7

Frictional Divergence

Herbert Riehl had given Bill Gray a spectacular first storm-flying experience with their low-level penetration of Category 4 Hurricane Helene in 1958. But it could hardly rival the experience that Gray passed on, almost like an heirloom, to Chris Landsea thirty years later.

Landsea had grown up in Miami and always had an interest in storms: Category 3 Hurricane Betsy hit in 1965, just two weeks after his family moved there. In 1988, after getting his undergraduate degree in atmospheric science at UCLA and working for a year at NOAA’s Aircraft Operations Center (home of the famous Hurricane Hunters), Landsea started graduate school under Gray. He’d only been out in Colorado a month when a hurricane named Gilbert came cruising across the Caribbean, and Gray told Landsea and fellow grad students Jim Kossin and Steve Hodanish, “I can get you guys onto a flight if you want to go.” They did. “We just showed up in Miami,” Landsea remembers. It probably didn’t hurt that he already knew everyone at the Aircraft Operations Center.

And so, on September 13, the three students found themselves hurtling toward an extremely intense Hurricane Gilbert on board one of NOAA’s long-range Orion P-3 turboprop research planes, equipped with high-tech instrument consoles for each passenger as well as plenty of barf bags. The timing couldn’t have been more perfect. Landsea, Kossin, and Hodanish were just “twenty-three-year-old kids having a good time on board this plane” (their idea of a “good time” being 199-mile-per-hour gusts and 40-mile-per-hour updrafts). But they had managed to get onto the historic flight that, at an altitude of 3 kilometers, recorded (by extrapolation) a central pressure of 888 millibars in Gilberts extremely compact eye, which was just ten kilometers across. It was the lowest pressure ever measured in the Atlantic basin and would remain so until the record was broken seventeen years later in Hurricane Wilma. Landsea, Kossin, and Hodanish had hitched their first ride into the strongest hurricane ever observed in the Western Hemisphere.

For Landsea—who would go on to work at NOAA’s Hurricane Research Division (the descendant of the National Hurricane Research Project) and later at the National Hurricane Center—this marked the first flight in a long career of airborne storm research. That includes sorties into Opal in 1995, Georges in 1998, Floyd in 1999, and finally Katrina even as it made landfall in Louisiana and Mississippi.

Much like Gray, Landsea is tall, genial, and entertaining—a crowd-pleaser. He sometimes opens his public talks with jokes that turn on his all-too-appropriate last name—for example, observing that if he had a son named “Aaron,” the kid’s full name would be Aaron Landsea (“Air and Land Sea”). Bespectacled but tanned and muscular, Landsea is the kind of scientist you’d expect to find living in Miami. At less formal scientific meetings, he wears Hawaiian (or as he calls them, “tropical”) shirts, and sometimes a shirt in which hurricane spirals replace floral motifs.

Above all, Landsea is the unrivaled master of Atlantic hurricane data. He heads up the HURDAT project, an ongoing reanalysis of Atlantic storms back to 1851 that aims to get the most precise statistics possible about their tracks and intensities (information that is obviously valuable to insurance companies seeking to assess risks for a particular region). Reanalysis led to the upgrading of Hurricane Andrew from a Category 4 to a Category 5 storm at landfall. But its value is perhaps even better epitomized by Landseas role in rediscovering the hurricane that affected San Diego, California, in October of 1858, when it was a mere settlement of a few thousand people (today the population of San Diego County exceeds 3 million).

Studying weather records and old newspaper reports in collaboration with an independent scholar named Michael Chenoweth, Landsea helped document a storm in which roofs had been torn off houses and ships had been driven up onto beaches. Los Angeles, too, experienced heavy rains and flooding. Drawing clues from reports of the damage, Landsea and Chenoweth inferred that San Diego had experienced Category 1-force hurricane winds, although the storm that produced them (presumably originating in the Northeast Pacific hurricane basin off the coast of Mexico) had remained just offshore. Such an event had never been recorded before or since, but Landsea and Chenoweth had forever distinguished the unlikely from the impossible. “The risk of a hurricane in southern California is now documented to be real,” they wrote. In conclusion, they linked the 1858 San Diego hurricane to seeming record-breaking storms like 2003’s Ana and 2004’s Catarina. “These are yet additional reminders that if we wait long enough, or dig deep enough in archives, that more surprises are in store to challenge our assumptions about the frequency, seasonality, and location of tropical cyclones,” they noted.

