After 11 years, U.S. Fire Program Analysis system still isn’t ready

Eleven years ago, federal agencies announced a bold strategy to battle the growing threat of catastrophic wildfires.

Across the West, vast expanses of forests had grown dangerously thick from decades of all-out fire suppression. At their edges, an army of urban refugees bored deeper and deeper into the woods, building dream homes in the trees.

The government’s planned response: a sophisticated new computer system — called Fire Program Analysis, or FPA — that would enable firefighting agencies to coordinate their efforts and maximize their resources.

It would help them weigh the benefits of fire suppression versus forest thinning, evaluate where to station people and equipment and decide how many planes to buy. It would be up and running by 2007, a powerful new tool in the yearly battle to save homes from the flames. In 2012 — while tens of thousands of Colorado residents have fled their homes from an alarming eruption of June wildfires — the program is still incomplete.

Federal agencies, led by the U.S. Forest Service, are still working on it. A peer review of the latest prototype called it the only alternative “to reduce risk and control costs at the national level” — but concluded it isn’t ready to use as intended.

And the former team leader on the project said it is a shadow of what it was supposed to be, the victim of forces opposed to a process that would take decisions about where to put resources out of their hands.

For years, the U.S. Government Accountability Office has warned Congress in reports and committee hearings that federal firefighting agencies were failing to develop the vital new tool they promised more than a decade ago.

“We think it’s very important,” said Anu Mittal, the GAO’s director of natural resources and environment.

The Fire Program Analysis system was intended not only to help agencies develop sensible budgets but also to identify “the most cost-effective mix of firefighting assets,” she said. “In essence, it’s supposed to help them make good decisions about asset allocations. The fact that it still is not functional after a decade is troublesome.”

Tom Tidwell, the Forest Service chief, said as FPA’s development has evolved, it has been significantly improved to include better data on not just historic fire patterns, but also current conditions and projections.

He also emphasized that the majority of Forest Service firefighters and crews are extremely mobile: When fires exploded in Colorado, Utah, Wyoming and Montana, the agency redeployed resources that had been fighting fires in the Southwest.

But Tidwell acknowledged that the Forest Service faces daunting challenges in its efforts to control wildfires, a task that has grown from 13 percent to 48 percent of the agency budget since 1995.

He said unnaturally thick growths of trees have accumulated in “65 million acres just in the nation’s forests,” which the Forest Service is hoping to treat with thinning and burning measures at a rate of up to 4 million acres a year.

A warmer climate has extended the fire season, complicating efforts, Tidwell said.

“This change in climate has just added to the severity of the fire seasons we’re seeing today,” he said. “You take that in addition to the amount of fuel that is available for these fires … it’s drying out, you’ve got dry, hot, windy weather and longer fire seasons, and you’re seeing the results of that today in Colorado.”

While this year’s early-season fires have destroyed record numbers of Colorado homes and killed at least six people, the burned area remains smaller than the 619,000 acres of Colorado scorched in 2002, the year of the Hayman fire.

Already this year, more than 200 large and small fires have burned more than 172,000 acres in Colorado — 70,000 acres more than the entire fire- season average for the previous five years.

FPA, the new system intended to gain ground against growing costs, was estimated in 2001 to “take about five years to complete,” the GAO reported.

FPA was established after officials in Congress and the White House Office of Management and Budget grew unhappy with the way federal agencies developed their wildland-fire budget requests, believing them to be wish lists rather than based on cost-effective analysis, said Stephen Botti, who led the project in its first phases before retiring as the National Park Service’s fire-program planning manager.

Botti and his team — which included Douglas Rideout, a forest economist from Colorado State University — went about developing the tool. Botti said that initial version took agencies’ objectives and then optimized them to determine how to best allocate resources for the greatest impact.

The idea was to figure out how much money to devote to fire suppression, and to reducing fuels to improve overall forest health, and where to do it.

But when the tool was used for a preliminary analysis in 2006, not everyone liked what it found, Botti said. The results showed which areas needed more resources and which needed less, throwing into uncertainty budgets used for staff programs and some administrative overhead, he said.

