8

Meet the Press

For a number of years, along with two of MIT’s other star scientists—oceanographer Carl Wunsch and ocean geochemist Edward Boyle—Kerry Emanuel has co-taught a first-year graduate course, “Climate Physics and Chemistry.” One year, it must have been 1999 or 2000, Emanuel found himself listening to Wunsch lecture on the oceans’ deep overturning circulation, which transports tremendous amounts of heat from the equator to the poles. Cold, salty water sinks in the North Atlantic because it’s more dense—what some call the “thermohaline” circulation—but in order to maintain a global circulation, less dense water must also rise to the surface again elsewhere. As Wunsch pointed out, this in turn requires a process of turbulent upper-ocean mixing that drives warm water downward into the ocean’s cooler layers. Most oceanographers thought the necessary mixing occurred through tidal processes, but Emanuel—who knew of hurricanes’ powerfully violent churning of the seas, and the cold wakes they leave behind after their passage—started to wonder whether they might play a significant role.

Up to that point, Emanuel hadn’t taken very seriously the notion that hurricanes might interact with the climate in two directions, both being shaped by and also shaping it—in short, that they might be a feedback. Wunsch’s lecture, though, got him thinking. “I did a very quick back-of-the-envelope calculation,” Emanuel remembers, “and then a more thorough publication.” And so he became the leading proponent of a fairly revolutionary idea: Far from being isolated and rare—if awe-inspiring—events with no greater planetary significance, hurricanes, by mixing the oceans, might help determine what kind of Earth we live on. Tropical cyclones, much like extra-tropical ones, could play a fundamental role in regulating the climate.

Its well known that hurricanes draw up cooler waters from below the sea surface, leaving a pronounced cold wake as they pass. But the violent mixing works both ways. Hurricanes also drive warm surface waters down into the colder ocean depths. After a hurricane passes and leaves a cold wake, the sun gradually warms the surface mixed layer (as it’s called) up again. The warm water pushed down to the depths, though, persists as an anomalous increase in the net heat content of that particular column of ocean water. So in order to restore equilibrium after a hurricane, Emanuel reasoned, the tropical oceans must redistribute that excess warm water to cooler regions.

This implied that hurricanes might contribute significantly to the overturning circulation, driving heat away from the tropics and toward the poles. The idea had at least two points in its favor. First, hurricanes have a remarkable capacity to churn up ocean waters to considerable depths. And second, like strategically placed turbines, their tropical location puts them in the right place to perform the sort of mixing that would lead to poleward heat transport.

Emanuel first published this big if also speculative idea in 2001, adding that if hurricanes intensified, this could result in a relative cooling of the tropics and a heating of the higher latitudes. He also soon situated this new conceit in the distant past, much as he had done previously by linking hypercanes to the extinction of the dinosaurs. Hurricane-driven ocean heat transport, he speculated, might help explain the oft-noted “cool tropics paradox” of the early Eocene era, roughly 56 to 50 million years ago.

The early Eocene was an exceedingly warm period in the Earth’s history that has long puzzled backward-looking climatologists—as well as frightened them, because they suspect that may be where we’re heading if global warming continues. Sea levels at the time were dramatically higher, as were atmospheric concentrations of greenhouse gases (carbon dioxide and methane) due to natural causes. The Arctic coastlines were rimmed by temperate forests, and populated by warm-weather plants and animals (including crocodiles in Greenland). Yet based upon the analysis of oxygen isotopes found in the fossilized shells of tiny marine organisms called foramanifera, scientists had found that even as polar temperatures during the early Eocene were dramatically warmer than today, temperatures in the tropics were only slightly higher.

How could this have happened? Emanuel proposed an explanation that, at least to some extent, could be tested: Stronger hurricanes had helped transport large amounts of heat out of equatorial regions. Emanuel’s attempt to investigate this idea, however, would lead to a very different kind of turbulent mixing: between scientists.