This underscored a central theme that would reemerge repeatedly from Landsea as the hurricane-global warming debate intensified in 2004 and beyond: “We don’t have good enough data to fully grasp the diversity of hurricanes that have existed in the past. As a consequence, we should expect to be continually surprised by apparent “record-breaking” storms or groups of them—like Catarina, or the Japanese typhoons, or the four Florida hurricanes—that would have prior analogues if only we could survey more of meteorological history. So don’t blame global warming, the argument goes, blame our skimpy observational records. A “record-breaking” storm, after all, surpasses only what has been recorded.

Yet even if Landsea’s argument is true—and to some extent it must be—climatic changes could also begin to surprise us with unprecedented storms, and it would take many years to determine which factor (poor measurements of the past or ongoing changes in the present) better explained them. The 1858 San Diego hurricane was, assuredly, just a rare event. It may have been made possible by El Niño conditions, which heat sea temperatures off the West Coast, allowing hurricanes to travel farther northward. El Niño represents a natural mode of variability; but if global warming sufficiently increases average sea temperatures off the coasts of San Diego and Los Angeles, it’s certainly conceivable that these cities could also face a greater risk of future hurricanes.

 

As suggested by his reanalysis research, his work on Gray’s hurricane forecast, and their collaborative studies of hurricanes and Sahel rainfall, Landsea is a strong meteorological empiricist. Like Gray, he digs down deep into the data—including getting dusty in the historical archives—rather than running models or designing theories reliant upon complex mathematics.

That’s not to say that Landsea denies global warming outright or entirely dismisses numerical modeling. He doesn’t go nearly so far as Gray on either of these fronts and is willing to criticize his mentor (“I wish he’d publish more,” he told me). Still, it’s easy to see the overlaps and influences. Landsea, like Gray, has long criticized attempts to link hurricanes and global warming, as well as attempts to study hurricanes in climate models. In one very widely cited 2001 paper, Landsea joined Gray and other hurricane experts in arguing that the uptick in Atlantic hurricane activity since 1995 involved an ocean-driven cycle, the Atlantic Multidecadal Oscillation, rather than climate change.

It came as no surprise, then, that in media interviews during and after the 2004 hurricane season, Landsea joined the other hurricane specialists in refuting any connection between the destruction and global warming. Yet even as they strove to do so, Landsea, Gray, and their colleagues couldn’t keep global warming entirely out of the picture. A scientist working within NOAA published a study in late September 2004, with Florida still reeling from Hurricane Jeanne, that made that impossible. It wouldn’t be the last time in the hurricane-climate debate that the scientific publication process and the weather seemed curiously synchronized.

In a tour-de-force piece of work, Thomas Knutson of the Geophysical Fluid Dynamics Laboratory and a coauthor, Robert Tuleya of Old Dominion University, went far beyond previous computer studies of hurricane intensity changes for increased concentrations of atmospheric carbon dioxide. They drew upon results from nine different global climate models whose scenarios raised tropical sea-surface temperatures by .8 to 2.4 degrees Celsius. The scientists fed the various climate-model results into the high-resolution GFDL hurricane model, and then ran experiments using four different cloud schemes, three different ocean basins, and so on. In total, the study involved nearly 1,300 simulations of hurricanes under enhanced greenhouse (or control) conditions. Any consistent results uncovered across such a large range of experiments would be very difficult to attribute to the quirks of a single model.

