Title: a climate Conundrum. (cover story) Authors

NO ONE FORESAW THE SHEER magnitude of tropical devastation that would paralyze the United States' Gulf Coast during the fall of 2005. In 2004, Florida's Ferocious Four--Hurricanes Charley, Frances, Ivan, and Jeanne--seemed to pummel every square mile of the Florida peninsula and panhandle, to the tune of at least $42 billion in damages. In early 2005, hurricane scientists were well aware that the Atlantic was still in the midst of a high cycle of storm activity, and forecasters predicted another busy hurricane season.

But how much worse could the situation get than in 2004, when four significant storms made landfall in a single state?

The 2005 Atlantic hurricane season broke a substantial number of records: most storms in a single season (27); highest number of Category 5 storms (three--Katrina, Rita, and Wilma); lowest surface pressure (Wilma, 882 mb); earliest season storm activity (five named storms in July, with Dennis and Emily as Category 4); and possibly an Atlantic record for most rapidly deepening storm (Wilma, 95 mb in 24 hours). Bright red splotches, representing steamy tropical ocean surface waters, dominated the evening newscasts.

Has the tropical atmosphere truly "gone mad," with tempest after tempest erupting in a seething hotbed of activity across the western Atlantic and Gulf of Mexico? Is global warming to blame? Have we finally crossed some type of "climate threshold" into the era of hypercanes and the doomsday maelstrom?

The hurricane scientist online chat group has been abuzz; we hurricane nuts can be quite excitable at times. But there is also a quiet restraint, a voice of reason that threads its way through this group's musings. While there is much speculation and debate on the causes of the vicious 2005 hurricane season, the climate machine is not broken. Global warming itself is not the sole cause of the unusually active 2005 hurricane season. Natural factors, and an element of chance, need to be considered.

Rhythms of the Tropics

Cycles of natural variation control rainfall and storminess throughout the global tropics. Many of these are wave phenomena, with names like "Kelvin Waves" and "Rossby Waves." There are also ocean-atmosphere rhythms, such as the North Atlantic Oscillation (NAO), the Pacific Decadal Oscillation (PDO), and the El Nifto Southern Oscillation (ENSO). Some of these control rainfall and even the birth of tropical cyclones on time scales of a few weeks to months; others, such as ENSO, operate on the scale of years and even decades, and determine whether a given year will experience drought or flood.

But one of the most important internal rhythms of the tropics is the Atlantic Multi-decadal Oscillation (AMO), a cycle that is tied to the frequency and intensity of rainstorms over tropical west Africa, where the wave disturbances of many U.S. landfalling hurricanes are born. The AMO is associated with periods of anomalous warming and cooling of surface water in the Atlantic Ocean. It is even possible that hurricanes in the Atlantic help drive the deep ocean currents that change these ocean temperatures.

The number of hurricanes in the Atlantic waxes and wanes on a coherent decadal timescale. A span of decades with relatively few storms, the low phase, alternates with several decades of heightened activity. Since about 1900, a major upswing in the number of storms occurred in the late 1920s through 1960s, followed by a period of low activity in the 1970s to early 1990s. In 1995, as if a switch had suddenly been thrown on, the Atlantic began lighting up with storms. In every year since 1995 (with the exception of 1997) the number of storms exceeded the long-term average of 8 to 10 storms per year. A total of 21 named storms occurred in 1995 and 2003, and between 13 and 17 developed during the intervening years.

It is likely that the superabundant seasons in 2004 and 2005 are in large part a reflection of the recent active phase of the AMO. But we do not have the data that reveal the behavior of this cycle much earlier than 1900. Some researchers have examined tree rings and African lake bed sediments, and their work suggests that the AMO has been operating for centuries. But the jury is still out on just how long-term and robust a cycle the AMO really is.

ENSO is a potent modulator of tropical cyclones in the Atlantic basin. When El Niño is active and strong (an event that occurs perhaps once every five years or so), the number of Atlantic hurricanes drops. This was the case in 1997, one of the strongest El Niños on record, when only eight named storms developed in the Atlantic. This dramatic reduction occurred in the midst of a very dominant AMO phase. From 1925 to 1997, for example, out of 24 hurricane events causing at least $1 billion in damages, 17 occurred in non-El Niño years. While El Niño is a process uniquely endemic to low latitudes of the Pacific, its influences flow into the Atlantic. It is likely that the El Niño wind circulation increases the amount of wind shear, which is deleterious to the health of embryonic hurricanes.

Hurricanes and Climate Change?

Hurricanes are a relatively rare breed of tropical disturbance, averaging roughly 80 to 90 a year around the globe. That is a tiny number compared to the estimated number of tropical thunderstorms active at any one time somewhere around the globe--a number that approaches 2,000. The total number of tropical cyclones has remained remarkably constant for many decades, seemingly immutable to the aforementioned cycles and any type of detectable longer term climate trend. Worldwide, and over time, there seems to be an equilibrium; when storms develop in one ocean basin in any given year, the numbers fall off in other basins.

