Computer models can help manage forest fires
February 1, 2022
Last year marked the 10th anniversary of the Las Conchas wildfire, which burned 150,000 acres in and around Los Alamos, blackening the sky, threatening the Laboratory, and prompting the evacuation of the town. Wildfires across the country are larger, hotter, and more frequent than in the past, in part due to climate change and in part due to mismanagement. For over a century, humans have dealt with wildfires by putting them out, but this approach, it turns out, was misguided. The exclusion of fire from the landscape results in dangerous accumulations of combustible biomass, overgrown understories, and overpopulated canopies, all of which have paradoxically led to more out-of-control wildfires, posing severe threats to people, property and ecosystems.
Fire kills diseased plants, keeps down crowding, and recycles nutrients back into the soil—it is an integral part of forest ecology. Rather than attempting to exclude fire, land managers now look for ways to trade high-intensity wildfires for low-intensity controlled fires, mimicking the natural fires of eons past. Prescribed burning, as the practice is known, is a tool being used increasingly throughout the country to keep forests healthy and productive, help manage terrestrial carbon stores, and reduce the risk of catastrophic conflagration.
In the mountains of Northern New Mexico and throughout the West, the summer months bring plumes of smoke and the question: Is that a wildfire, or a prescribed burn? Los Alamos is intimately familiar with both. The Cerro Grande wildfire, pictured here in May of 2000, began as a prescribed burn, but it got out of control and burned 43,000 acres in and around Los Alamos, causing an estimated billion dollars in property damage. More recently, in 2011, the record-breaking Las Conchas fire was ignited by a tree falling on a power line and burned over 150,000 acres in the Los Alamos area.
Prescribed burning must manage smoke and ensure safety while maximizing effectiveness, so it requires robust scientific guidance. Here in the dry and rocky West, forests generally contain less fuel than their Eastern counterparts, but they also have much less water and often much more complex topography, complicating the practice of prescribed burning. Laboratory scientists are working along three avenues to improve prescribed burning in the West: building a science-based foundation of knowledge on which to base decisions, integrating fire models with other relevant models, and including the processes of climate change in fire management.
Modeling tools
“Seasoned prescribed-burn practitioners have amazing intuition that comes from their extensive experience,” Los Alamos scientist Rod Linn explains, “but accelerating the learning curve for less experienced practitioners is important for the success of prescribed-fire practices.” Linn and his colleagues are building a science-based foundation of knowledge—a sort of database of different sets of conditions with known outcomes—that will help land managers and fire experts know what to expect in a given scenario, even if they have never seen it before. The database includes data from real fires as well as computer models—complementary sources of invaluable information that together enable the exploration of “what if” scenarios in new landscapes with evolving conditions.
The composition of the forest affects the fire, and the fire affects the composition of the forest.
In collaboration with the United States Forest Service and Florida-based Tall Timbers Research Station, the Los Alamos team has developed a suite of computer models that are helping build the database and are being deployed in real-time fire-management situations. The first of these models was FIRETEC, which was designed to couple with a Los Alamos computational fluid-dynamics model called HIGRAD that models airflow over various terrains. The result is HIGRAD/FIRETEC, a coupled atmosphere-fire model based on physics principles such as conservation of mass, momentum, atomic species, and energy, that produces high-resolution, high-fidelity, 3D simulations of wildfires. The model has been used to explore how topography, fuel structure, and interactions between these factors drive fire behavior, and to study how fire severity is influenced by bark beetle outbreaks. HIGRAD/FIRETEC is useful not only for studying hypothetical scenarios, but also for working backwards from real-world fires to help explain their behavior and delineate cause-and-effect relationships.
However, because HIGRAD/FIRETEC runs on a Laboratory supercomputer, it’s not a field-deployable, real-time tool. To fill this niche, the scientists developed a laptop-capable, fast-running tool called QUIC-Fire. The Laboratory hosts a family of QUIC (Quick Urban & Industrial Complex) models, developed by physicist Mike Brown and his team, to model how airborne chemical, biological, and radiological agents are transported and dispersed around buildings. QUIC-Fire uses algorithms originally developed for other QUIC models, which makes it both fast and nimble.