 

Emanuel couldn’t directly observe many of the central processes stipulated by his hypothesis. He couldn’t go out and track ocean heat fluctuations from a boat during or immediately after every strong hurricane. But he could use models and theoretical considerations to calculate how much heat hurricanes might help to transport, and whether it sufficed to bring about a significant climate-scale effect. This, in turn, depended on how extensively these storms mix the ocean; and that depended on the collective strength of hurricanes summed over their lifetimes.

Traditional measurements of storm frequency and intensity were ill-suited to such an inquiry, however. Simply focusing on the total number of storms lumped together weak tropical storms with Category 4 and 5 monsters, even though the latter, with gigantic waves like those observed in Hurricane Ivan or Typhoon Tokage, had a much greater capacity to mix the oceans. Simply focusing on a storm’s peak intensity, meanwhile, told you nothing about how long it managed to maintain that intensity. The longest-lived tropical cyclone on record, Hurricane/Typhoon John of 1994, lasted for 31 days, during which time it traveled 8,000 miles from its point of origin in the Northeast Pacific basin all the way across the world’s largest ocean to the Northwest Pacific basin, and then back east into the Central Pacific again. Along the way, John strengthened and weakened multiple times, spanning the entire gamut of hurricane intensity and at one point becoming a Category 5 storm. Obviously, then, John had mixed the oceans far more than a weak tropical storm that fizzles out within days of its formation.

Frequency and intensity measurements also ignored the important issue of storm size. All else being equal, larger hurricanes mix the oceans more than smaller ones. For all these reasons, traditional frequency and intensity metrics failed to capture how much total hurricane power existed—or, for that matter, the true magnitude of the potential hurricane threat to human beings. For a strong overlap exists between hurricanes’ capacity to stir the seas and their capacity to destroy outposts of civilization. Both depend on wind speed, with longer periods of high winds making it more likely (all else being equal) that storms hit something when they’re at their most dangerous.

To account for all this, Emanuel turned to his equations. He derived a new measurement of hurricane strength that he called the “power dissipation index” (PDI)—some have since labeled it a “potential destruction index”—which measured the total amount of energy released by a hurricane over the course its life. The calculation began with peak sustained wind speeds taken at regular six-hour intervals over the entire history of a storm, and then converted these measurements into a reflection of total storm power by multiplying each wind-speed measurement by itself twice, or cubing it, and adding all the cubed measurements together. Though this might sound like a surprising mathematical move to make, it’s actually fairly uncontroversial among hurricane specialists, who know very well that a storm’s destructive capacity goes up “as the cube of the wind speed,” as it’s often put.

When Emanuel examined the new hurricane data, he suddenly saw something very surprising. Increasing storm lifetimes, coupled with increasing intensities, had apparently resulted in a doubling of the amount of power dissipated by Atlantic and Northwest Pacific storms over the past thirty years. “The trend sort of jumped out of the data,” Emanuel remembers. Most significantly, he found that increases in PDI correlated closely with rising sea temperatures—themselves generally thought by scientists to reflect the impact of global warming.

And this was no minor trend. Emanuel’s data suggested a much bigger increase in storm power than predicted either by Knutson’s latest modeling work or Emanuel’s own theory of a hurricane’s maximum potential intensity. For the first time, actual data suggested that hurricanes were intensifying; simultaneously, a huge gulf had opened between that data and preexisting theory, which had postulated a much lower sensitivity of hurricanes to the relatively modest changes in climate that had been seen so far. And now Emanuel, who’d come up with the theory in the first place, listened to the data, and found himself in a position much closer to the one that had been staked out by Kevin Trenberth. As his paper stated: “My results suggest that future warming may lead to an upward trend in tropical cyclone destructive potential, and—taking into account an increasing coastal population—a substantial increase in hurricane-related losses in the twenty-first century.” Or as he put it in an interview with Discover magazine: “For the first time in my professional career, I got alarmed.”

