An Experiment in Co-Producing Fire and Smoke Science

In early October 2023, nearly fifty research scientists and technicians collaborating with the USDA Forest Service-sponsored Fire and Smoke Model Evaluation Experiment (FASMEE) gathered on the Fishlake National Forest to collect measurements from a rare stand-replacing prescribed fire. Developing new approaches to predict fire and smoke behavior, scientists representing the USDA Forest Service, the National Aeronautics and Space Administration (NASA), Tall Timbers, Desert Research Institute, and universities from across the country, partnered to collect fire-related data from belowground to space. These synergistic research projects characterized fuels, measured radiant heat and energy, evaluated smoke concentrations, and documented fire effects on vegetation and even bats.

Connecting Fire and Smoke Observations to Models

Photo Credit

Vita Wright

Researchers from Rocky Mountain Research Station, Michigan Tech University, and Tall Timbers Research Station measure fuel moisture, fuel loading, and unburned vegetation. 

Wildland fire managers require information and tools that accurately predict fire behavior and smoke dispersion so they can make decisions about how to plan, respond to, and manage wildfires and prescribed fires. Although today’s predictive models are extremely useful tools, fire behavior and smoke models are inherently complex and still include uncertainty. To improve predictive models for wildland fire management, scientists need detailed observations of fire and smoke; however, it is logistically difficult to collect real-time data during active wildfires. While prescribed fires are planned for wildfire risk reduction and ecosystem restoration, they provide opportunities for data collection under well-documented environmental conditions.

Adam Watts, USDA Forest Service Research Biologist and Principal Investigator of the Fire and Smoke Model Evaluation Experiment, spent much of 2023 coordinating with managers and scientists to prepare for collecting data from a complex stand-replacing prescribed fire planned for Fall 2023. By incorporating data collected during prescribed burns into predictive models, scientists can help wildland fire managers make more informed decisions based on knowledge such as where fires spread, how smoke disperses, and the effects of fire on vegetation, soil, and wildlife.

The Fire and Smoke Model Evaluation Experiment is an example of co-production, where scientists and managers partner in the development of scientific knowledge. Scientists affiliated with this project have invested the time to develop long-term relationships with managers, such as those on the Fishlake National Forest, who implement the burns these scientists study. The District Ranger and fire managers – from the burn boss to the engine foreman – value the science being produced and so cooperate with researchers to facilitate data collection.

Jason Kling, Richfield District Ranger, explained, “For the last several years, we've been experiencing some of the worst drought we've ever experienced. And, so, for me, it's important to have that connection between management and research so that we can stay up on the latest and greatest. What are they learning and where can we apply that so that, as managers, we're making the best-informed decisions and applying science so that we can be responsible and make decisions that are sound.” 

Linda Chappell, retired Intermountain Regional Fuels Specialist and a fire ecologist, also described the benefits of partnering with researchers, “For the Forest to have research here helps us meet and greet… we get a chance to ask questions that we otherwise would not have any chance to ask... So, our questions help drive what researchers are interested in, in many cases, and that's really useful to us. We get direct feedback… And the forest has better burners and better fire users for having that information and those relationships.”

To support the Fire and Smoke Model Evaluation Experiment, the Forest Service partnered with the Tall Timbers research station to implement an organizational structure that mimics an operational Incident Management Team and minimizes the impact of research on fire managers. With a substantial investment in planning and organizing both ahead of and during the burn, the scientists work through a liaison to obtain the logistical information needed to safely collect data. Given the risks and complexity of managing a stand-replacing prescribed fire, it is imperative that scientists conduct research without threatening burn operations.

Key Science Partnerships

NASA, known for exploring air and space, sent scientists from the NASA FireSense Project, who used aircraft mounted with high-tech instruments to collect data measuring fuel conditions, active fire dynamics, post-fire effects, and air quality and forecasting from high altitudes. NASA will be comparing data collected by Forest Service and academic scientists on the ground with data collected by aircraft and satellites. They are ultimately working to develop better remote sensing tools and techniques that support fire management agencies.

Mike Falkowski, Program Manager of the NASA Applied Sciences Wildland Fires and FireSense programs, explained, “One of the unique things about being here on the Fishlake [National Forest] today is we're able to work not only with the national forests, but we also have Forest Service Research here that are doing the detailed ground measurements and taking other data to help improve fire behavior and smoke models. And here at NASA, we're collaborating really closely with the Forest Service and bringing in our aircraft and our satellite observations to improve what they're doing… Also, the data they collect helps us improve our techniques and our algorithms.”

Andy Hudak, Research Forester with the Forest Service’s Rocky Mountain Research Station and a long-time partner on the Fire and Smoke Model Evaluation Experiment, worked alongside Nancy French of Michigan Tech University to lead a crew of fuels and fire effects technicians gathering data about fuel moisture, fuel loads, and fire effects on the ground. In partnership with NASA, they will compare data collected on the ground to data collected by LiDAR, a remote sensing method that uses lasers attached to trucks or airplanes, and by NASA’s satellites, to improve the accuracy of remote sensing tools used for measuring fuels and fire effects. 

Susan O’Neill, a Research Air Quality Engineer with the Pacific Northwest Research Station and architect of the BlueSky Framework, collected smoke particulate and transport data near the fire. Data and tools developed by O’Neill are frequently used by the Wildland Fire Air Quality Response Program’s Air Resource Advisors to create daily smoke forecasts and inform the public of smoke risks during wildfires.

The Fire and Smoke Model Evaluation Experiment also provides opportunities for graduate students and postdoctoral researchers to conduct cutting edge research.

In a recent scientific development, University of Idaho scientists and students used uncrewed aircraft systems (UAS), also referred to as drones, to collect data about living microbes – bacteria and fungi – in smoke. They will use these data to estimate how many microbes are released from a given fire and how far the microbes are transported downwind. With scientists and managers working side-by-side and learning from each other, Forest Service drone pilots also collected thermal infrared imagery over the burn to identify hot spots and visualize the fire’s spread. In addition to the research on living microbes in smoke that University of Idaho graduate student Phinehas Lampman conducted, Pratik Shukla, a graduate student at the University of Maryland, used a ceilometer to measure aerosol concentrations up to 15 kilometers above the ground. Pratik also recorded and mapped carbon detected in the smoke to quantify the fire’s contributions to atmospheric carbon.

Kathleen Clough, a PhD student with the Desert Research Institute’s Atmospheric Sciences Division was able to use coupled fire-atmospheric modeling to predict fire and plume behavior, as well as PM2.5 transport, faster than real-time.

Datasets developed through the Fire and Smoke Model Evaluation Experiment will be archived in a permanent, publicly accessible database, under the oversight of a data management team led by Leo Calle at the University of Maryland. “The goal of coordinated measurements, leading to compatible datasets that are useful for multiple modeling needs, is one of the most challenging concepts of FASMEE—but its importance to our continuing work to support fire managers makes it worth the effort,” says Watts.