QUIC-Fire is the team’s field-deployable, real-time fire model designed for site-specific use by members of the fire community. This QUIC-Fire simulation, set in the Valles Caldera National Preserve near Los Alamos, illustrates how a hypothetical fire might progress under specific conditions. In the second frame, the fire splits into two paths, as a result of how the wind travels over the terrain. By the fourth frame the two paths have converged and the fire continues its spread up the hill. This kind of real-time information helps fire managers know what to expect so they can plan accordingly. CREDIT: David Robinson/LANL
“Careful choices were made to balance the need for accuracy and the tool’s usability by the fire community,” says Sara Brambilla, one of the lead developers of QUIC-Fire. “We now have a tool that brings in real terrain, evolving weather, and complex ignition patterns to simulate fire spread and smoke transport. And we can run it on a laptop.” Because QUIC-Fire can run hundreds or even thousands of simulations in a relatively short time, it is especially useful for exploring a wide range of outcomes for prescribed-fire scenarios and site-specific strategies.
As well as atmospheric and transport models, the team is also integrating ecology and hydrology computer models with their fire models. For example, vegetation is a strong driver of fire behavior, and accurately linking vegetation-specific effects to fire behavior is essential. Los Alamos scientists Adam Atchley and Turin Dickman are working on understanding how the moisture levels and distribution of fuels impact the spread of fire.
The parameters in fire models become moving targets when viewed through the lens of climate change.
“Better representations of the distribution and attributes of live vegetation will improve our ability to accurately model prescribed fires, which are more sensitive to small changes in these quantities than wildfires are,” explains Atchley. This is because wildfires typically occur under ideal conditions, when things are hottest and driest, while prescribed burns are deliberately done under marginal conditions when not everything will burn. Dickman adds, “Prescribed fire is about killing some plants to save others. The fire community is just starting to recognize the importance of plant-carbon and water cycles in fire behavior and post-fire plant survival, particularly for low-intensity burns.”
In other words, it’s not just about how the composition of the forest affects the fire, but also about how the fire affects the composition of the forest. Some plant species, like the lodgepole pine, thrive in post-fire environments and have evolved so that their seeds will only germinate after having gone through low-to-moderate intensity fire. Capturing the biophysical feedbacks between vegetation and fire means scientists can explore the influence of different forest treatments, as well as drought and other warming responses, on fire outcomes, including vegetation recovery or mortality.
Fire forecast
But many of the parameters being included in the fire models—atmosphere, hydrology, vegetation, etc.—become moving targets when viewed through the lens of climate change. Longer summers with higher peak temperatures and lower precipitation can turn Western forests into tinderboxes. The Los Alamos team is using future climate projections to understand how climate change will impact land managers’ ability to put more low-intensity fire back onto the landscape. Atmospheric scientist Alex Jonko works with Atchley to study the climate-induced expansion and contraction of prescription windows—the range of conditions under which prescribed fires can be safely implemented.
HIGRAD/FIRETEC is the team’s physics-based atmosphere-fire model, which runs on a supercomputer. It produces 3D simulations of fires, which are used to understand fire behavior in general. Here a HIGRAD/FIRETEC simulated fire (left) is compared to an actual experimental burn in the Northwest Territories of Canada (right). By combining simulation data with real-world data, scientists are building a foundation of knowledge that can help decision makers know what to expect in a variety of circumstances.
Jonko explains, “Rising temperatures reduce the opportunities to burn during increasingly hot summers, and variable precipitation can lead to either drier or wetter fuels in the future, so we’re using information from climate models to evaluate how well current prescription windows might hold up.” Here in the West where the air is already dry and the terrain can be treacherous, even a small change in these variables could necessitate substantial revision of current prescribed-fire practices.
The Laboratory’s fire modeling team is also investing locally in wildfire-related projects. Local Native American communities have long understood that intentional fire is a way to keep the landscape resistant to extreme fire. To engage these communities, the fire modeling team set up a Wildfire Simulation and Visualization summer program and year-long internship targeted at Native American undergraduate students. Additionally, as Los Alamos itself is no stranger to wildfire, the team is working with the Laboratory’s Wildland Fire Program to improve understanding and find ways to reduce fire risk across the Laboratory’s 22,200 acres.
Longer fire seasons and more extreme fire behavior are the new norm. The prescription, it would seem, is to develop tools that can keep up with these changes to ensure safe communities and healthy forests.
