Thomas Keever saw the telltale signs of trouble on his computer a month before it hit. A trajectory traveling northwards out of the Caribbean, curving toward the southeast picking up microscopic spores on the backside of winds from Hurricane Ivan that eventually brought Asian soybean rust to the United States.

You'll likely come to think of him from now through October as your thrice-weekly weatherman regarding the threat of soybean rust in the United States in 2005. The forecast gives growers and others in the industry a 24- to 72-hour heads-up regarding rust infestations.

As the meteorologist at the North American Plant Disease Forecast Center on the campus of North Carolina State University in Raleigh, N.C., Keever brings together information from a myriad of sources over the Internet to write forecasts in next-to-real time about the what, when and the risk to an disease epidemic in the North American Hemisphere.

The forecasts are listed on the Web at http://www.ces.ncsu.edu/edu/depts/pp/soybeanrust/.

The Center, the brainchild of a semi-retired, white-bearded plant epidemiologist Charlie Main, got its start in the summer of 1995 thanks largely to an outbreak of blue mold in burley tobacco in the mountains of North Carolina. That first year, Main, Keever and other scientists “cut and pasted” the forecasts and faxed them out to growers and Extension agents.

Today, taking advantage of the real-time aspect of the Internet, the forecasts can do exactly what they're meant to do, tell about something that's coming, instead of something that's already happened.

Keever has been working on soybean rust with various state and federal agencies for at least two years. His first official forecasts of the year hit the soybean rust Web site the first of March.

“Right now, I'm looking at sources that may be present in the Caribbean or perhaps along the Gulf Coast of the U.S.,” Keever said in late January while hanging the meat on the Web site. “We're running some daily trajectories out of those areas to see what might be happening.”

In the world of plant epidemics, fungal diseases have an over-wintering site. They spread from there. Keever has been checking winter freezes throughout the U.S. He lists Louisiana and portions of Florida as the only areas immune to the hard winter freezes that kill the fungus.

In addition to possible over-wintering sites in the extreme Southern U.S., Keever has his eye trained on sites in western Cuba, Jamaica and the Yucatan.

Working at his computer, Keever picks a number of meteorological parameters, including moisture, and weather from the recent past, present and future.

He's already worked out a “for instance” trajectory, assuming that soybean rust over-wintered in Mobile, Ala., showing where the spores may have gone on a particular day in January. No cause for alarm now.

“If I see that pattern sometime in March or April, it's really going to get my attention, moving north through the lower Mississippi Valley and curving northeast through the Ohio Valley,” Keever says. “By that time, kudzu would be flourishing and the soybean crop would be in the ground. That's the kind of things I'll be looking for in my forecasts.”

Using the Web site of the Air Resources Laboratory, an agency of the National Oceanic and Atmospheric Administration, Keever picks a set of data to run trajectories from specific longitude and latitude or picks a spot on the map.

“Western Cuba,” he says. “I'll be running trajectories out of there for blue mold and soybean rust this year. We'll use that as one sentinel, so to speak.”

Using the Air Resources Laboratory's Hy-Split model takes an air concentration or deposition forecast at 3,000 meters. The air concentration forecast mimics the number of spores and their movement through the atmosphere.

A deposition simulates the path of the spores as they move through the atmosphere, taking into account rainfall. “Looking at the rainfall gives you a picture of where the spores are deposited on the ground, which is very helpful to me in making my forecast,” Keever says.

It's this kind of information Keever will use in making his thrice-weekly forecasts. He looks for “dangerous situations for the transport of spores and the weather that accompanies them.”

The forecasts will contain reports from the sources of soybean rust spores; meteorological conditions that could contribute to its spread; the likelihood of that spread; and the associated risk. The site will also have the recommended fungicides, rates and equipment for management of soybean rust.

Using disease information at the “source,” combined with rainfall at the target area and conditions for infection, Keever puts all of the information together and comes up with a risk assessment for those crops.

“One of the most important pieces of information is what's going on with the disease source,” Keever says. “I cannot forecast the source of the disease that I don't know about. That local information is important.”

Because the plant disease forecast center began with another fungal disease, Keever considers the scientists have a slight jump on soybean rust. “We're in the same sort of situation with soybean rust as we are with blue mold,” Keever says. “I suspect early in the season, we'll see soybean rust established on kudzu and move to soybeans as the crops come up.”

While soybean rust only hit the U.S. in the fall of 2005, the center in North Carolina, as well as others, has been working on the disease for about two years.

Keever and the scientists at the center were asked in the fall of 2003 to do historical tracking of soybean rust's trek from Africa to South America. “Our group and some other groups did analysis work that concluded soybean rust arrived from Africa around New Year's 2000-2001 in the southern part of Brazil. From there, over the next two to three years, it moved into portions of Colombia. That brought us to the work we were doing last year when there was great concern that soybean rust might spread from South America into the Caribbean and into the United States.”

Keever was on loan to the USDA-APHIS doing work on soybean forecasting last year during the height of hurricane season. “There was an enormous amount of hurricane activity last year, so there was a good deal of concern as to whether these hurricanes might be bringing spores up to the U.S. from South America.”

Charlie and Ivan got Keever's attention. “As it turns out, the threat from Charlie was relatively low,” Keever says, pointing to a list of hurricane names during the 2004 season.

“Ivan turned out to be a different story,” Keever says.

He started getting concerned about the trajectory of Ivan the first week in September. “Some of our forecast pathways were showing that there might be some interaction with the hurricane” and the spores that were active in Colombia,” Keever says.

“One of the disturbing things was the path the hurricane was taking. Once it had moved to the west, the winds on the east side were pushing spores farther north,” Keever says.

This event turned out to be the winds that pushed soybean rust into the U.S., with the first confirmation being in southeast Louisiana, and subsequently in nine states at multiple locations.

Keever describes his work as “gathering information from the Internet and various other sources and applying the weather information at hand.”

“Thomas is being a little modest,” Main says. “I don't mind tooting his horn. Thomas gave a pretty good indication of when and where this disease might occur.

“I often say that either we're getting better at this or we're damn lucky — one or the two. We're maybe a little bit of both.”

e-mail: cyancy@primediabusiness.com