Oregon Public Broadcasting

Fire crews around the state are preparing for what they say could be the busiest few days of the 2010 wildfire season so far. Reporter David Nogueras traveled to Prineville to find out how fire officials make those predictions and then use them to prevent fires from spreading.

David Nogueras / OPB

Mapping fire probabilities at the Central Oregon Interagency Dispatch Center.

Dispatcher: "Central Oregon Dispatch, JeffCo Fire, ODF Grizzly?"

It's 3pm at the Central Oregon Interagency Dispatch and things are just starting to pick up here.

When wildfires break out in Central Oregon, COIDC can call on a network of spotters to determine where the fire is and then dispatch fire crews to that location to take part in what they call the "the initial attack". 

But it's been slow so far this year. 

Dispatcher: "Hey Don. I just got a call from JeffCo fire they've got a fire east of Deer Ridge…deer Ridge is reporting it. It's in the juniper and it's rolling…"

COIDC provides support for the Bureau of Land Management, The Oregon Department of Forestry and the Deschutes and Ochoco National Forests.  All told, it's responsible for more than 9-and-a-half million acres of land.

Megan Teaford, a dispatcher here says part of reason COIDC is able to cover that much turf is because it has access to resources.

Megan Teaford: "Central Oregon is a big aircraft hub, so we can host a lot of stuff here…."

Teaford explains that the center can dispatch jump crews, lead ships, air attack platforms, smokejumpers and repel crews…lot of vital tools to fight fires.  She says that can mean the difference between whether a brush fire is stamped out early,  or whether it grows into really big and expensive fire.

So how do managers know how to best deploy these resources? Grant Kemp lets me in on the secret: data-driven guesswork.

Grant Kemp: "It's not an easy task but the predictive services folks up at the Northwest Interagency Coordination Center, they've come up with these formulas and models and these are the results of it."

Firefighters have been using meteorologists in the field since World War II.  Only in the last 10 years has the meteorologist seen that role greatly expanded to include both regional and longer-term forecasts.

Kemp directs my attention to a large screen. It's a webpage that looks like a weather report. But instead of listing the likelihood of sunshine or rain, it lists the likelihood of wildfire.

Grant Kemp: "So you see today it says the probability of ignition is only one fire today in Central Oregon. But as we get toward Saturday and Sunday, the probability of eight to nine fires on both of those days and the probability of a large fire is 20 percent"

That means there's a one-in-five 5 chance a timber fire could develop into a fire that are 100 acres or more. 

Kemp is ahead of the game thanks in part to this guy.

John Saltenberger: "My name is John Saltenberger and my title is Fire Weather Program Manager for Predictive Services at the Northwest Coordination Center".

John Saltenberger is a meteorologist by training, but for the last 25 years he's worked alongside firefighters, first in the fire camps and more recently behind a desk at the Northwest Coordination Center in Portland.

Saltenberger says his job isn't only concerned with current conditions, but past conditions as well.

John Saltenberger: "Heah accurate record keeping, thorough record keeping is an integral part of this business, because when people ask me to do fire predictive forecasts I become a historian."

Trying to look past a 10 day window, says Saltenberger is a little bit of art and little bit of science.

This year, Saltenberger made his long-term prediction using objective and subjective techniques. He looked through past records to find a year that had similar hallmarks to the early part of this year. 

He checked the number of lightning strikes, number of acres burned and El Nino and La Nina weather patterns, Saltenberger found one year that looked more promising than the rest. 

1998.

John Saltenberger: "So it looked very similar to what's happening in 2010, in fact I went back and did some objective comparison of other past years to the 2010 season and when I did that objective correlation I found the 1998 season was the number 1 most closely correlated season to the 2010 season."

So with the data in hand, Saltenberger based his long-term forecast on the 1998 fire season.

But '98 started out with a hot and dry July and so far though, that hasn't come to pass.

John Saltenberger: "Yeah, it is disappointing".

But, says Saltenberger it's work that's still worth doing.

John Saltenberger: "Even if we find a tool that doesn't work, there's value in knowing that it doesn't work. That it may not have any further value and therefore not pursuing it any further which just opens the doors for looking at other types of things."

Saltenberger's says he hopes that by opening those doors, he can do his part to keep firefighters and keep more roofs over people's heads.