Lessons in the Ashes

Firefighter with a drip torch

A firefighter uses a drip torch to apply fire during a prescribed burn. (Photo by Swanson Scott, U.S. Fish and Wildlife Service.)

Before Crystal Kolden became an assistant professor of Geography at the University of Idaho she fought wildfires. Originally a history major, Kolden says she had no interest in the career paths of lawyer or history teacher that lay before her. So after graduation, she headed off to the El Dorado National Forest near Lake Tahoe. There, she worked in timber and recreation until someone handed her a giant metal can with a curled prong at the end. The can, called a drip torch, is used to light fires during prescribed burns. “And that was it,” says Kolden. “I fell in love with fire.”

In more than a decade that Kolden spent on the ground in fire management, she witnessed first-hand the impacts of fire on a range of different landscapes. She also heard many different perspectives on fire’s role in ecosystems. One thing she observed across the board, was that fire will rarely burn through an area completely. Instead, fires usually leave behind unburned islands of vegetation within the perimeter of the burned area. To an ecologist, these unburned islands are valuable assets. They serve as refugia for wildlife and as landscape features that can accelerate forest regeneration after fire. But to a firefighter, unburned islands pose a threat.

“An island of unburned trees could torch out and launch an ember across the fire line that could create a spot fire that might perpetuate the wildfire,” explains Kolden. For this reason, she says, firefighters often intentionally burn to ashes any unburned vegetation near a fire’s perimeter.

Kolden’s curiosity about the relative merits of practices like this and like suppressing fires in wilderness areas eventually led her to enroll in a master’s program in geography at the University of Nevada in Reno, and later, in a geography Ph.D. program at Clark University in Worcester, Massachusetts. In graduate school, Kolden increased her understanding of fire’s impacts by honing her skills using geographic information systems (GIS) and information from satellites, including USGS’s Landsat satellites, and planes to analyze landscape-level patterns of wildfire over time. Using data like this is known as “remote sensing.”

While still a graduate student, Kolden joined a US Forest Service Enterprise Team devoted to fire ecology and fire behavior research. Enterprise teams are units within the Forest Service that act somewhat like private contractors to help find ways for government agencies to do better work. Kolden’s team primarily focused on collecting real time fire behavior data on wildfires.

“Ever seen the movie Twister?” asks Kolden. “We’d do something similar to that, but with more safety measures.” She and her team would work with the incident team to predict where fire was likely to go the next day or two and try to get ahead of it to place data collection equipment in its path.

Today, Kolden manages teams that collect information after fires all across the Northwest. She combines the on-the-ground data they collect with information from Landsat data to answer several key questions relating to wildfire and unburned islands. One of these questions is whether fires have been or will become more intense due to climate change. To date, there is no consensus on the answer to this because fire intensity can be a hard thing to measure, or even define.

Kolden had the idea of using Landsat data to locate unburned islands in order to see whether they have been getting smaller or fewer in number over time. She reasoned that if this were the case it would suggest that wildfires are getting more intense. If they are getting more intense, as many suspect, then future fire management strategies will need to be adjusted accordingly.

With Kolden’s teams traveling all over the Northwest to get fire behavior data, Kolden is able to test and improve mapping capabilities. Once she has validated her results from remote sensing by comparing them with the data collected on the ground, Kolden will be able to apply her improved model to 32 years of remotely sensed data to map all unburned islands for the entire Northwest. 

Another goal of Kolden’s research is to better understand the variables that contribute to the creation of unburned islands—things like invasive versus native grasses, tree density, forest structure, and natural barriers, such as streams and rock. As she puts it, “Once we understand the factors that determine where unburned islands form, we can use that information to manage and manipulate vegetation across the landscape to cause the intentional formation of unburned islands to protect places we don’t want to burn: cultural resources, homes, critical habitat, and endangered species.”

Building fire resilience into these features of the landscape in advance may save Forest Service and other groups the huge expense of emergency measures down the road.


About 300 miles away from Kolden’s office at the University of Idaho in Moscow, is the Boise office of the US Geological Survey (USGS)’s Western Geographic Science Center and the base of Jason Kreitler, a USGS research geographer. Like Kolden, Kreitler has spent considerable time thinking about the policies that shape wildland fire management. However, Kreitler is examining the problem with a different lens, using economics and social science.

Kreitler explains his research focus like this: “We have fixed budgets for most, if not all, of our public land management, so the question is, how do we optimize the use of those funds to best meet our conservation goals, like protecting biodiversity or ecosystem services? I hate to use buzzwords, but it’s really how can we manage our public lands in a more ‘holistic’ way?”

As Kreitler’s research is showing, one key tactic is to incorporate costs into fuel treatment planning. Although it’s not common practice, incorporating costs can substantially increase returns on fuel treatment expenditures.

“By incorporating cost effectiveness you can treat more acres for the same cost than if you just consider benefits,” says Kreitler. “When you can reduce fire risk for more acres of forest, that translates at the landscape scale to a reduction in expected fire intensity and probability, or fewer of what are considered negative fires from a hazards perspective.”

Rather than adding yet another independent tool, Kreitler has been careful to incorporate a cost analysis tool into existing Forest Service software that managers use already. This helps estimate expected revenue or cost to treat any given tree stand by allowing users to input factors, ranging from the price of timber at the local mill and how much biomass is being harvested to how steep and remote a harvest site might be.

“The overall goal of using all these inputs,” says Kreitler, “is to use resources as effectively as possible to bring forests into a more fire resilient state.” However, he is quick to point out that quantifying resilience can be tricky, pointing out that different people may define resilience differently. Semantics aside, there’s an excellent case to be made for building fire resilience into our Northwest landscapes. As Kreitler’s research indicates, resilience is important to protect both our forests and our watersheds.

To better understand the impact of fire on water quality, Kreitler is currently working as part of an interdisciplinary team funded by the Northwest Climate Science Center to model expected changes in sedimentation caused by future wildfire in watersheds of the Western US. His collaborators on the project are Joel Sankey, a USGS research geologist based in Arizona; Todd Hawbaker, a Colorado-based USGS research ecologist; Nicole Vaillant, an Oregon-based Forest Service fire ecologist; and Scott Lowe, a professor of economics at Boise State University in Idaho.

“So far the biggest surprise for us has been the projected increases in sedimentation that our ensemble of climate, fire, and erosion models show for future scenarios of potential fire across the west,” says Kreitler.

The research team found that a full three-quarters of the watersheds modeled show an increase in sedimentation of ten percent, while one quarter show a whopping 100 percent increase in sedimentation. As Kreitler put it, “A quarter of these watersheds across a large portion of the west could see pretty drastic increases if the future turns out to be anything like our scenarios.”

Kreitler says he’s optimistic that this is a problem we can get ahead of, pointing out that the city of Denver recently saw the development of an ecosystem service-based market when the city’s water utility began levying an additional voter-approved fee on their users. This fee covers forest treatments for hazardous fuels and forest restoration to restore watersheds damaged by severe wildfires during the late 1990s and early 2000s.

“That’s one great example of wildland fire interacting with the provision of ecosystem services,” says Kreitler. “We’re interested in the circumstances that set that up. Where else could that occur? Seeing residents recognizing the problem and being willing to pay moves us in a much more
positive direction.”

Unburned islands left behind in the wake of wildfire play an important role as refugia for wildlife and can help speed forest regeneration. (Photo by Arjan Meddens​.)

Members of Crystal Kolden’s field crew collect data on an area that was burned in the previous year. (Photo by Arjan Meddens.)

Jason Kreitler. (Photo by Susan McIlroy.)