 

Figure showing the correlation between Atlantic sea-surface temperatures and hurricane power dissipation. Courtesy of Kerry Emanuel.

 

Nature published Emanuel’s results online on July 31, 2005. The study, which made pretty much everybody’s list of top papers for that year, constituted a thunderclap. Here was perhaps the world’s leading hurricane theorist—the scientist who had single-handedly generated the concept of hurricane intensification due to global warming as a theoretical possibility, but who had remained unconvinced until now that such intensification had yet made itself manifest—undergoing a very public conversion. “I changed my mind in a big way after 2004,” Emanuel would later remark. “I was very open about it, and I think it’s a strength of science that when ideas you have are challenged by the data, you be prepared to modify them.”

 

Even before Emanuel’s paper came out, the scientific realignment had begun. In July 2005, just as an unprecedented hurricane season began in the Atlantic, Emanuel had his name removed from a group paper downplaying global warming-hurricane concerns and, in essence, seeking to reiterate the 1998 consensus. Emanuel tried to bring his co-authors around to his new view behind the scenes, but they weren’t convinced. So he e-mailed one of them, noting that “the problem for me is that I cannot sign on to a paper which makes statements I no longer believe are true.” He added, “I see a large global warming signal in hurricanes.”

The more skeptical paper read, in part, as a rebuttal to hurricane-climate links made following the 2004 season by Trenberth and others. It ultimately appeared in the Bulletin of the American Meteorological Society. Most of its authors worked for NOAA or in the hurricane-forecasting community; they included Landsea, Mayfield, National Hurricane Center forecaster Richard Pasch, Climate Prediction Center director Jim Laver, and lead author Roger Pielke, Jr., a disaster policy expert at the University of Colorado who argues that population growth and economic development in coastal areas are the chief cause of our vulnerability to hurricanes, with or without global warming. (Emanuel doesn’t necessarily disagree, but thinks the impacts from global warming will become more pronounced with time.)

The timing of Emanuel’s study, his intellectual conversion, and his withdrawal from the paper—together they foreshadowed a coming scientific tempest. The weather also contributed to the organization of this political storm. Emanuel’s Nature paper appeared online a month before the landfall of Hurricane Katrina, but at a time when the Atlantic had already been behaving oddly when judged by historic standards. Traditionally, July has not been a very active hurricane month in the basin, though its well within the bounds of the official season. But in 2005, July featured a record five tropical cyclones, of which three became hurricanes^* and two, Dennis and Emily, were extremely intense hurricanes that impacted populated areas.

Both Dennis and Emily developed from tropical waves, as the most powerful Atlantic storms usually do. One of those scary hurricanes that come up through the Caribbean seeming to ricochet off islands, Dennis kept flirting with early-season intensity records held by 1957’s Category 4 Hurricane Audrey, a June storm that devastated the eastern Texas and western Louisiana coasts and killed more than five hundred people. “The bayou folk swam, clung, gasped and prayed for their lives,” reads a contemporary Time magazine account of Audrey’s assault on Cameron, Louisiana. “Those lucky enough to reach specks of dry land found only more terror: with them were alligators and water moccasins, tossed out of the torrent, snapping and striking in their fury.” But Dennis surpassed Audrey’s record for hurricane strength before the month of August, with a pressure drop down to 930 millibars (27.46 inches) over the Gulf. The storm weakened before coming ashore as a Category 3, but its surge, six to seven feet above normal tides, entirely overwashed parts of Santa Rosa Island.

Dennis overlapped with the season’s next extremely intense storm, Emily, whose path through the Caribbean pointed toward the Yucatan Peninsula, which the storm smacked at Category 4 strength on July 18. In retrospect, however, the remarkable development came a day earlier. In a postseason reanalysis, the National Hurricane Center noted that as Emily passed southwest of Jamaica, its maximum sustained winds had briefly topped out at over 160 miles per hour, even as central pressure dropped to 929 millibars (27.43 inches). These measurements meant Emily had snatched away Dennis’s record while also setting another one: earliest-forming Category 5 storm in the Atlantic. No previous Category 5 hurricane had ever been recorded in the basin in July.