For instance, one recommendation was to move resources from coastal Alaska, where wildfires are relatively rare, to California, where they regularly wreak havoc in populated areas, Botti said.

“We’re talking about a couple of billion dollars in federal wildland-fire funds here,” he said. “Any time you tinker with that, it becomes political in a hurry. There was pushback from the bureaus that the answer was not acceptable. This was mainly the Forest Service objecting to that.”

Botti said the Department of the Interior’s budget officer welcomed the model.

Even so, a team of scientists was brought in to analyze FPA and found faults. Botti said that in his view, some officials didn’t like anything independent that would cut them out of decision-making.

The original model was scrapped and replaced with one that Botti said focused on simulating fire scenarios and gave fire managers license to manipulate the system to achieve results they wanted.

The version of FPA now under development offers suggestions about how to balance suppressing fires and reducing hazardous fuels, but it appears it will not be used as a budgeting tool for every individual geographic fire unit in the U.S. as originally envisioned, he said.

“There are a lot of people in the bureaus who are dedicated to doing the right thing,” Botti said. “When you get higher up the food chain, it gets more political and, obviously, some people are interested in protecting turf and organizations.”

John Phipps, a Forest Service official working on the revised model, said the agency has begun using it for national budget planning and hopes to deliver desktop versions to field offices across the country, enabling them to use resources more effectively.

“Quite honestly, I don’t think there was any plot” to scuttle the original system, he said.

But he agreed that people in Forest Service field offices feared — and still fear — a computer model that could deprive them of people and equipment.

“Their fear was FPA was going to be a black box,” he said. “We’d turn the crank and spit out the numbers, and they’d become victims of some scientist’s planning model.”

Tidwell, the Forest Service chief, said results from the initial version of FPA “were not credible, to the point where even a nonfire person would look at it and say, ‘That just doesn’t make any sense.’ “

But Rideout, the CSU wildland-fire economist and researcher, defended the methodology behind the original FPA, noting it passed a peer review published in a scientific journal last year.

“It was a new effort, and in many respects it seemed to be working,” Rideout said. “It had its critics, it had its own warts and wrinkles, but it was a huge step.”

FPA was supposed to coordinate firefighting efforts of five federal agencies — the Forest Service, the Bureau of Land Management, the Bureau of Indian Affairs, the Fish and Wildlife Service and the National Park Service — that manage more than 1 million square miles of land in the U.S.

But for years, government auditors have repeatedly reported that federal agencies delayed its development and monkeyed with its original intent.

• In 2008, the GAO reported to Congress that federal wildland-fire costs had tripled since the mid-1990s to more than $3 billion a year, citing three factors: “uncharacteristic accumulations of vegetation” from fire suppression; increasing human development in wildlands; and severe drought “in part due to climate change.”

Federal efforts to combat these growing threats had been hampered by the delayed development of FPA, “which will drive billions of dollars in federal expenditures each year and directly affect millions of citizens living in fire-prone areas,” the GAO reported.

• In 2009, the GAO testified that federal officials “told us they will not rely substantially on FPA’s results until needed improvements are made.”

• In 2011, the GAO testified that federal agencies expected to use FPA in making 2013 budget requests but that “these efforts are still under way.”

Asked how this year’s fire outbreak might be different if the original FPA were in place as planned, Rideout said: “I think the responses to fire would be more cost-effective. I’m not sure whether we would have gotten to these fires any faster or later or better, or with less expense.”

This year, four years after the GAO asked for the study, federal agencies received an independent peer review on their proposed FPA model. It concluded that progress has been made but that improvements are needed before “an integrated, interagency tool” can be put to use.

For one, the feature “that supports allocation decision-making is not ready,” the review concluded, “because it does not exist yet.”

At the same time, the report stressed the importance of completing this project.

“There is no alternative to FPA,” it concluded, “for national coordination of fire investments to reduce risk and control costs at the national level.”

David Olinger: 303-954-1498 or dolinger@denverpost.com

Eric Gorski: 303-954-1971, egorski@denverpost.com or twitter.com/egorski