Katrina and the waves: the row over hurricanes

A year after New Orleans' devastation, opinion is divided over why hurricanes have become so destructive. Andrew Buncombe meets the experts locked in a stormy row

It was late afternoon inside the low-ceilinged room on the edge of Miami and a bank of computer monitors was showing a mass of throbbing colours - green and blue and baby-sick yellow - steadily march northeastwards across the upper half of Florida.

This swirling mass was Alberto, the first tropical storm of the 2006 season, and it had already moved across the western Caribbean and the Gulf of Mexico, deluging Cuba and Grand Cayman with massive rainfall and winds of around 70mph. Tracked by the experts from the National Weather Service's Tropical Prediction Centre, it would ultimately head up over Georgia, North Carolina, and then Virginia before its energy dissipated and finally fade the following day.

Hunched over a telephone in the centre of the room, the receiver pushed into his ear as he sat in what appeared to be painful concentration, forecaster Richard Pasch was involved in a conference call with colleagues across the region also monitoring Alberto's course. It was clear that Alberto's power was already falling and there was a brief discussion as to whether it should be downgraded from a tropical storm. In the end - as the official advisory issued at 5pm that afternoon showed - they decided to maintain the storm-warning for a few hours longer and issued the following troubling statement: "There is a threat of flooding due to locally heavy rain over portions of the southeastern United States."

The prediction of tropical weather systems is a vast and complicated matter. After the 2005 hurricane season - the busiest on record - never before has so much attention been given to the warnings and estimates provided by forecasters about the likely strength and course of the storms that batter the Caribbean and the south-eastern seaboard of the US. With the numbers of people moving to coastal communities increasing, never has so much money and so many lives been at stake. Never has the need been greater to get those predictions correct.

And yet never before has the world of tropical weather prediction been riddled by such heated controversy and disagreement - a controversy with huge implications for anyone fearful as to the impact of such tropical systems in the future. The debate is part of a broader discussion about the extent and implications of climate change and in essence it is about whether storms are getting stronger and more powerful as a result of man-made global warming. Some say there is no convincing evidence of such a claim while others say the evidence is obvious and clear for those prepared to look. Both sides are equally convinced they are correct.

"It's the coffee table topic of the moment," said Gavin Schmidt, a climate modeller at NASA's Goddard Institute for Space Studies and moderator of RealClimate.Org, a website run by climate scientists. "The big problem is that this is a big issue and there are some interesting scientific questions, but because this is a brand new issue the work has not yet been done that is required to reconcile all the different ideas."

It was a year ago that a storm system called Katrina made us all start thinking about hurricanes in a very different way. Katrina was a category five "storm" that had weakened to category three by the time it struck the Gulf coasts of Louisiana and Mississippi on the morning of 29 August. Its storm surge led to the deaths of more than 1,800 people and caused massive, extraordinary damage estimated at around $80bn. The day before the storm struck, Ray Negin, the Mayor of New Orleans belatedly ordered an evacuation of his city. "We are facing a storm that most of us have long feared," he said. "This is a once-in-a-lifetime event."

In the immediate aftermath of Katrina I spent a week in New Orleans, witnessing the efforts to evacuate its stranded and beleaguered residents, witnessing the stinking, wretched conditions of those people forced to seek shelter in the city's Convention Centre and Superdome. It was an unsettling experience and a troubling time that exposed some of the faultlines that run through American society; for every example of human kindness or largesse of spirit there would be another example of humanity's ability for meanness and cruelty. There was generosity alongside thoughtless racism, boundless community spirit rubbed shoulders with selfishness.

There was also widespread awe that an act of nature could have caused such devastation. Among the stinking flooded homes and buildings, amid a city brought to its knees by the storm, it was not hard to find people voice the opinion that something was out of kilter, that something unusual had befallen. This, after all, was a city and a region that was used to hurricanes. The afternoon drinkers in Johnny White's pub on Bourbon Street would brag that the bar had never shut, not even when Hurricane Betsy tore into the city back in 1965. The question was soon being asked in the media and by the people of New Orleans - could global warming be to blame?

Among the science cited to back such a claim was a report in Nature magazine by Professor Kerry Emanuel, a climatologist from the Massachusetts Institute of Technology (MIT), which argued that the strength of hurricanes had increased in recent years and that this was linked to climate change. Although sea temperatures had only increased by around half a degree over the last 30 years, the destructive power of hurricanes had doubled in that period, he said.