Dennis and Emily appeared to lend urgency to Emanuel’s paper, whose publication followed their landfalls within weeks. Journalists linked the storms and the science, and they also turned to Landsea and Gray for their reactions to the work. Landsea went far easier on Emanuel than he had on Trenberth; he took a critical but measured tone. He praised the study, but said he didn’t trust the wind speed measurements at its foundation (a critique he would articulate in much more detail soon enough). Gray, however, let loose, describing Emanuel’s work as “a terrible paper, one of the worst I’ve ever looked at” to the Boston Globe.

The substance of Gray’s criticism was similar to Landseas. Emanuel had cubed wind-speed estimates that were themselves prone to considerable error, Gray complained. Emanuel countered that while the data surely did have errors, the strong correlation he had observed between increasing storm strength and increasing sea-surface temperature was very striking (and worrying). These were two different data sets that measured different things. It seemed unlikely to Emanuel that errors in the wind-speed measurements would fortuitously skew the data in such a way that it just happened to closely track an independent rising trend in ocean heat.

Gray’s description of Emanuel’s work as a “terrible paper” was among the earliest of a volley of barbs, sometimes issued publicly through the media, that would be exchanged as the hurricane-climate debate ramped up in late 2005. Soon Gray told the Los Angeles Times that “the people who have a bias in favor of the argument that humans are making the globe warmer will push any data that suggests that humans are making hurricanes worse.” “People are jumping out of the woodwork to say that storms are stronger because of global warming, but they’re mistaken and most of them don’t know what they’re talking about,” he told yet another paper. And speaking of jumping—Gray told USA Today, “If I’m proven wrong, I will jump off the highest peak in Colorado.”

Reading these quotations, it sounds as though Gray talked to reporters in the same unguarded way he talks when you’re sitting in his kitchen. Saying one’s scientific peers are biased and don’t know what they’re talking about tends to anger them, though—and Gray said more than that. He also both implied and stated that Emanuel, and the authors of another soon-to-be-published hurricane-climate paper, were coming up with these results to advance their careers and get increased research funding. On CNN, Gray said of those linking hurricanes and global warming, “There are all these medicine men out there who want to capitalize on general ignorance on this subject.” In an interview with Discover magazine he was asked why some scientists support a hurricane-climate link and replied, “So many people have a vested interest in this global-warming thing—all these big labs and research and stuff. The idea is to frighten the public, to get money to study it more.”

If Gray called Emanuel’s work a “terrible paper,” it soon became clear that his long-range forecast for 2005 also had its shortcomings. In December 2004, Gray announced that he did not expect “anything close to the U.S. land-falling hurricane activity of 2004” for the coming season, predicting just eleven named storms, six hurricanes, and three major hurricanes, a slightly above-average year by historic standards. By late May 2005, Gray had upped the forecast to fifteen named storms, eight hurricanes, and four major hurricanes, while NOAA, which had been releasing its own seasonal forecasts since 1998, had similarly predicted twelve to fifteen named storms, seven to nine hurricanes, and three to five major hurricanes. But these final preseason estimates were still too low, almost by half; 2005 ultimately shattered all records, producing twenty-eight storms, fifteen hurricanes, and seven major hurricanes. Four were Category 5 storms, and three were among the six most intense hurricanes ever measured in the Atlantic. Such a large gap between the forecasts and reality clearly called into question whether Gray’s or NOAA’s statistically based techniques could predict such a record year—in advance of the actual season, at any rate.