And Knutson and Tuleya found a number of consistent results in response to carbon dioxide-induced global warming. On average, maximum hurricane wind speed increased by 6 percent, central pressure dropped by 14 percent, and, most dramatically, precipitation increased by 18 percent near the storm center (due to the fact that in a warmer climate, the air converging into a hurricane holds more water vapor). This translated into a shift toward stronger storms by “half a category” on the Saffir-Simpson scale, hardly trivial considering that hurricane damage levels spike upward with increasing storm intensity. And the strengthening would surely have been greater but for another partially offsetting effect of climate change: The carbon dioxide increase in the models had the effect of producing an enhanced warming of the upper troposphere relative to the surface in the tropics. This in turn increased atmospheric stability and made it somewhat harder for air to rise in the hurricane eye wall. Finally, the study found that although sea surface and upper tropospheric temperatures rose consistently in the climate models, changes in vertical wind shear were far more erratic. Knutson and Tuleya thus questioned whether this key dynamic factor would shift under global warming in such a way as to systematically affect hurricane intensity. It looked instead like the most important expected changes would be caused by thermodynamics.

All of this seemed to greatly clarify how global warming might change hurricanes. But the upshot for the present moment was simply that there wasn’t one. The study made it explicit: Due to their relatively modest magnitude, the carbon dioxide-induced changes to hurricanes “are unlikely to be detectable in historical observations and will probably not be detectable for decades to come.” At least according to this work, the effect of global warming on hurricanes seemed a problem for the future. It certainly couldn’t be blamed in any meaningful way for the storms in Florida.

Landsea, however, thought even this overplayed the likely effect of climate change. Before long he teamed up with longtime global-warming contrarian Patrick Michaels to critique the Knutson study. They began by posing a question—“Should we trust models or observations?”—and went on to answer in defense of the latter. The Knutson and Tuleya work, they argued, was too idealized, too simplistic. Even its modest results for hurricane intensification were overstated because “observations, rather than models” suggested a more tenuous relationship between sea temperatures and hurricane strength.

In the face of these criticisms, Knutson and his coauthor stood firmly by their study and defended the assumptions they had used to study changing storm intensities in a changing world. “If we had observations of the future, we obviously would trust them more than models,” they dryly responded. “But unfortunately observations of the future are not available at this time.”

 

Landsea could counter Knutson’s position in the peer-reviewed scientific literature by writing a critique. That’s just what he eventually did. But when a group of climate scientists went straight to the media, staging a press conference to make an even stronger linkage between hurricanes and global warming in October 2004, they couldn’t be countered in the same way.

When Landsea heard about the upcoming Harvard press conference, he contacted Kevin Trenberth, one of the listed participants, and asked him to rethink. “I said, please don’t do this, because there’s no science behind such a link,” Landsea recalls. But Trenberth—a short, thin, mustached scientist who originally hails from New Zealand and is one of the most widely cited researchers in the climate field—did the event anyway. In fact, during it he stated: “I think one of the reasons we’ve got this press conference is to perhaps try to add a little bit to other statements that have been made by hurricane forecasters.”

Trenberth heads the climate analysis section of the National Center for Atmospheric Research in Boulder, Colorado, and specializes (among other areas) in studying how global warming will change the global water cycle and the character of precipitation in its various forms, ranging from snowfall to the intense downpour generated by hurricanes. It was largely from this vantage point that he approached the hurricane question. In a warmer world, Trenberth reasoned, the atmosphere ought to hold more water vapor, which in turn should bring about stronger precipitation in storms. Warmer seas should also evaporate more water, providing more fuel for the hurricane heat engine. In short, Trenberth was largely relying on basic physical reasoning about how the climate system responds to change on a large scale. On the Harvard conference call, he was accompanied by Harvard biological oceanographer James McCarthy, an expert on the impacts of climate change; Dr. Paul Epstein, associate director of the sponsoring Center for Health and the Global Environment; and Mathias Weber, a senior vice president at the U.S. division of the reinsurance company Swiss Re, a Europe-based firm that has been very open to the notion of a possible connection between climate change and an increase in insured losses due to extreme weather.