Recently, however, several peer-reviewed scientific papers have addressed hurricanes and global warming in prestigious journals. One of these papers, by Kerry Emanuel in Nature, presents evidence that the energy dissipated by hurricanes has doubled since the mid-1970s in the Atlantic and western North Pacific basins. The author asserts that storms have become stronger and/or more long-lived in these basins. Another paper by Peter Webster and colleagues, in Science, shows that while the number of Category 1, 2, and 3 storms has remained nearly constant worldwide, the number of Category 4 and 5 monsters has doubled over the past 35 years. What is interesting is that the increase in intensity/duration of the Atlantic and Pacific storms coincides with an abrupt increase in ocean surface temperature of about 0.5 Celsius. Webster feels that even when accounting for natural cycles such as ENSO and AMO, the correlation between the upswing in hurricanes and ocean temperature in the past 30 plus years can be explained by global warming.

At first blush, it seems tenable that as ocean temperatures rise in a human-induced global warming world, hurricanes would become more powerful. After all, the oceans are the fuel tank, water vapor is the fuel, and the water temperature loosely corresponds to an octane rating. In more precise terms, as water temperatures rise by only a small amount, the amount of water vapor evaporated increases exponentially. However, many factors other than ocean temperature influence the intensity of hurricanes, such as the strength of wind shear and the amount of moisture available in the middle layers of the atmosphere. How these variables will change and interact in complex ways with the thunderstorms that give rise to hurricanes in an enhanced greenhouse world remains uncertain.

Additionally, some scientists acknowledge that while global warming may already exert some degree of influence on the intensity of hurricanes, the change is barely detectable, an increase of just a few percent. The full impact of global warming may require decades to create any noticeable impact. The changes that may be occurring are still very small relative to large, natural "mood swings," for instance, of the AMO.

The unusual destructiveness of the 2004-2005 hurricane season also has much to do with where the strong storms were being steered. Since 1995 there have been other years with many strong storms, but these generally re-curved over the Atlantic, east of the United States mainland. In 2004 and 2005, most of the powerful storms intensified closer to Florida and the Gulf of Mexico, where they were repeatedly propelled inland by the larger flow. You can partially blame the past two years on bad luck, after a string of many active Atlantic basin years in which the United States has been extremely fortunate.

There are many measures of hurricanes, including intensity, strength, duration, size, and number. You can further break these down, basin by basin, or address them on a global scale. A few studies have begun to address some of these characteristics, specifically in changes that are seen in some locations but not in others. There are many cycles operating within cycles that control tropical storminess; trying to sort out cause-and-effect among the oscillations and detecting steady trends (such as global warming) can be daunting. Our record of reliable data is quite short and there are many different types of datasets that must be carefully related to one another and cross-calibrated. For instance, hurricane intensity has been gauged by aircraft penetrations in some years, while in others satellites have been used to estimate intensity. Do these vastly different techniques of measurements provide equivalent results? Also, changes in weather and changes in climate need not reflect the same underlying physical processes. A single fierce tropical cyclone, or even one or two seasons' worth of especially virulent activity in one basin, are discrete events. These are distinct from longer holistic trends, such as global warming, which play out gradually over decades and simultaneously across many ocean theaters.

Why Was 2005 So Bad?

In the wake of a record-breaking storm season, scientists are eager to uncover the reasons for 2005's superabundance of hurricanes. Against the backdrop of the continued high phase of the AMO, there was no El Niño during 2005. Destructive wind shear was weak throughout the Gulf of Mexico, Caribbean, and Atlantic. Storms repeatedly developed healthy outflows in the upper atmosphere that were not beaten back by strong winds. (The more air that can be exhausted from the top of a hurricane, the more mass can rise up from below, allowing the pressure near the surface in the eye to fall.) There is emerging evidence that the general sea level pressure throughout the Caribbean region was unusually low, a condition that became established months in advance of the 2005 summer season. Such a persistent region of low pressure has been found to favor the formation of Atlantic hurricanes.

It is likely that the ocean also played a key role in the rapid and sustained intensification of many of this year's storms. Oceanographers have suggested that the waters in the Gulf of Mexico were unusually warm--not just on the surface but down through a very deep layer, particularly within the large clockwise-spinning gyre of a current called the Loop Current. There was plenty of stored energy, such that the effects of one storm mixing up cold water from the ocean depths would not create a long-lived cold wake that would weaken the next storm to follow. In fact, the tracks of Katrina and Rita moved along the axis of the Loop Current for many days, rather than quickly crossing over it at a sharp angle--a chance alignment that allowed these storms to draw the maximum amount of energy for an extended duration, contributing to their virulence. With such a synergistic coupling between the storm and its energy supply, and in the absence of wind shear, there really was nothing to hold back these monsters from pegging the intensity meter at maximum. Perhaps this is further evidence that 2005's bounty of highly destructive storms boils down in part to just plain bad luck.