Meanwhile, a newer brand of forecasting that Gray would surely view with considerable skepticism—dynamical seasonal hurricane forecasting, which employs global-climate models to project storm numbers—had shown the glint of an ability to perform as well as, or possibly somewhat better than, the statistical techniques. On May 1, 2005, the coupled ocean-atmosphere climate model run by Meteo-France, the nation’s weather service, predicted 22 named storms in the Atlantic. By June 1, models run by the United Kingdoms Met Office and the European Centre for Medium-Range Weather Forecasts in Reading, England were also predicting active seasons, and if their results are taken together, these three European models predicted 16.2 storms—still well short of reality, but slightly closer to it than the preseason statistical forecasts.

The dynamical seasonal hurricane forecasting technique relies upon running ensembles of global climate models to get a sense of the probability, across many model runs, of how many hurricane-like storms will appear. In short, the technique hones in on the same computer-generated storms that empiricists had so criticized for not adequately resembling real-life hurricanes. At present, this method remains in its infancy. Yet just as was the case for standard weather forecasts or forecasts of hurricane tracks, it’s possible or perhaps even probable that numerical models will get better and better over time and eventually eclipse statistically based techniques for the purposes of predicting seasonal hurricane activity. After all, the numerical models have an advantage: They’re not tied to past history. They simply solve the equations. If global warming is dramatically changing hurricanes, as Emanuel’s work suggested, such models may stand a far better chance of producing accurate forecasts.

 

And so with a high-profile new paper linking global warming to strengthened hurricanes, and with statistical and dynamical forecasts alike predicting a busy season, August arrived for the Atlantic basin. Hurricane Irene, a long-lived Cape Verde-type storm, spent nearly half the month dissipating its store of power over open ocean. And then came Katrina, the most damaging hurricane in U.S. history, with an estimated cost of over $80 billion and more than 1,500 killed. “When it hit land—while Chris Landsea rode through the storm in a NOAA plane—Katrina also drove a record storm surge, twenty-four to twenty-eight feet in some places in Mississippi, higher even than the surge driven by Hurricane Camille (a fact attributable to Katrina’s enormous size). And there were other shocking measurements: sustained winds of nearly 175 miles per hour over the Gulf produced during a rapid bout of intensification once Katrina hit the infamous Loop Current; and a minimum central pressure measurement of 902 millibars (26.64 inches), temporarily placing the storm fourth among the most intense hurricanes ever measured in the Atlantic basin, ahead of Camille and Mitch (both 905 mb; 26.72 in.) and just behind 1980s Hurricane Allen (899 mb; 26.55 in.). But Katrina’s intensity ranking would drop to sixth after hurricanes Rita and “Wilma. If the 2005 hurricane season had been an Olympic race, it would have been one in which multiple runners set world records.

A brace of books about Katrina and its aftermath have since been written. As a result, the appalling incompetence of the Federal Emergency Management Agency following the storm has been well established. So has the negligence of the U.S. Army Corps of Engineers (and, let us not forget, the members of Congress who ultimately oversee it). Not only had the Corps constructed faulty levees; it had redirected the Mississippi to starve protective wetlands and built worse-than-worthless projects like the Mississippi River Gulf Outlet (“MR. GO,” it’s not-so-amusingly called), which helped channel Katrina’s storm surge straight into New Orleans. By contrast, one government agency that did a stellar job during Katrina was the National Hurricane Center. Its stressed and overworked forecasters had nevertheless produced track and landfall predictions that were stunningly accurate many days in advance—giving other agencies, like FEMA, ample time to prepare. But the hurricane center’s performance was the exception, not the rule.

The broad failings by our government helped ensure that Katrina had a far more devastating impact than it would have had otherwise. For contrary to the words of George W. Bush, what occurred could not simply be chalked up to the “whims of nature.” Had Katrina hit New Orleans directly as a Category 5 hurricane, surely nothing could have prevented the utter destruction of the city. But that’s not what happened. Instead, the storm weakened before landfall and swerved away. It can’t be emphasized enough: Katrina missed. “Overall, it appears likely that most of the city of New Orleans experienced sustained surface winds of Category 1 or Category 2 strength,” the National Hurricane Center concluded.