In his remarks—which later became the source of fierce controversy—Trenberth did not claim that global warming had caused the 2004 hurricane destruction. However, he sought to explain that, as global warming is clearly happening, it’s unlikely it would not be having an effect on hurricanes. “We can’t say anything really about the tracks which make the hurricanes hit the U.S. or miss the U.S.,” Trenberth said. “What we can say is that the high sea-surface temperatures [and] water vapor make for more intense storms and so this is consistent with the evidence that we’re seeing.” In a press release announcing the call, Trenberth was further quoted as stating that “the North Atlantic hurricane season of 2004 may well be a harbinger of the future.” Mathias Weber also included this cautionary remark (which Trenberth echoed): “We believe it is actually impossible to associate a single event such as Hurricane Charley, or even a series of events such as the series we have seen this year, to a climate change trend.”

Yet despite such caveats, Trenberth’s statements did go beyond where the hurricane-climate debate had wound up by the close of the 1990s by suggesting that a discernible influence on storms already existed. For example, Trenberth called Catarina “the first of its kind and [clear] evidence that things are changing.” Citing the Atlantic hurricane increase since 1995, he added that “this kind of evidence is pointing more in the direction that these extremes are occurring and are having a real impact on society.” So if Trenberth’s remarks were later misinterpreted as suggesting a direct causal connection between global warming and the Florida hurricanes, that’s partly understandable in light of how the Harvard press conference—both by its timing and its theme—broadly “linked” the two.

In contrast to Trenberth, Emanuel at this point in time could be found expressing his own position with stronger reservations, and hewing more closely to the 1998 consensus. While still convinced global warming ought to change hurricanes, Emanuel added in a 2004 article that no such change had manifested at the present moment—nor should it have, in light of an average increase in tropical sea-surface temperatures of only .3 degrees Celsius (about half a degree Fahrenheit) since 1950. Using his maximum potential intensity theory, Emanuel calculated that storms should only have strengthened by about 1.3 percent, too small an increase to be detected given large natural variability and the error margins inherent in storm intensity measurements. “I think it’s extremely difficult to pin the last season on global warming,” Emanuel later said of the 2004 destruction. “That does not preclude that there may be a global-warming signal buried in there somewhere, but nobody in my field thinks that we’ve seen it.”

Knutson’s study on hurricane intensification under global warming also argued that the phenomenon would be undetectable for years. But Trenberth suspected that such modeling studies were probably underestimating the sensitivity of hurricanes to climatic changes. Meanwhile, at the Harvard event, Paul Epstein went even further, implying a direct link between global warming and the deaths caused by Jeanne in Haiti. No wonder that in response to such statements, Bill Gray fired back in his typical fashion. “They are all smart guys—I admire them for their talents,” Gray told one reporter. “But on this topic, I feel many of them have sort of sold their soul.”

Heightening tensions further was the 2004 presidential election that lay on the horizon. No one knew what effect the devastation in Florida might have on the outcome in this most pivotal of states, which had twenty-seven electoral votes compared with twenty for the other key swing state, Ohio. Certainly it was enough of a wild card that both George W. Bush and, later, John Kerry made a point of personally surveying the damage and offering support to those whose lives had just been shattered. In this context, talk of global warming and its impacts on hurricanes could easily be parlayed into an attack on Bush, who had struggled on the climate issue during his first term in office, and whose administration had been repeatedly accused of altering and suppressing scientific information related to climate change and its impacts.

Just a few days after the Harvard press conference, this political subtext ceased to be a subtext at all when two advocacy groups called Environment 2004 and Scientists and Engineers for Change announced they would be displaying billboards along the 1–4 corridor linking Tampa and Orlando, as well as in the cities themselves. “Global Warming = Worse Hurricanes,” the billboards read. “George Bush just doesn’t get it.” Alongside this slogan, the billboards depicted a hurricane barreling toward the shell-shocked Florida peninsula.

 

Billboard put up along highways in Florida during the 2004 presidential election by the groups Environment 2004 and Scientists and Engineers for Change. Credit: Environment 2004.

