Faced with the onset of devastating wildfires and their rising costs, federal officials 19 years ago announced an ambitious plan to create a new data-driven approach they said would blunt fire's dreadful impact on the Western landscape.
The solution: A sophisticated computer program that would redirect resources across the country. That program would show where more firefighting aerial tankers and firefighting crews were needed. It would identify the most vulnerable landscapes where controlled burns and the thinning of dried out trees would best promote forest health.
Federal officials predicted in 2001 that it would take five years to build the computer program, which they named Fire Program Analysis. They said it would work across the five federal agencies responsible for fighting wildfires and coordinate their efforts into a more sensible approach.
But after spending more than $50 million on the initiative, those promises still have not come to fruition. Federal authorities abandoned the Fire Program Analysis system in 2014, telling the U.S. Government Accountability Office at that time that the system “only delivered inconsistent and unacceptable results.”
The former team leader of the project said the effort was largely doomed because of political infighting, the victim of forces opposed to losing control of spending priorities to a new system that would objectively analyze where the needs were greatest.
“There was this institutional inertia to continue on with the way things had been before,” said Steve Botti, who once was in charge of developing the FPA system.
Botti said the effort to build the FPA system was launched after officials in Congress and the White House Office of Management and Budget expressed fears that federal agencies were developing their firefighting arsenals based on wishful thinking and historical patterns that weren't grounded in effective science and data. The critics feared the agencies were establishing their own fiefdoms, and that data and computer modeling was needed to help guide the federal government’s response, especially when wildfires were becoming a growing problem for the nation, he recalled.
Since the FPA program has been disbanded, efforts to pull off something similar still have moved forward, though not to the degree originally envisioned.
Researchers in the U.S. Fish and Wildlife Service built another computer modeling program after the FPA system was retired, but that more recent program still isn’t driving budget decisions, according to the researcher who led that effort. And the National Park Service months ago decided to stop funding a similar computer program that was providing firefighting and fire mitigation analysis for the nation’s parks, including Rocky Mountain National Park.
As a result, the five federal agencies in charge of fighting wildland fires still haven’t coordinated efforts even as states like Colorado struggle with ever larger forest fires that are growing in intensity, said Botti, who retired from the National Park Service program in 2007 and now is the mayor of Stanley, Idaho.
For decades, the GAO, Congress’ watchdog agency, has highlighted the need for an analytical system to help determine the most cost-effective way to reduce the risk of wildfires and determine where firefighting crews and equipment are most needed.
A 2012 GAO report stated that the five federal agencies in charge of fighting wildfires still hadn’t coordinated their efforts despite more than a decade of criticism from Congress and government watchdogs. That report described the federal officials' efforts to build a robust analytical system as marked by “delays and revisions, with the agencies several years behind their initially projected time for using it to help develop their budget requests.”
“The idea of having a tool or a suite of tools that would be helpful in the allocation process would be something that is useful, but we’re not in a position to say what that should look like or what it should consist of,” said Anne-Marie Fennel, the GAO’s director of natural resources and environment.
Colorado’s not the only area of the nation that’s vulnerable. A December GAO report found that five federal land management agencies — the Department of Agriculture’s Forest Service and the Department of Interior’s Bureau of Indian Affairs, Bureau of Land Management, Fish and Wildlife Service and National Park Service — estimate that more than 100 million acres they manage are at high risk from wildfires in any given year. But those agencies only have the resources to address a fraction of those areas, mitigating just 3 million acres in 2018, according to that report.
Botti said he thinks the federal agencies would be better prepared if federal officials had followed the original recommendations of an early prototype of the now-defunct Fire Program Analysis system released in 2006.
He said the modeling from that early version of the computer program found vast inefficiencies in where federal firefighting resources had been deployed. The early modeling suggested billions of dollars in firefighting resources needed to be moved from less vulnerable areas of the nation to other states with higher wildfire risks, Botti said.
“There was some opposition to that, quite a bit really, because it was attempting to change the way, fundamentally, resources would be allocated throughout the United States,” Botti said.
Part of the pushback came from federal officials who argued that they knew better how to deploy resources than what the modeling from the computer program suggested, he said.
“It was said that if the model suggests that if even 5% of resources had to be reallocated from one location to another, that was going to be unacceptable,” Botti said. “In some cases, it was much more out of whack than that. These subjective budgets had been built up over decades and some areas had far more resources than they needed, and others were deficient.”