This was hardly a worst-case-scenario hurricane, meaning that only human failure can explain how Katrina still managed to take such a toll. No one knew at the time about the poorly constructed levees, but New Orleans’s general vulnerability had been discussed ad nauseum, especially during Ivan only a year earlier. There was no excuse for Bush or for FEMA to be so fundamentally unprepared, so clueless.

Without a doubt, humans can make themselves starkly vulnerable to extreme weather events based upon where they live (often at dangerous sites where land meets sea) and how they alter their environments (such as by destroying wetlands). But nature must also cooperate to engender true catastrophe. Usually it will do so eventually, provided favorable conditions prevail—and an extremely hospitable atmospheric and oceanic environment had made Dennis, Emily, Katrina, and later Category 5 storms Rita and Wilma possible in the first place. That environment had been supercharged for hurricanes due, among other factors, to its anomalously warm sea temperatures. Might humans have had something to do with that as well?

In a later retrospective, a team of NOAA scientists that included Landsea pointed out the many hurricane-ripe conditions that prevailed in 2005: record sea-surface temperatures in the main hurricane development region, La Nina-like conditions in the tropical Pacific (the opposite of El Nino), and low levels of wind shear. NOAA didn’t say anything about a role for global warming, however; rather, the analysis favored an explanation that invoked natural cycles. (Emanuel might have called it a reiteration of the agency’s party line.) The 2005 hurricane season also featured relatively weak easterly trade winds across the Atlantic. Stronger trade winds cause more evaporation from the sea surface and thus more cooling; weaker winds leave the oceans warmer. So it seems that in 2005, hurricanes took up some of the slack for the trade winds in ventilating the oceans.

But why had the oceans gotten so hot in the first place? In the region of the tropical Atlantic central to hurricane formation, the June-to-October temperature anomaly broke all records: .92 degrees Celsius above normal. If we step out of the chronological sequence of events for a moment, we can consider the scientific paper that has linked this most closely to climate change: a 2006 study by Trenberth and his colleague Dennis Shea, who tried to break down how much of the 2005 anomaly could be connected with a global trend of ocean warming attributable to human influences. Through statistical analysis, they credited .45 degrees Celsius, or roughly half of it, to the global trend. The rest they attributed to the aftereffects of the 2004–05 El Nino and the Atlantic Multidecadal Oscillation, which in their analysis had only a small impact.

Trenberth had statistically tied a central environmental factor underlying the record hurricane year—sea-surface temperature—to global warming. Assuming that’s correct, it’s significant. But it’s still a far cry from asserting that any single storm, such as Katrina, could be attributed in any meaningful way to climate change. For members of the public, however, who often conceptualize global warming in a visceral rather than a statistical sense, and who invariably connect it to whatever weather extremes they might be experiencing at a given time, this point hardly came across so clearly in Katrina’s aftermath. It didn’t help that many journalists and commentators stoked the confusion.

First came the various pundits and politicians who went overboard. In an opinion article published in the Boston Globe on the day after Katrina’s final landfall, veteran global warming reporter Ross Gelbspan wrote, “The hurricane that struck Louisiana yesterday was nicknamed Katrina by the National Weather Service. Its real name is global warming.” The German environment minister Juergen Tritten went even farther with an attack on President Bush, writing (according to the Washington Post’s translation), “The American president closes his eyes to the economic and human damages that are inflicted on his country and the world economy by natural disasters, like Katrina, through neglected climate protection.”