 

Judged from the standpoint of political advertising, this certainly constituted a gripping message. It was also a direct, if unsuccessful, attempt to sway an election. Science and politics had merged inextricably in this instance, and that had a serious consequence for the hurricane-climate debate going forward. Scientific divides would harden, while the Bush administration in its second term would dismiss claims that climate change had intensified or might intensify hurricanes, even when some of those claims came from the administrations own scientist employees.

Whatever the impact of the Scientists and Engineers for Change ad campaign in Florida, it did not manage to win the state for John Kerry—we all know how the 2004 election turned out. But the Harvard press conference would have serious repercussions nonetheless.

 

Shortly prior to the Harvard event, Trenberth had invited Chris Landsea to contribute to a section of the Intergovernmental Panel on Climate Changes Fourth Assessment Report, due out in 2007, which addressed Atlantic hurricanes. The IPCC process marshals the collective contributions of hundreds of global scientists in a loosely hierarchical format, and Trenberth, relatively high on the totem pole, was “coordinating lead author” of his section, a job that entailed getting the worlds foremost experts to contribute. Landsea was inarguably an expert on hurricane data, and he had accepted Trenberth’s invitation.

But several months after the press conference, Landsea publicly resigned from the IPCC process, e-mailing out a denunciation of Trenberth’s Harvard statements in the form of a letter to forty-five colleagues that then made its way onto the Internet. “It is beyond me why my colleagues would utilize the media to push an unsupported agenda that recent hurricane activity has been due to global warming,” Landsea wrote. Later he added: “I personally cannot in good faith continue to contribute to a process that I view as both being motivated by pre-conceived agendas and being scientifically unsound.” Landsea also questioned whether the scientists involved in the Harvard event had any business speaking about hurricanes: “To my knowledge, none of the participants in that press conference had performed any research on hurricane variability, nor were they reporting on any new work in the field,” he observed.

If Landseas resignation did not immediately follow the Harvard event, it was because first he’d sought internal reassurance that when it came to hurricanes, “what will be included in the IPCC report will reflect the best available information and the consensus within the scientific community most expert on the specific topic.” When Landsea complained to the IPCC leadership about Trenberth, however, he did not find a particularly sympathetic response. “Individual scientists can do what they wish in their own right, as long as they are not saying anything on behalf of the IPCC,” chairman Rajendra Pachauri had written to him. But Landsea protested that Trenberth had been identified as affiliated with the IPCC at the Harvard press conference. It was only after the IPCC did not deal adequately with his concerns (in Landseas opinion) that he chose to resign.

After Landseas open letter appeared in January 2005, the spat between him and Trenberth drew major media coverage. Trenberth called Landseas charges “ridiculous” in one interview, but said he would welcome him back to contribute to the IPCC report. Landsea said he would work with the IPCC again, but not on a section that had Trenberth as a coordinating lead author. Perhaps the most cutting remark in the conflict came from Landsea in an e-mail later made public:

 

The sad thing about this is that it did not have to turn out this way. I did try to caution [Trenberth] before the media event [and] provided a summary of the consensus within the hurricane research community . . . Dr. Trenberth wrote back to me that he hoped that this press conference would not “go out of control.” I would suggest that it was out of control the minute that he and his fellow panel members decided to forego the peer review scientific process and abuse science in pursuit of a political agenda.

 

In the context of the ongoing global warming debate, the dispute between Trenberth and Landsea had significant political implications. With each of its successive five-year assessment reports, the IPCC had strengthened the conclusion that global warming is happening due to human activities. Meanwhile, each time the climate “skeptics” had sought to undermine the panel’s credibility and objectivity. In this sense, Landseas resignation represented a public relations bonanza to them. He had handed over a big stick with which to beat the IPCC. Among many others, Republican senator James Inhofe of Oklahoma, a leading global warming “skeptic” in Congress, would later cite the Landsea-Trenberth episode as a “stark example of how the [IPCC] process has been corrupted.”

