Forecasting Fires All in a Day’s Work

Associated Press Writer

It’s barely dawn when Mike Fitzpatrick starts his shift with a blur of colorful maps, figures and endless charts, but already he knows what the day will bring. Lightning will strike in places he expects. Winds will pick up, moist places will dry and flames will roar.

But where will the next wildfire hit? Are homes and lives threatened? Where should crews go to stop it? Could it get out of control?

This is Predictive Services of the Northwest Interagency Coordination Center, and Fitzpatrick is in the business of forecasting wildfires.

It’s crucial in the fire-prone West. The decisions made here can save time, money and lives.

Near Portland International Airport, in a nondescript office building cluttered with stacks of graphs, coffee mugs and maps bursting with tacks, Fitzpatrick, head of the six-person prediction unit, gets the morning’s status report. The evening shift ended at midnight, and Fitzpatrick has to quickly find out how wildfires in Washington and Oregon behaved overnight.


A fire that began the day before near an Oregon Indian reservation is still burning. Conditions are ripe for a bad day. Brush and trees are everywhere, the ground is flat, the wind steady. The small community of Simnasho is in the fire’s path. Already, 1,500 acres have burned.

“Wind is going to be the deciding factor over there today,” Fitzpatrick tells the center manager, Gerry Day. Fitzpatrick thinks about the resources tied up fighting fires in California and Alaska. He hopes that he won’t need them.

He glances at the dry-erase board of fire details, and sees that 130 firefighters and five engines are on the scene. Fire managers want four more engines.

“This thing is going to turn into a big fire,” Fitzpatrick says.

It’s 7:35 a.m.


After the 1988 fires in Yellowstone National Park, government agencies suggested a change. Instead of reacting to fires after they begin, what about trying to predict them?

It was a new way of thinking about wildfires, and it would take 10 more years before Predictive Services was created here, the first in the country. It wasn’t until 2001, after the hiring of two meteorologists, that the unit really took off. Across the country, there are now 10 other prediction units like this one.

It’s midmorning when Paul Werth, Pacific Northwest fire weather program manager, hands Fitzpatrick graphs that show rivers of air moving over Washington and Oregon. Wind, unstable air patterns and dryness are perfect conditions for large fires.

“It looks like we’re going to get into more thunderstorm activity,” Werth tells Fitzpatrick.

While Fitzpatrick worries about the future, it’s Werth’s job to study the past. He pores through more than 30 years of records from different countries to come up with a scenario for the Pacific Northwest. He can predict weather patterns 10 days out by studying temperature, rainfall, humidity, fuel moisture and what happened in the past.

“Before we had Predictive Services, we only knew what the fire danger was for the next day,” Werth says. “But fire managers really want to know what it’s going to be next week.”

The Pacific Northwest averages 4,000 wildfires a year; most are caused by lightning. This year has been mild, with just more than 36,000 acres burned in Washington and Oregon, but scientists know that can change. In Oregon now, all but one area -- the central section of the coast -- is at high risk for fire. All of Washington is at risk.

The next cubicle over, meteorologist Terry Marsha goes over lightning predictions. The day before, he forecast lightning in five areas; he was right on four. Lightning sparked 150 fires, but they were quickly put out.

He talks with meteorologists from the National Weather Service, plugs weather data into equations and comes up with a probability for wildfires starting.

Miles away in Mt. Hood National Forest, a remote area weather station pumps out data every hour: rainfall, wind speed and direction. Marsha relies on 200 such stations to help him make his forecasts.

“The fuel conditions are ready to go,” he says. “All you’re waiting for is something to set it off.”

It’s Marsha’s job to find out where that might be.

Almost noon, and the six fires burning in Washington and one in Oregon are holding. Fitzpatrick predicts that the Log Springs fire near the Warm Springs Indian Reservation will burn another 1,500 acres today.


On the other side of the coordination center, Steve Dickenson, a former smokejumper and Hotshot crew leader, uses the intelligence reports to determine which fires should get resources. When fire managers run out of their own crews, they call here and ask for help.

Dispatchers can quickly assign crews, engines and air tankers, but Dickenson makes tough decisions on which fires to make a priority. A wrong decision could leave an area vulnerable.

A fire manager in Washington wants to know if he can let extra crews go, but Dickenson, the emergency operations manager, wants him to keep them. The threat is still there. “In the past, when we have not had Predictive Services, decisions were made just anecdotally,” Dickenson says. “We need to be making decisions on the best science.”

Two days before, air tankers were positioned in Moses Lake, Wash., when Dickenson decided to move them to Redmond, Ore. The same morning, a fire broke out on the California-Oregon border, and tankers put it out.

“I try to pre-position six hours before,” he says. “Two hours is too late. Having that resource there before it’s needed is crucial.”

Marsha can forecast lightning strikes within a few hours. Dickenson reviews the forecast, compares it to area fuel conditions, surveys how many crews are available and dispatches resources to the areas most at risk.

“We have not got science to the point where you can tell where the fire is going to be started,” he says. “We just try to put it as close to the lightning track as [it’s] been forecasted.”

There are times, Dickenson says, when he moves resources and nothing happens. But that doesn’t happen often.

Mt. Hood National Forest fire manager Reggie Huston says the forecast “gives us the ability to do quite a bit of pre-planning for any type of event that may be coming at us.” She says if she knows an area is at risk, she can assign more crews and helicopters, and alert the public.

When Dickenson’s resources are tapped, the National Interagency Fire Center in Boise, Idaho, takes over, able to dispatch crews from all over the country.


Despite the science of wildfire forecasts, knowing where to put crews is still a gamble. Will this be the fire that explodes? Are there enough crews or too many? Are other areas at risk?

“Guessing will always be a part of it because of weather,” Dickenson says.

Fitzpatrick and Marsha say some fire managers prefer instead to rely on instinct and experience. Many just snap a stick to see if an area is dry.

“We’re trying to bring more science into a discipline that doesn’t accept science,” Marsha says. “It’s kind of a macho profession.”

In the future, they hope that Predictive Services becomes standardized nationwide, with units, fire managers and agencies working together.

It’s been a good day, and Fitzpatrick is hoping that the luck continues. “Every day that goes by that we don’t get the big fire, the probability of having a big fire lessens,” he says.

But they remember the Biscuit fire of 2002. It had been one of about 500 in late July, burning just a few acres. It threatened nothing. By the end of August, it burned 500,000 acres, becoming the largest in Oregon history.

“These other fires got the resources. It wasn’t a mistake. It’s just the way it worked out. Your forces are overwhelmed,” Fitzpatrick says.

He makes a final round of calls, checking once more on the seven fires. The Log Springs fire did what Fitzpatrick expected -- gained 1,500 more acres. Eight days after it began, the fire had grown to 13,500 acres. The fires look good, but the threat still looms. Tomorrow is still to come.