On a scientific level, such assertions are indefensible. As Greg Holland says of those who claim global warming strengthened or otherwise affected Katrina: “It’s just unadulterated garbage. That cyclone was reacting to the immediate environment that it was in.” When it comes to global warming, Holland adds, “All you can say is that as time goes on, if you were hit by a tropical cyclone, the chances are now higher it’s going to be a Category 4 or 5 than they were before.” Emanuel has similarly stated that attributing the Katrina disaster to global warming would be “absurd.” But given that Gelbspan was writing only a day after Katrina and amid the national alarm and horror over the event, we should probably forgive him for overstepping a bit. No one was thinking in a very nuanced way at the time.

Talking heads weren’t the only ones connecting Katrina to global warming. In a considerably more subtle but perhaps also more powerful way, the major media did so as well. During prior active Atlantic storm seasons, journalists had written occasional “he said, she said” stories about global warming and hurricanes, quoting the scientists on either side of the debate. This coverage had often been highly episodic in nature, frequently following in the aftermath of destructive land-falling storms like Hurricane Andrew. Such a pattern of reporting, by its very nature and timing, helped to foster the misleading impression of a direct causal link between individual weather events and long-term climate trends.

Katrina took this pattern to a new extreme, triggering unprecedented media discussion of the hurricane-climate relationship. The storm produced saturation news coverage of people stranded on rooftops and suffering in the Superdome and at the New Orleans Convention Center. Amidst this round-the-clock stream of information, the hurricane-climate relationship represented just one theme among many—including race, poverty, leadership failings, and government incompetence. Nevertheless, the attention to it far exceeded anything that had come before.

If you plot the number of articles over the years discussing the hurricane-global warming relationship in two major agenda-setting newspapers, the New York Times and the Washington Post, the resultant figure bears a striking resemblance to the wind-speed measurements of a hurricane that remains relatively weak for many days, then suddenly runs over deep warm water and explodes into a Category 5 storm. From 1985 through 2003, discussion of this subject, whether brief or in-depth, remained at a relatively low level. Those slight blips that did occur were perhaps partly tied to the occurrence of noteworthy hurricanes, or to higher levels of attention to the broader global warming issue (for example in 2001, when President Bush withdrew from the Kyoto Protocol). But there were no major media feeding frenzies, not even following destructive land-falling storms like Hugo and Andrew.

In 2004, however, coverage started to rise precipitously, linked to the Florida storms. The next year that rise became meteoric, almost certainly thanks to the combination of alarming new research (like Emanuel’s) and Katrina. Some forty articles discussing the connection between hurricanes and global warming appeared in the month of September 2005 in these two publications alone. Other highly influential media outlets reflected the trend as well. In October 2005, a Time magazine cover asked, “Are We Making Hurricanes Worse?”

 

Combined articles in the Washington Post and the New York Times discussing the relationship between global warming and hurricanes, plotted over time.

 

Emanuel himself missed much of this frenzy. Just after publishing his Nature paper, he headed off for a long-planned sabbatical to the Mediterranean island of Majorca. The experience, he would later note, helped provide a nice “filter.” But in Emanuel’s absence, a major new environmental factor invaded the hurricane-climate argument: the media. Journalists made the debate public as never before, and the scientists involved became newsmakers whose every word seemed fraught with significance. They in turn started worrying about how they came off in news articles, and what other scientists were saying about them—particularly Gray, who frequently slammed others’ work.

Media coverage that folded global warming into the Katrina story line also seemed to parallel changes in public opinion. In a survey by the Pew Research Center released on September 8, shortly after Katrina, 25 percent of respondents said the storm’s severity had been “caused by global climate change.” Democrats and Independents were far more likely than Republicans to make such an attribution in the poll, underscoring the partisan nature of the global warming issue.

In turn, widespread links between global warming and Katrina—sometimes articulated directly by misguided commentators, sometimes merely implied by the timing and tenor of media coverage—triggered a fierce backlash. To cite just one example, on September 9, 2005, the conservative Washington Post columnist Charles Krauthammer baldly asserted, “There is no relationship between global warming and the frequency and intensity of Atlantic hurricanes. Period.” In light of the emerging science, such an absolute decree was at least as incautious as Gelbspan’s claim that Katrina’s “real name is global warming.”