Advocates generally took either Landseas or Trenberth’s side in the dispute according to their political predilections. What few recognized, however, was that at the heart of their argument, once again, lay a split between scientific approaches. On the one hand, in his remarks on hurricanes Trenberth had spoken from the perspective of a climate scientist interested in how the overarching system works. Given that global warming is happening, he reasoned, how could that not be having at least some effect on storms reliant upon heat and moisture? But Landsea, speaking from the perspective of an Atlantic hurricane analyst and a master of storm data, declared that no evidence could be adduced to demonstrate a “long-term trend up in the frequency or intensity of tropical cyclones.” On at least one level, Trenberth and Landsea weren’t so much disagreeing as speaking past each other. Had their debate unfolded through the scientific process, rather than through the amplifying, oversimplifying, and often distorting medium of newspaper and other media reports, they might well have been able to find some common ground.

After the dust-up with Landsea, Trenberth talked to the press somewhat, but did not put out a public statement of his position to rebut critics in the media or on the blogs (which were beginning to play a pivotal role in shaping discourse about hurricanes and climate). Instead, he began working on a scientific article detailing the reasoning behind his statements. The paper, which appeared in Science in June 2005 with the hurricane season already underway, laid out the basic logic. After citing the 2004 Florida storms and Japanese typhoons, and the global-warming questions that had been raised in relation to them, Trenberth spent most of the paper on caveats. Sure, there’s lots of hurricane variability, within and between ocean basins. And sure, hurricanes are affected by anything from El Niño to vertical wind shear. So far, so good.

But nevertheless, Trenberth continued, no one can deny that sea-surface temperatures are rising and that rise has been linked to human-induced global warming. Further, the warming means there is more water vapor in the atmosphere. These two trends should theoretically provide more energy to fuel hurricanes and increase their rainfall. In the face of such changes, Trenberth wrote, it would be surprising if there weren’t changes to hurricanes as well, even if no trends had been firmly documented yet. Such a situation often occurs in science, he observed: Researchers lack the data to show that something is truly happening that they suspect ought to be, based on physical reasoning. “Although variability is large,” Trenberth wrote, “trends associated with human influences are evident in the environment in which hurricanes form, and our physical understanding suggests that the intensity of and rainfalls from hurricanes are probably increasing, even if this increase cannot yet be proven with a formal statistical test.”

It was a fascinating piece of scientific speculation. The tide of Trenberth’s article—“Uncertainty in Hurricanes and Global Warming”—underscored as much. Trenberth had advanced a scientific hypothesis based on fairly simple physical reasoning, but he had no body of hurricane data to support him. He was also going beyond existing hurricane modeling and theory, as represented by the work of Knutson and Emanuel, which agreed that hurricanes ought to change under global warming but not enough for the change to be detectable yet.

“I was out there by myself,” Trenberth remembers. But if he’d gone out on a limb, his statements and his paper also challenged other scientists to dig into global data on hurricanes, in order to determine whether to saw off his support or to clasp his hand.

 

Meanwhile, even as the Trenberth-Landsea controversy rippled across the media and blogs, another cluster of intense tropical cyclone activity hit a part of the world vastly distant from both Japan and Florida.

In the space of a mere month in February 2005, two Category 4 and two Category 5 storms—Meena, Nancy, Olaf, and Percy—tore through the Cook Islands in the Southwest Pacific basin, to the east of Australia and northeast of New Zealand. Collectively, these cyclones trashed Raratonga, Tokelau, American Samoa, and numerous other small islands. The worst of them, Percy, had a record low central pressure for the region of 900 millibars, and reportedly damaged nearly all standing structures on Pukapuka and Nassau, home to 670 people.

Just two years earlier, Cyclone Zoe had ripped through the Solomon Islands, also in the Southwest Pacific basin, with winds approaching 180 miles per hour and a satellite estimated minimum central pressure of 879 millibars—which, if accurate, would have made it the strongest known Southern Hemisphere hurricane. And now, according to the Cook Islands Meteorological Service, so many intense cyclones appearing in the area, in such a brief period of time, represented yet another “record.”