In support of his claim, Emanuel had gathered statistics dating back to 1930 relating to the power of hurricanes and - having made adjustments to counter what was widely considered an inaccuracy in some earlier measurements - worked out a figure to measure their annual destructive power, which he called the power dissipation index (PDI). He then set these figures against data showing the average September sea surface temperature for each of those years.

He claimed to have discovered a remarkable link between the two. Furthermore, he showed that in the last 30 years storms were lasting longer and were more intense. He wrote: "My results suggest that future warming [of the oceans] 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 21st Century."

Emanuel was not the only person making such claims. In 2004 - which until last year was the busiest hurricane season on record - George Trenberth, of the National Centre for Atmospheric Research in Boulder, Colorado, had argued in another paper that "trends in human-influenced environment changes are now evident in hurricane regions". He added: "These changes are expected to affect hurricane intensity and rainfall but the effect on hurricane numbers remains unclear. The key scientific question is how hurricanes are changing."

A third piece of evidence came in the form of another study, this one by Peter Webster and Greg Holland, who said their examination of data from 1970 onwards suggested the number of Category 4 and 5 hurricanes worldwide had nearly doubled over the past 35 year, while sea temperatures had continued to rise. "Our work is consistent with the concept that there is a relationship between increasing sea surface temperature and hurricane intensity," said Webster. "However, it's not a simple relationship."

As the extent of the devastation from Katrina became clearer, the British government's chief scientist, Sir David King also entered the debate. "We have known since 1987 the intensity of hurricanes is related to surface sea temperature and we know that, over the last 15 to 20 years, surface sea temperatures in these regions have increased by half a degree centigrade," he said. "So it is easy to conclude that the increased intensity of hurricanes is associated with global warming."

Take your pick - there is no difference between a hurricane, a typhoon or a cyclone. Rather, they are varying names for the same tropical weather systems created in, respectively, the Atlantic and eastern Pacific oceans, the western Pacific ocean and the Indian and southern Pacific oceans.

These storms begin as areas of low pressure within a band of the tropical oceans known as the Intertropical Convergence Zone (ITCZ) or, more romantically, the Doldrums. In the simplest of terms they form like this: hot damp air rises from the ocean, spirals, cools and then condenses. Rain falls, energy is released and the air is warmed - by this means reinforcing the updraft. An area of low pressure develops and moisture-laden air rushes in at an angle determined by the rotation of the earth.

The spiral spins increasingly faster and draws in air from the upper atmosphere. This air is warmed by compression and adds to the heat pool. Soon the system has momentum and mass of its own. Powered by heat from the sea, these systems are then steered by winds as well as by their own energy. Around their core, winds grow with great velocity, generating violent seas. When they hit land they can sweep the ocean inwards with them while creating tornadoes and producing torrential rains and floods. It had been the storm surge of Katrina that destroyed the levees of New Orleans.

There is a musicality to the naming of these storms that belies their extraordinary and terrible destructive power. In Heaven's Breath, his natural history of winds, Lyall Watson writes: "In the North Atlantic they are called hurricanes in memory of the Mayan storm God Hunraken; in the Indian Ocean they have been known as cyclones from the Greek kuklos meaning circular, ever since the President of the Marine Courts at Calcutta first used this term in the middle of the 19th Century; in the China Sea they call them typhoon, from ty fung or "great wind"; in the Philippines it is baguoi; in Japan reppu; and asifa-t in the Persian Gulf."

Tropical weather has its own precise set of standards for classification. In the North Atlantic, the area of tropical weather most studied, a tropical depression, which has no "eye", is a weather system with sustained winds of no more than 38mph. A tropical storm is a system with maximum sustained winds of 39-73 mph. Above this the system becomes a hurricane, graded 1-5 on the Saffir-Simpson scale, according to their strength and intensity. The strongest, a category five hurricane, has sustained winds of more than 155mph.

The sheer might of a hurricane is belittling. Researchers have found that a typical storm can release more than ten million trillion joules of energy - the equivalent of around one million Hiroshima bombs - in a single day. Most of this energy is ultimately dissipated upwards.