Those opposing mandatory cuts on greenhouse gas emissions had good reason to fear that Katrina could prove a pivotal event, forever altering the views of Americans on climate change. And in the wake of the storm, it certainly looked as though they were trying to do something about it. The think tanks of the political right, ever battling over the science of climate, became hurricane experts. The George C. Marshall Institute, partly funded by ExxonMobil and long a hub for global warming skepticism and contrarianism, fired out a news release on September 6 declaring, “Linkage Between Hurricanes and Global Warming Tenuous.” Tech Central Station (as it was then called) and the Competitive Enterprise Institute put out similar messages.

The next month, the Marshall Institute followed up with an event at the glitzy National Press Club in Washington, D.C., entitled “Atlantic Hurricanes: The True Story.” The event featured Florida State University oceanographer James O’Brien, an El Niño specialist, as the expert on hand. O’Brien proceeded to propose the most narrow metric conceivable for assessing storm changes: He would examine whether a trend could be seen in the intensity of U.S. land-falling hurricanes, which comprise only 3 percent of the global total of storms. It’s true virtually by definition that we have more reliable data on land-falling storms than on any others. But global warming should affect storms regardless of whether they happen to make landfall, which meant that focusing arbitrarily on such a tiny subset of hurricanes could greatly obscure the real issue.

Sure enough, O’Brien could find no trend in land-falling storm data. But this absence could hardly be said to invalidate Emanuel’s vastly different approach. And then O’Brien went farther, acknowledging a global trend of ocean warming but questioning what might have caused it. “It is warming up in many places in the ocean. Whether it is man-induced or whether it is natural variability—who knows?” he remarked.

In fact, the scientific literature strongly demonstrates that, just as for warming surface-air temperatures, the current trend of warming sea temperatures cannot be explained without invoking a significant role for human-induced changes to the climate. Only a few months before O’Brien’s pronouncement, around the start of the Atlantic hurricane season, a major study attributing the warming of the world’s oceans to human influences had been published by leading oceanographers and climate scientists in the journal Science. Noting a substantial ocean-by-ocean warming trend, the scientists had run various climate-model studies seeking to replicate the signal. They found that only models that factored in greenhouse gas and sulfate aerosol emissions could do so successfully. Models that merely took into account natural variability, whether springing from causes internal to the system or external phenomena (such as volcanic eruptions and changes in solar intensity), failed to explain the data. “The immediate conclusion is that human influences are largely responsible for the warming signal,” the study found.

At the time, this was the definitive study attributing the heating of the oceans to human influences. O’Brien didn’t mention it. But the hurricane-climate skeptics would find themselves increasingly on the defensive as the 2005 hurricane season progressed.

 

What had begun, then, as Emanuel’s attempt to test a somewhat esoteric hypothesis about a hurricane—climate feedback had suddenly translated into the biggest scientific story of the year—and it was about to get bigger. Scientific papers get published in the thousands, but with this particular one all the stars had been aligned. Emanuel already knew the hurricane-climate issue carried with it a unique political sensitivity; the 2004 hurricane season and its aftermath had amply demonstrated as much. But he couldn’t have known that his intellectual conversion would go hand in hand with an even more staggering hurricane year, leading the media to take up the subject as a new obsession. Moreover, another scientific bombshell would soon explode, this one with direct origins in the Trenberth-Landsea argument from the previous year.

As a result of all of these independent developments, the issue of hurricanes and global warming, once a side subject among the vast array of concerns that fall under the rubric of global change, was about to become virtually coterminous with global warming itself. The sublime image of a hurricane as glimpsed from space would become the new icon of climate change wielded by environmental groups and their supporters. Meanwhile, industry and conservative think tanks would harden their skepticism, and a group of not-entirely-suspecting scientists would soon be swept into the fight of their lives.