Consensus Growing for Active Hurricane Season

(April 23) -- When the private forecasting company WSI increased the number of predicted tropical storms and hurricanes in its updated Atlantic hurricane forecast this week, it added to the growing consensus among forecasters that the 2010 season will be an active one, unlike last year.

WSI increased the number of predicted hurricanes from seven to nine. The overall WSI seasonal forecast, with 16 predicted named storms (hurricanes and tropical storms), is similar in intensity to the forecasts prepared by Colorado State University (15 named storms) and AccuWeather (16 to 18 named storms). WSI indicated it's more likely that these numbers would be increased, not decreased, in any future updates.

The long-term (1950-2000) average is 9.6 named storms, and the short-term (2000-2009) average is 15 named storms.

The WSI forecast cited some of the same reasons for concern about an active season that Colorado State and AccuWeather had given: exceptionally warm Atlantic water, which is warmer now than it was at the same time in the record-breaking 2005 season, and the expected weakening of the Pacific El Nino. Atlantic hurricane seasons that follow an El Nino tend to be more active than normal.

Skeptics of the accuracy of seasonal forecasts are numerous, often citing the inability of forecasters to predict the weather for next week as evidence that a seasonal forecast is impossible to make with any accuracy and is therefore useless. While that might seem like a logical argument, short-range and long-range forecasts are vastly different.

Short-range forecasts focus on details, such as whether the high temperature will be 80 or 74 degrees or whether it will snow six inches or two inches. Forecasters look at small-scale factors to make those determinations. Any slight error in one of these details makes for an inaccurate forecast.

A long-range forecast, such as a hurricane season forecast, is a general forecast based on large-scale weather factors. The large-scale weather factors don't change nearly as quickly as the small-scale factors needed to make a short-range forecast, and long-range forecasts can be accurate even if a short-range forecast is not.

This is not to say that long-range forecasters are always correct, but since the large-scale factors that result in an active hurricane season are fairly well understood, forecasters have shown some skill.

Long-range forecasts are more often about trends rather than specifics. For example, with a seasonal hurricane forecast, a forecast might focus on how active this season will be compared with last season, how storm totals will compare to an average season or which areas are at greater risk for landfall.

In fact, the three private outlets mentioned here, along with the federal government's Climate Prediction Center, all correctly predicted that the 2009 season would be less active than the 2008 season. The initial forecast for all of the entities was for a greater number of storms than actually occurred -- the seasonal total was nine -- but they were correct that the trend would be fewer storms. There were 16 named storms, including eight hurricanes, in the 2008 season.

Even though the specific numbers might not be right this season either, the trend is clear: Private forecasters are expecting an active hurricane season, and given the current large-scale weather factors, it would not be surprising if any updated 2010 seasonal forecasts are for more storms, not fewer.

The official U.S. government forecast issued by the Climate Prediction Center will be issued May 20. We'll see if it adds to the consensus.