Mark Finney, a researcher in the U.S. Forest Service who was involved in the FPA initiative, said political infighting wasn’t the cause of the controversy. He said the early efforts lacked good data, and that the FPA computer program produced inconsistent and “counterintuitive results.
“The pushback that I saw was that, in the early days, the system was not producing logical results,” he said. “It assumed perfect knowledge. It started producing nonsensical results that were the exact opposite of the way an organization would respond if it got more money.”
Douglas Rideout, a Colorado State University forest researcher and another member of the FPA team, defended the work on the system, which produced a favorable peer review in 2011.
“It was starting to produce results,” he said. “And I think it scared people. It was suggesting that budgets ought to be reallocated, but the appetite for the federal government changing budgets around was not very good.”
Finney and Rideout each have separately tried to find ways to rekindle the original vision of FPA through new computer programs they’ve played a role in developing. Both of their initiatives borrow some of the concepts of FPA, but they’ve also run into roadblocks.
Rideout helped create for the National Park Service a program called STARFire, which he said in recent years began showing promising results. The program analyzed Rocky Mountain National Park, he said, and also was used to develop recommendations on how to lessen the risk of wildfires throughout the state of Idaho and other areas of the nation.
“This system did everything anybody wanted it to do,” Rideout said. “It would locate potential fuels treatments across entire landscapes. It would do it based on return on investment. It would show how to allocate budgets between fuels treatments and fire preparedness. It was entirely strategic, and it was getting better and better.”
Several months ago, National Park Service officials told him they were defunding the program.
“The federal government needs something like this,” he said. “When they defunded it, I don’t really believe they knew as an organization what it is that they had let go.”
National Park Service officials said they used STARFire to analyze about 50 national parks, leaving about 350 other parks unaddressed. Nothing is scheduled to replace the program, they said.
“We were trying to work toward that national view so we would have data and outputs coast to coast and border to border to look at it from a national perspective and then down to a regional perspective to help inform decisions,” said Mark Koontz, a National Park Service wildland fire operations leader. “But we weren’t there, and it was going to take a considerable amount of more time to get us there because of some of the challenges of getting the data and inputs to run the model. So it was decided to refocus funding towards other smaller efforts that will provide greater benefits sooner as far as informing decisions as we move forward.”
For Finney, of the U.S. Forest Service, the efforts continue toward refining his brainchild: a computer program called the Wildland Fire Investment Planning System, which he described as the next generation of FPA. He said that newer computer system is capable of analyzing tradeoffs in fire program investments for firefighting preparedness, large fire suppression and hazardous fuel treatment.
But it still has limitations. Efforts still are underway to try to gather data on state and local firefighting resources so the new computer program can conduct a comprehensive analysis on where the federal government’s firefighting resources are most needed. And although the new program can analyze landscapes to determine which forests and landscapes need thinning and other fuels treatment to minimize wildfire risks, the program isn’t driving budget decisions in any federal agencies yet, he said.
He said the program shows that the federal government needs to invest dramatically more, at least 10 times more, to help address overgrowth that has built up in the nation’s forests. But such investments, which often require controlled burns to reduce overgrown, dried out vegetation vulnerable to ignition, aren’t always popular with the public, he said.
Finney said that while controlled burns are effective, people often distrust them and complain that they contribute to pollution and view them as putting forests at risk. In reality, he said, they keep forests healthy and help to avoid larger, devastating blazes.
He said that several years ago he produced a study on Colorado’s wildfire risks calling for a series of changes and mitigation strategies to forests and landscapes throughout the state. Most of the strategies identified in the report remain unrealized, he said.
On the 10th anniversary of the Hayman fire, which in 2002 burned 133 homes in Colorado and engulfed 138,144 acres and forced the evacuation of more than 5,000 people, Finney gave a speech in Colorado with a warning.
“I gave this speech on the Hayman fire,” he recalled. “And I said we haven’t seen the worst fire in Colorado history yet, and how do I know that? We haven’t done anything differently. Why would you expect to have something better in the future?”
He said he could give the same speech today, even as Colorado has endured a fire season with record-setting blazes when it comes to the amount of acres burned.
“I’m just one voice among many, many saying the same thing,” Finney said. “And this is just one fire season again. And once it’s over, people will go, ‘Whew, we survived that one.’ But it’s just going to keep going like this unless we start making the necessary changes.”