By Richard Black
Environment correspondent, BBC News website
Increases in hurricane intensity are down to humanity's greenhouse gas emissions, according to a new analysis.
Katrina drew attention to a possible link from emissions to storms
Scientists calculate that two-thirds of the recent rise in sea temperatures, thought to fuel hurricanes, is down to anthropogenic emissions.
Research published last year found there had been a sharp rise in the incidence of category 4 and 5 storms - the strongest - in recent decades.
But other scientists caution there may be errors in historical storm records.
Hurricane formation is strongly linked to sea-surface temperature, with warmer waters more likely to form storms.
Sea-surface temperature and hurricane strength vary naturally, and deciphering a clear impact of human greenhouse gas emissions has been difficult.
However, the last two years have seen several major pieces of research which have at least increased understanding of the issue, without settling it conclusively.
Peak of intensity
In July last year, Kerry Emanuel, from the Massachusetts Institute of Technology, published research showing that the duration, maximum wind speeds and energy released in tropical storms has increased markedly in both the North Atlantic and the North Pacific oceans since the mid-1970s.
A few months later, Peter Webster from the Georgia Institute of Technology documented a rise in the incidence of category 4 and 5 storms; the 15-year period from 1975 to 1989 saw 171 severe hurricanes, but the number rose to 269 for the subsequent 15 years.
He told the BBC News website at the time: "What I think we can say is that the increase in intensity is probably accounted for by the increase in sea-surface temperature, and I think probably the sea-surface temperature increase is a manifestation of global warming."
Then in June this year, Kevin Trenberth of the US National Center for Atmospheric Research analysed the exceptionally active 2005 North Atlantic hurricane season.
Sea-surface temperatures (SST) had been 0.9C above the long-term average, he found; and by comparing the North Atlantic with other regions of the ocean, he deduced that human greenhouse gas emissions accounted for about half of this rise.
The latest research takes things another step further, using 22 computer models of climate to examine a possible link between SSTs and human-induced global warming.
These models typically deal in projections and probabilities, which is inevitable with a huge and chaotic system such as global climate.
Benjamin Santer, Tom Wigley and colleagues conclude: "There is an 84% chance that external forcing [human activities] explain at least 67% of the observed SST increases" in the Pacific and Atlantic zones where hurricanes form.
"The important conclusion is that the observed SST increases in these hurricane breeding grounds cannot be explained by natural processes alone," said Dr Wigley.
"The best explanation for these changes has to include a large human influence."
On a political level, the debate over hurricane intensity has become a poster child for climate sceptics.
They object to what they regard as overblown, opportunistic links made by some commentators between the devastation of Hurricane Katrina, in particular, and rising greenhouse gas concentrations.
Their arguments are given scientific underpinning by problems which some researchers have identified in historical records of storms.
Satellite observations date back only about 35 years, and even then there are issues of calibration. Were early instruments measuring things in precisely the same way as their successors? Before that, researchers have to rely on written records from observations on land and at sea.
"I've seen examples where historically a storm was given a certain strength, and then it's re-analysed and it comes out as having been much stronger," commented Julian Heming, tropical prediction scientist with the UK Meteorological Office.
"And even if we take it as read that there is an increase in the baseline of sea-surface temperature, there are complexities in the way that cyclone formation reacts to that," he told the BBC News website.
"The hurricane now approaching Bermuda [Florence] struggled and struggled to get to hurricane intensity; so there will be year-to-year differences, and even storm-to-storm differences."
Sceptical observers also maintain that computer models are far from perfect.
But as Benjamin Santer and Tom Wigley point out, what other tools are there for projecting the future?
"In the real world, we are performing an unprecedented and uncontrolled geophysical experiment," they write. "We know, beyond a shadow of doubt, that [human] activities have changed the chemical composition of the Earth's atmosphere.
"In a post-Katrina world, we need to do the best job we possibly can to understand the complex influences on hurricane intensity."