Quotes of the Day

It's no accident that the documentary An Inconvenient Truth opens with a satellite image of Hurricane Katrina bearing inexorably down on a helpless New Orleans. Since hurricanes draw their destructive power from heat in seawater, you would expect that global warming would intensify these terrifying storms and multiply their number, leading to increased devastation on land. All other things being equal, that's probably what would happen.

But all other things are not equal, which is why the relationship between climate change and hurricanes is anything but settled. Two new studies released this week have moved the ball significantly forward, however. The first, appearing in Science, says the frequency of Atlantic hurricanes will actually decrease during this century but that the most powerful Category 4 and 5 storms will likely double in number.

The second study, presented at a conference of the American Meteorological Association in Atlanta, says that whether or not the nature of hurricanes changes, the property damage they wreak in the U.S. will rise an average 20% over the next two decades because of the rising sea level caused by global warming.

The idea that there might be fewer Atlantic hurricanes in a warming world is not a new one. Earlier studies led by Thomas Knutson of the National Oceanic and Atmospheric Administration's Geophysical Fluid Dynamics Laboratory (GFDL), a co-author of this week's Science paper, had reached the same conclusion. The explanation for this seemingly paradoxical finding: hurricane activity is governed not only by ocean temperatures, but also by factors such as ocean currents and the speed and direction of wind in different layers of atmosphere. It turns out, says Knutson, that the key to hurricane frequency is not simply how much the Atlantic warms, but how it warms in relation to the rest of the tropical oceans. If it warms more than average, he says, you have an increase in storms. If it warms less, you have a decrease. The rise in Atlantic hurricanes that we have seen since 1980 or so, he says, is probably the result of exactly that kind of differential warming, not so much global warming overall.

The significance of the current paper, says GFDL's Morris Bender, the lead author, is that it simulates the statistics of the most severe future hurricanes — something previous studies couldn't do because of an inability to accurately reproduce a hurricane's structure. A new modeling approach used in this week's study remedies that problem, he says, and suggests that Category 4 and 5 storms will become relatively more common by 2100 — with the important caveat that the change will not become clearly detectable until the second half of the century. The locus of the biggest increase, continues Bender, is projected to be in the western Atlantic, north of about 20 degrees latitude — about where the southern edge of Cuba lies. And because these stronger storms are so damaging, the increase in their number would likely outweigh the decrease in overall numbers, in terms of destruction.

The authors warn that their findings, while important, are not the final word and must be validated by others. Bender, Knutson and their colleagues ran the simulation using an average of 18 climate models at once, then reran it with four models individually. One of the models, Bender says, showed a decrease in the most powerful storms, demonstrating once again that the effects of climate change are complex and not easily predicted.

While the Science paper looked at hurricane activity and strength, the second paper looked at an entirely different question. Much of a hurricane's destructive power comes not from winds or rain but from the bulge of seawater it pushes ahead of it and crashes into shore. If sea level rises, these so-called storm surges become more damaging. In order to put a dollar figure on how much more damaging, a group of scientists looked at climate models, hurricane databases and so-called catastrophe models that evaluate the potential destruction of storms in specific places. Then the researchers wove the information together to project what hurricane-driven storm surges might cost Americans 20 years from now.

Before talking about the results, the authors emphasize that they are not making an actual forecast. Their study is more of a thought experiment that includes a long list of necessary oversimplifications. "We're not projecting changes in population, property values, building codes or zoning regulations," says lead author Ross Hoffman of Atmospheric and Environmental Research Inc., a private firm that does climate and ocean modeling, among other things, for companies and government agencies. "We're simply asking, If nothing else changed but sea level, what would the effect be on property damage?"

Even oversimplified, it's an extremely complicated thing to figure out. Take sea level, for example. Most studies on climate change talk about the average rise worldwide. But things can look very different when you zoom into specific stretches of coast. Ocean currents can make local sea level higher or lower than the world average. So can the continuing rebound of land from the weight of glaciers from the last ice age, even though they melted more than 10,000 years ago. Factors like the extraction of oil and gas, like in the Gulf of Mexico, can also make the shore slump.

Take into account property damage, and things become even more complex. The amount of destruction depends on the density of development, the mix of commercial and residential property, various construction characteristics and the location of a building's contents, says co-author Peter Dailey of AIR Worldwide, which specializes in catastrophe modeling. Surge damage typically affects lower floors; in a single-story warehouse, for example, most of the contents are at risk, while in a hospital they would be distributed over many floors.

Furthermore, how far a storm surge penetrates inland depends on the local topography, both offshore and on — damage to property on a seaside cliff is a lot different from that on a beach. And even if property lies at a generally low elevation, it makes a difference whether the region is hilly or flat, or close to a bay or estuary.

What makes this study unique, says Dailey, is that it takes all of these effects into consideration. The result is a map of the Eastern U.S. seaboard that calculates plausible, albeit oversimplified, numbers for the increased damage that storm surges could cause. An overall round number, says Hoffman, is a 20% increase. But it would be higher in some areas — as much as 100% — and lower in others. In general, in a place like Louisiana, which is already at high risk, the increase would be smaller, and vice versa for places like Long Island and Connecticut, where the relative rarity of hurricanes makes today's risk to property low, but the potential increase very significant.

If some of the study's oversimplifications turn out to be wrong, as they well may — if hurricane activity increases or sea level begins rising faster than expected or a new real estate boom puts more property in harm's way — the damage estimates could rise.

Neither of the new studies is the final word, as the authors plainly acknowledge. In both cases, though, our ever-evolving understanding of what climate change will look like has become a little clearer.

Lemonick is the senior science writer at Climate Central.

• Michael D. Lemonick
• You would expect that global warming would lead to an increase in hurricanes, since hurricanes are fed by warm sea water, but you would be wrong