AIR Worldwide Corp., the Boston-based catastrophe modeling firm, said it is at work on a new hybrid hurricane storm track model.

The company said it made the announcement at a recent conference it held where it discussed a near-term projection for hurricanes striking the U.S. coast over the next few years. AIR said it sees the risk increasing 10 percent on average with the largest relative increase in the southeast.

The firm said its near-term projections, which consumer advocates have attacked for skewing insurance rates upward, was presented to climate scientists and meteorologists at its International Summit on Hurricanes and Climate Change held in May at Hersonissos, Crete.

AIR said Peter Dailey, director of atmospheric sciences at AIR, presented the company’s near-term hurricane risk methodology and related research at the meeting.

The firm methodology used to develop the 2007 near-term model was reviewed by Kerry Emanuel of Massachusetts Institute of Technology, Timothy Hall of National Aeronautics and Space Administration, and James Elsner of Florida State University, who co-organized the summit in Crete, AIR said.

According to AIR’s statement from Mr. Dailey, news coming out of Crete “is that the wider scientific community is also now beginning to focus on how potentially increased hurricane activity in the Atlantic basin may translate to activity along the U.S. coast.”

During his presentation, Mr. Dailey discussed the statistical analysis undertaken by AIR to develop a near-term view of hurricane risk constructed by hurricane intensity and by coastal region.

The company said the concept is conditioned on the assumption that sea surface temperatures will likely remain elevated above the long-term average for the next several years.

AIR said it presented preliminary findings at the conference from a hybrid physical-statistical storm track model that it is developing as a complement to its ongoing statistical analysis.

The current research, AIR said, is examining where storms form in the North Atlantic under conditions of warm sea surface temperatures and to what extent the genesis location impacts the probability of landfall.

“The effect of climate on tropical cyclone activity is complex,” said Mr. Dailey. “Various climate signals, as well as naturally occurring feedback mechanisms, affect the likelihood of storms forming and intensifying into hurricanes and the nature of their track–and sometimes these effects compete with one another.”

He noted that in 2006 the onset of mild to moderate El Ni?o sea current conditions in the tropical Pacific led to increased wind shear in the Atlantic, which is detrimental to hurricane development.

In addition, Mr. Dailey said storms over Africa’s Sahara Desert “carried significant amounts of dry, dusty air westward over the Atlantic Ocean, depriving incipient tropical cyclones the moisture and heat they need to mature.”

In his review of AIR’s projections, the company said that M.I.T.’s Mr. Emanuel stated that AIR follows a “wise strategy, allowing one to make inferences about the effect of very basic sea surface temperature variability on hurricane activity without unduly stressing the statistical significance of results based on the limited best track data.”

Mr. Hall of NASA, according to AIR, said, “I agree with AIR’s decision to switch from a sea surface temperature point forecast to a category forecast (warmer vs. colder).”

He added that “[sea surface temperature] has a significant influence on tropical cyclone landfall rates, but the forecasts are not reliable enough to warrant precision beyond broad categories. Overall, I judge AIR’s North American tropical cyclone analysis to be sound and of high quality.”

“A big advantage of AIR’s approach is that it is entirely reproducible,” said Mr. Dailey. “The data is in the public domain and the analytical techniques used are well known and well accepted. That means that it can be peer-reviewed and presented at scientific forums like the Summit.”