Even being on the edge of a modest hurricane is thrilling, if slightly frightening, experience. In September 2004 I was in West Palm Beach in Florida when Hurricane Frances, a vast, swirling category two storm the size of Texas, swept through. The first signs of the storm's approach came at around 10am on that Saturday morning as power and electricity lines were knocked out and it gradually built during the day, with the sky growing darker and darker. By 7pm winds of more than 70mph were bending trees and hurling debris.

The worst of the hurricane came between midnight and 4am when the centre of the eye passed a few miles to the north. I looked out of my leaking hotel window and listened as the storm howled and roared, driving sheets of rain at all angles. Palm trees were bent almost double by the force of the wind. Five people in Florida and one in the Bahamas lost their lives as a direct result of the storm while it was also indirectly blamed for 42 more deaths.

Why Frances? Why Katrina? Why Hurricane Mitch, which caused such terrible devastation in Honduras back in 1998? Such names come from a pre-prepared list of alternate male and female names running through the alphabet, missing out the letters Q,U,X,Y and Z. A plan exists - used for the first time ever last year - to name storms after the letters from the Greek alphabet when the first list is completed. In the prediction centre in Miami, this year's list of names had been spelled out on the large map of the Atlantic fixed to one wall using magnetic letters. If - as predicted - there are 15 named storms this year, Oscar will be the last of the season.

Hurricanes that are especially deadly or destructive are routinely "retired" - a process overseen by the World Meteorological Organisation - and those names are never used again. The name Katrina was formally retired during the organisation's regional meeting last April in Puerto Rico.

The claims that global warming is directly linked to the increased destructive might of hurricanes has caused its own storm. Back in October 2004 when Trenberth first claimed a linkage, a fellow scientist with whom he was collaborating on a chapter for the report of the Intergovernmental Panel on Climate Change, resigned in outrage.

That climatologist, Chris Landsea, said at the time: "I personally cannot in good faith contribute to a process that I view as both being motivated by pre-conceived agendas and being scientifically unsound." He added: "It is beyond me why my colleagues would utilise the media to push an unsupported agenda that recent hurricane activity has been due to global warming."

Landsea is a climatologist at the tropical prediction centre, administered by the US National Oceanic and Atmospheric Administration (NOAA), and he was present the day that his colleagues were plotting the course of Alberto. Friendly and patient, part of Landsea's job is to help produce the government's twice-yearly seasonal hurricane forecast.

Like most of the experts in this field, Landsea believes that more work needs to be done to analyse the record of storms within the Atlantic and to try and ascertain whether there is any trend. But based on the data currently available he believes no such trend can be ascertained, a view he summed up in a paper he co-authored in the November 2005 edition the journal of the American Meteorological Society. He wrote: "To summarise, claims of linkages between global warming and hurricane impact are premature¿..Yet claims of such connections persist, particularly in support of a political agenda focussed on greenhouse gas emission reduction."

In his office Landsea handed me a bright red graph that plotted the major Atlantic hurricanes from 1944-2005, the spikes marking the years that saw the most powerful storms. The graph, including a line that followed the five year average, showed there was an above-average number of category three, four and five storms between the years 1948-1954 and that the number than gradually fell. During the 1970s and 80s, the graph showed, there was a lower than average number of powerful storms. Then, in the early 90s the numbers started to rise again.

Landsea disputes Emanuel's findings and says there is insufficient evidence to conclude that the this recent increase in hurricane activity since the early 90s is anything other than natural "multidecadal oscillations", or put otherwise, part of a natural cycle. He has strongly questioned Emanuel's methodology, claiming that the adjustments Emanuel made to the data from the 1930s resulted in a false conclusion.

He understands why Emanuel made those adjustments - because of research Landsea himself carried out, the scientific community had believed that hurricane measurements from the 1930s overstated the strength of those storms compared to the storms of the 70s, 80s and early 90s. But Landsea said it was now believed the bias may in fact lie the other way, and that the strength of the storms of the 70s, 80s and early 90s was under estimated. He said there was evidence for this; Hurricane Andrew, which struck in 1992 and is second only to Katrina for the total damage it caused, was originally measured as a category four hurricane. Ten years later it was recategorised as category five.

"If you used the adjusted data there is a trend, if you use the original data it is just a cycle," said Landsea. "[My problem] is with his methodology."

Another scientist who questions the conclusions of Emanuel and others is Roger Piekle, of the University of Colorado's Centre for Science and Technology Research Policy. Piekle has spent time investigating the damage caused by hurricanes, arguing that if - as Emanuel claims - hurricanes have increased in their intensity and potential destructive power one should be able to quantify that damage.

