COLORADO State University hurricane forecaster William Gray and his team are predicting an above-average hurricane season next year.

"We foresee a slightly above-average hurricane season for the Atlantic basin in 2005," wrote Gray and Philip Klotzbach. "Also, an above-average probability of US major hurricane landfall is anticipated. We do not, however, expect anything close to the US landfalling hurricane activity of 2004.

"We believe that 2005 will continue the trend of enhanced major hurricane activity in the Atlantic basin that we have seen over the past ten years," said Mr. Gray last week.

In their annual "Extended Range Forecast of Atlantic Seasonal Hurricane Activity and US Landfall Strike Probability", Gray and Klotzbach conceded that recent hurricane activity had forced them to change their forecasting methodology.

While their previous method had demonstrated "hindcast skill" for the period 1950-90, it did not "give skilful results when utilised on a real-time basis for forecasts between 1995-2001. This was due to the discontinuation of the strong relationships we had earlier found between West African rainfall and the stratospheric Quasi-Biennial Oscillation with Atlantic basin major hurricane activity."

Even forecasters are allowed to blame the weather, and the researchers displayed a little pique with systems which didn't stick to the script.

"We did not expect these relationships that had worked so well for 41 years to stop working from 1995 onward," Gray huffed. "We do not yet have a good explanation why these relationships failed."

This led them to "discontinue" their previous forecast scheme and develop a a new "pool of six predictors" which show strong statistical relationships with "combinations of these predictors and Atlantic basin hurricane activity occurring the following year".

None of the predictors appears to involve bottles of clear shark oil, the clouding of which is held by St. David's islanders to herald severe unpleasantness.

"We estimate that 2005 will have about six hurricanes (average is 5.9), 11 named storms (9.6), 55 named storm days (49), 25 hurricane days (24.5), three intense (category 3-4-5) hurricanes (2.3) and six intense hurricane days (5.0). The probability of US major hurricane landfall is estimated to be 30 per cent above the long-period average," wrote Gray.

In 2004, by contrast, there were nine hurricanes, 15 named storms, 91 named storm days, 45 hurricane days, six intense hurricanes, and 23 intense hurricane days.

Gray and Klotzbach describe the increase in major hurricane activity since 1995 as the result of North Atlantic sea surface temperatures "having become about 0.4 to 0.6? C warmer than normal since 1995", and tropical Atlantic August-October "upper tropospheric" winds, six to nine miles above the surface, having increased from the east. These changes, and changes in upper atmospheric vertical wind shear and low-level horizontal wind shear were not the result of global warming.

"It was the strengthening of the Atlantic Ocean thermohaline circulation (THC) which led to these Atlantic basin changes," Gray and company asserted. According to Tim Osborn of the Climate Research Unit at the University of East Anglia, the density of sea water is controlled by its temperature (thermo) and its salinity (haline), and the global oceanic circulation driven by density differences is therefore called the thermohaline circulation.

"The Gulf Stream, and its extension, the North Atlantic Drift, bring warm, salty water to the northeast Atlantic, warming western Europe," writes Mr. Osborn. "The water cools, mixes with cold water coming from the Arctic Ocean, and becomes so dense that it sinks, both to the south and east of Greenland."

Cold, dense water currents gradually warm, and return to the surface, throughout all of the world's oceans. The surface and sub-surface currents, the sinking regions, and the return of water to the surface form this closed loop called the thermohaline circulation.

Osborn believes that "increased rainfall and warming are both expected as a result of increased greenhouse gas concentrations", but Gray is not convinced. They have both seen paleoclimatic evidence that this circulation system has fluctuated both recently and in the distant past.

Gray reported that "historical and geographic evidence going back thousands of years indicate that shifts in the thermohaline circulation tend to occur on periods of about 25 to 50 years".

He suggested that the period of diminished hurricane activity experienced between 1970 and 1994 was a result of an earlier "shift" in global oceanic circulation, changed again since 1995 by this period of warmer and saltier Atlantic water, and that therefore "enhanced Atlantic basin major hurricane activity will persist through the early decades of the 21st century".

Gray and Klotzbach have what appears to be a very sound argument why "global warming" was not the deadly instigator of the 2004 hurricanes.

"Many individuals have queried whether the unprecedented landfall of four destructive hurricanes in a seven-week period is related in any way to to human-induced climate changes," they wrote. "There is no evidence that this is the case.

"If global warming were the cause of the increase in US hurricane landfalls in 2004, and the overall increase in Atlantic basin major hurricane activity of the previous nine years, we would expect to see an increase in the other storm basins as well (West Pacific, East Pacific, Indian Ocean, etc.). This has not occurred.

"When tropical cyclones world-wide are summed, there has actually been a slight decrease since 1995. In addition, it has been well-documented that the measured global warming of about 0.5? C (between) 1970 and 1994 was accompanied by a downturn in Atlantic basin hurricane activity."

They attributed 2004 and the previous nine seasons to the oceanic circulatory fluctuations.

"Major hurricane activity in the Atlantic has been shown to undergo marked multi-decadal fluctuations that are directly related to North Atlantic sea surface temperature anomalies.

"It was the environmental steering currents that drove four of the six major hurricanes of 2004 on such long, low-latitude westerly tracks that made this season so special," Gray and company proposed. "The very damaging Atlantic 2004 hurricane season was simply a low probability event resulting from unusual natural variability in the ocean-atmosphere system.''