. Valentine and the scientific team were in a small laboratory filled with computers, gas chromatographs, and other equipment. Three flat-screen monitors were mounted over a wooden desk in the center of the room, and Valentine was seated in front of them. He was unshaven, and looked tired. Just outside the lab was a large probe outfitted with electronic sensors and plastic cannisters to capture water. Under powerful spotlights, the Gulf ’s surface seemed pristine and impenetrable.

In “Moby-Dick,” Ishmael perceives the sea to be “indefinite as God,” but in fact the ocean is a place of considerable structure, with discrete sections layered across it horizontally. The sea’s uppermost layer is its brightest and smallest part, extending about two hundred metres from the surface. It is called the sunlit zone, and the vast majority of life in the ocean inhabits it. Tiny plants there generate half of Earth’s oxygen, much of which remains in the water. When they die, they fall several hundred metres, where vast colonies of bacteria consume their carcasses, using up tons of oxygen in the process. In this zone, Valentine told me, “respiration dominates,” and the water’s oxygen content reaches its minimum. Deeper still, the ocean descends into total darkness: the midnight zone, which makes up ninety per cent of the sea, though it contains only two per cent of its life. The water is as cold as thirty degrees. Oxygen-rich water from Earth’s poles, moving toward the equator, flows to this part of the sea.

Valentine ordered the probe down into the midnight zone. The framework entered the water, and feedback on one of the monitors showed the temperature dropping. The readings were normal until eleven hundred metres, when the oxygen levels dipped sharply. “This is a pretty clear oxygen anomaly here,” Valentine said. The probe fluoresced chemicals in the water, revealing the presence of hydrocarbon molecules. It appeared that the team had found a plume.

“What we’ve noticed is that much of the hydrocarbons that emanated from the BP leak ended up in the thousand-to-twelve-hundred-metre water depth,” Valentine told me. He looked at the anomaly again. “This is very typical for dissolved hydrocarbons,” he said. “Not typical for the ocean.”

When the probe was back on deck, the scientists drew water from the cannisters with large plastic syringes. The word “plume” was scientifically accurate, but the amounts of hydrocarbons in the samples were invisibly minute. In the syringes, the water was frigid and crystal clear.

Research in the Gulf has in many ways been encouraging. At the shoreline, pockets of oil will certainly linger. Although certain species may be at severe risk from the remaining oil, many others, such as sea turtles, do not seem to be under great threat, and the marsh as a whole does not appear to be ecologically devastated. All told, the spill killed fifty-six hundred birds, a dismaying number, but a small fraction of the quarter million that died in the Exxon Valdez spill. Oysters have suffered gravely, though this appears to be from the change in salinity caused by allowing the Mississippi to flow more forcefully into the Gulf. Luck certainly played a role in sparing large portions of the coast—a turn in the weather could have made the impact much worse—but a strategy based on dispersing the oil offshore appears to have helped prevent a great deal of crude from hitting land.

On the Pisces, Kessler told me that there was still oil in the ocean, “but it is being consumed—a lot of it has been consumed already.” Researchers have observed oil plumes being degraded by microorganisms, and Kessler and Valentine concluded that bacteria had consumed nearly all the methane from the spill. Along with a scientist from the Woods Hole Oceanographic Institution, Valentine also found small concentrations of dioctyl sodium sulfosuccinate, an ingredient in Corexit, suspended in the midnight zone; it is unclear whether this will be harmful. Ongoing research suggests that a layer of petroleum-based chemicals—apparently lethal to coral and other creatures—has coated portions of the Gulf sea floor, but no published study has verified that the layer includes large amounts of the BP oil.

On an expedition in June, Kessler and Valentine noticed that masses of dead plankton were floating at the water’s surface. Other researchers worry that the oil might be carried through the marine food web, and thus persist in the environment for years. At the same time, the spill seems to have been a gift for some species. During the cleanup, NOAA closed tracts of ocean to industrial fishing, which causes immense damage to the sea, and certain fish populations have tripled, possibly as a result. NOAA and the F.D.A. found that dioctyl sodium sulfosuccinate, the Corexit ingredient, was not building up in the tissue of seafood species. Although satellite data suggest that twenty per cent of the Atlantic bluefin tuna larvae in the Gulf were killed, a harsh blow, a main spawning “hot spot” was spared. Clearly, it will be years before the oil’s full ecological impact—especially the sublethal effects on plants and animals—is fully understood. Recent studies in Prince William Sound suggest that, in small ways, the ecological legacy of Exxon Valdez persists to this day.

The Pisces was now within sight of the wellhead, and Kessler walked on deck to gather phytoplankton samples from the sunlit zone. “It is about as easy as they come,” he said. “You throw a bucket over the side.” In June, he and Valentine were at the wellhead, and oil surrounded them in sheaths, some black, some reddish and emulsified, some iridescent sheen. “I mean, you just went in and out of it,” he said. “It was very apocalyptic. It’s like there were skyscrapers on the surface of the ocean trying to cap the spill.” Enormous flares of gas erupted as oil was brought to the surface. “We were under a literal black cloud for the entire expedition—it wasn’t the entire time, but there were huge, thick, black clouds of just soot from the burning operations.”

The Pisces moved quietly through the water. In the distance, one could still see the vessels that had been working to insure that the wellhead remained capped. They seemed unreal—industrial apparitions, gray like haze. Kessler tossed a bucket into the sapphire-colored water, and said, “You get this blue, kind of Caribbean color.” The sunlight was warm, and a breeze moved across the deck, and he added, “It’s hard to believe that there was really nasty thick crude all over this just a few months ago.” ♦