Yet he argues no such conclusion can be drawn. "If hurricanes are indeed getting more destructive over time then this trend should manifest itself in more destruction. However, my analysis of long-term data of hurricane losses in the US shows no upward trend once the date are normalised to remove the effects of societal change."

This debate has been played over the last couple of years in the scientific journals, at conferences and on the internet. But is the ordinary person to make of it? Furthermore, how can professional climatologists look at the same data and come to such different conclusions?

Both sides accuse the other of a narrow-minded fundamentalism. "We are looking at the same data," confirmed Emanuel. "What is very striking is how highly linked hurricane action is to ocean temperatures."

Emanuel does not name names but asked why others looking at the same evidence reach different conclusions, he said: "The problem is that some of my colleagues have allowed their political and even their religious convictions to influence what they do. You can tell this because they only ever argue about a single piece of evidence¿I think scientists ought to weight the evidence. When you see something before you have seen the evidence, that is prejudging."

Of Piekle's findings, Emanuel says his own measure showing an increase in the destructive power of storms measure the entirety of the hurricane's lifespan - not simply a measure of its windspeed when it strikes land. "Although Atlantic hurricanes do most of their damage 6-12 hours after landfall, they last for an average of 180 hours; moreover they only a fraction of hurricanes ever affect the US coastline," he wrote in a paper that rebutted Piekle's claims.

In response to Landsea's criticisms, Emanuel accepts that his application of "smoothing" some of the data may have exaggerated the intensity of recent storms in the Atlantic. Yet he says such smoothing did not affect his analysis of the western Pacific storms. Furthermore he argues that there remains a high correlation between storm intensity and sea temperatures. " I maintain," he wrote in a paper last year, "that current levels of tropical storminess are unprecedented in the historical record and that a global-warming signal is now emerging in records of hurricane activity. This is especially true when one looks at global activity and not just the 12 per cent of storms that occur in the Atlantic."

In an interview, he also warned: "We probably won't see a quiet decade again in the Atlantic. We may see some quiet years - this year may be quiet - but I don't think we will see a quiet decade like the 70s and 80s."

Asked if that was the result of climate change, he said: "As far as we can understand it, that wold seem to be the case."

This May, the experts at NOAA predicted that the 2006 season would produce 13-16 "named storms" and eight to ten hurricanes, of which four to six could be major systems. Two weeks ago it trimmed its projection, predicting a total of 12-15 names storms, seven to nine hurricanes and three to four major storms.

The official revision followed that of the forecasters at Colorado State University, who reduced their forecast . They believe there will be three to five major storms this year. So far there have been just three named systems - Alberto, Beryl and Chris, which caused minimal damage - but the season formally stretches from June 1 to November 30 and there are plenty more opportunities for chaos and damage. August to October is usually the busiest time.

Experts say they are getting increasingly better at predicting hurricanes. In terms of the twice-yearly seasonal forecasts which have been issued only over the last 10 years, forecasters are seeking to predict the number of named storms, the number of hurricanes, the number of major hurricanes and finally something called the accumulated cyclone energy index - a measure of the combined strength of all the storms of any season. The forecasters aim to be correct on at least three out of four of these measures.

In terms of the shorter-term forecasts that are made once a storm is active and people are desperately anxious to know whether it is going to run through their town or else pass them by and wreak havoc elsewhere, officials say they can measure their improvement. "Our five days forecast [where forecasters predict a storm will be in five days time] is now as good as our three day forecast was 15 years ago," said Frank Lepore, a spokesman for the tropical prediction centre.

In terms of the longer predications, those forecasts by the scientists seeking to decide whether this last decade of high storm activity is part of a natural cycle or else something more ominous, the only thing about which there appears to be agreement is that more work needs to done. No one doubts that since the early 1990s storms have increased in their intensity and no one doubts that average sea temperatures have increased slightly, though measurably, over the last 30 years. Whether there is link between these two phenomenon remains the unanswered question.

Landsea admits that such a lack of hard, provable answers and such dispute between scientists over data must be bewildering for ordinary people. "I'm sure it's confusing to the general public because it's confusing to us," he said. "We are trying to figure out what happened in the past and what will happen in the future. And when you're talking about hurricanes it's a pretty important thing."