The analysis, which was developed by researchers at the U.S.D.A. Forest Service and Oregon State University, also shows that the northwestern part of the country will face forest and rangeland fires this year that could be unusually severe and generally the worst of any area in the nation.
Wet weather has recently predominated in the Southwest, Central and Atlantic coast regions of the United States. As a result, 2005 should be a fairly mild fire season in most areas of the nation.
But the same forecasts for a period from now to August show some pockets of drought severity and associated fire risks in southern Florida, Maine and southeastern Arizona. And there is a major problem of historic proportions developing in a six-state region that includes Oregon, Washington, northern California, Idaho, Montana and Wyoming.
"We project that the drought severity the northwestern states are now experiencing will only get worse in coming months, and reach levels that were generally seen during the Dust Bowl of the 1930s," said Ronald Neilson, a bioclimatologist with the U.S. Forest Service and professor of botany at OSU.
Severe drought does not automatically translate into major fires, said Neilson and colleague James Lenihan, a fire and ecosystem modeler.
"It takes ignition sources such as lightning storms to trigger multiple fires, and that doesn't always happen," Lenihan said. "But those events are fairly common and, because of that, fires will often occur if vegetation, moisture and climatic conditions are right, which it appears they will be this year."
The latest consensus forecast for fire risk this spring and summer indicate huge fo rest and rangeland fire outbreaks in northeastern Oregon and southeastern Washington, and more isolated but severe fire potential in forests near Eugene, Roseburg, Bend and the Portland area in Oregon.
Other severe fires are forecast for parts of southwestern Idaho and parts of Montana.
Unlike short-term weather forecasts, these projections are ultimately based on long-term, global climate models and a "general vegetation model" created by researchers from the Forest Service and OSU, including the work of associate professor of geosciences Chris Daly and the OSU Spatial Climate Analysis Service.
These systems have simulated drought and fire in the American West, for instance, fairly accurately backwards in time to 1895, and can also be used to make both near-term and longer projections into the future. They are constantly updated, and now include the latest actual weather information through the end of last January.
In terms of the current projections for a tier of states in the northwestern U.S., it appears the situation is going to go from bad to worse. It bears some similarity to conditions last seen in 2001.
"We use five different global climate models as the underlying basis for our projections, and they are all showing the same thing," Neilson said. "It is going to become extremely dry in many parts of the Pacific Northwest and northern Rocky Mountain states, and the fire risk is going to be significantly higher than normal. There is nothing to indicate a wet spring."
Fire is one obvious implication, the researchers said, but stream flows, fisheries, agriculture, recreation and industry may all be affected.
According to Lenihan, the major storms that inundated southern California and other parts of the Southwest this year have significantly reduced the fire risks in that region, at least for this year. Vegetation growth that occurs during years with heavier rains, however, can sometimes set the stage for major fires in following years if drought resumes.
The models that simulate the droughts and the likelihood of fire are complex and sometimes counter-intuitive, the researchers said, involving such things as vegetation growth, fuel loads, soil moisture, climatic trends and other factors.
"Last year was a pretty severe drought over quite a bit of the nation, but for various reasons our model didn't really suggest a bad fire year for 2004, and in fact it was a pretty mild year," Lenihan said.
"That was fortunate, because there's some evidence that the low-risk fire years are more difficult to predict than those with higher risk," he said. "When our models show a very high level of fire, as they do now, they are usually pretty accurate."
Graphic images mapping the U.S. drought and fire risk conditions are available to illustrate this story, at the web site of OSU News and Communication Services, at http://oregonstate.edu/dept/ncs/photos/index.html