Of the Baker City Herald

Smoke billows, eyes drip, lungs burn.

And Roger Ottmar smiles.

To Ottmar, smoke is more than an odor and more than an irritant.

For him every blue tendril is a specimen whose secrets he needs to study.

Ottmar, who works for the U.S. Forest Service, is not content just to watch smoke drift through the trees at a 550-acre prescribed fire at the south end of Baker City's watershed, 10 miles west of town.

When he and his fellow smoke scientists work on a fire they also erect 20-foot-tall towers fitted with sensors that measure the volume of smoke.

They hook up battery-powered pumps that push smoke into stainless steel cylinders, which they'll haul back to a laboratory so they can find out what sorts of gases are actually in the smoke.

And when they've assembled all the data, the researchers will use the information to refine computer models that predict how much smoke a particular fire will emit and, equally important, where that smoke will wind up.

Those predictions are crucial, Ottmar said, because with them in hand, forest managers can light prescribed fires with more assurance that the smoke won't settle like L.A. smog into a nearby populated area.

No predictions about fires or smoke are foolproof, he emphasizes.

But researchers have andquot;improved upon the models considerably over the past 20 years,andquot; he said.

To refine those models Ottmar and his team from the Forest Service's Forestry Sciences Lab in Seattle have traveled thousands of miles to study smoke.

Their eyes have watered all over the United States, from Alaska to Florida. They've even flown to Brazil to find out whether smoke is different south of the equator.

Last weekend, though, Ottmar was much closer to home.

He was working in the Elkhorn Mountains above Baker City with most of the 10 members of his team, which has the official title (take a deep breath) of Fire and Environmental Research Applications Group.

The team was joined by about three dozen other Forest Service scientists from Seattle and from the agency's Rocky Mountain Research Station in Missoula, Mont.

Ottmar has studied smoke since 1978.

That was a time, he said, when the Forest Service and other federal agencies were beginning to consider ways to reduce the effects of the smoke their prescribed fires created.

The agencies' interest was spurred in part, Ottmar said, by Congress' 1972 passage of the Clean Air Act.

andquot;People were, and are, very concerned about the health effects of smoke, both for firefighters and the public,andquot; he said.

Smoke hard to predict

But predicting the behavior of smoke, an ephemeral substance whose movements are affected by temperature changes of a few degrees or even the lightest puff of wind, is difficult.

Ottmar's focus, though, is not so much on where the smoke will go, but on how much of it there will be.

Estimating the volume of smoke a specific fire will generate requires three types of data, he said.

The first is the amount of fuel how much combustible stuff there is in the area to be burned, including standing trees, fallen logs, twigs and needles.

Second, Ottmar needs to know about how much of that fuel actually will burn.

One way he does this is by wrapping lengths of wire around logs on the ground. After the fire is out he collects the wires and, based on how much of the log is left (if any), he can measure how much of it burned.

From those figures Ottmar can calculate how many tons of stuff per acre the fire actually consumed.

The third and final element he needs to calculate the amount of smoke from a particular fire is what experts call the andquot;emission factor.andquot;

This factor, which varies depending on the type of forest involved, basically is an estimate of how much smoke a fire would produce, given a specific amount of fuel burned.

andquot;That's where the towers come in,andquot; Ottmar said.

By collecting actual volumes of smoke from fires in which both the amount of fuel available and the amount burned are known, the towers help Ottmar and other researchers improve their computer models.

Those models can predict smoke volume from any fire, whether it's a prescribed blaze or a wildfire.

Although the Forest Service and other agencies have employed computer models since the late 1970s, those early versions were crude compared to today's, Ottmar said.

During the past two decades researchers have created new models that not only more accurately estimate smoke volume, but also better predict where the smoke will drift.

Besides collecting cylinders of smoke from the Baker City watershed prescribed fire, members of Ottmar's team also measured the blaze's heat, both on and above the ground.

andquot;That's very important, because the amount of heat determines how high the smoke will go,andquot; Ottmar said.

Think of the andquot;stop, drop and rollandquot; lesson from elementary school when air is heated it rises.

Thus, Ottmar said, the hotter the fire, the higher its smoke will climb.

This concept is a crucial one for people who plan prescribed fires, because the higher smoke rises, the more likely it is to be dispersed by the wind, which usually increases with altitude.