THE IMPORTANCE OF INFORMATION – CONNECTING RESEARCH AND OPERATIONAL NEEDS IN THE WILDFIRE COMMUNITY
BY SAMANTHA KRAMER, BRIAN POTTER, CRYSTAL MCCLURE, NATE BENSON, SHIHMING HUANG, AND MEGAN SCHILL
On April 15, 2024, 39 researchers, fire weather forecasters, and wildland fire managers gathered in Boise, Idaho, for a two-hour workshop with a clear goal: to close the gap between fire weather research and the people who rely on it every day in the field.
The workshop, Gathering User Input for Long-term Fire Weather Outlooks, was hosted at the 7th International Fire Behavior and Fuels Conference by Brian Potter, a research meteorologist with the USDA Forest Service, and Samantha Kramer, an air quality and fire weather researcher now at the Atmospheric Interdisciplinary Research Institute, with funding support from National Oceanic and Atmospheric Administration’s Office of Oceanic and Atmospheric Research (OAR) Weather Program Office. The aim was to improve communication between fire weather researchers and the wildland fire and fuels management community, and connect research priorities directly to the needs of the people using that research on the ground. Scientists Crystal McClure and Nate Benson also supported the workshop as dedicated observers and recorders.
WORKSHOP STRUCTURE
The workshop began with a presentation on the state of fire weather forecasting, the tools, displays, and forecast models available, and the needs and misconceptions researchers often encounter across the wildland fire and fuels community.
From there, participants broke into six discussion tables and then reconvened as a full group, working through a series of wildland fire forecasting scenarios. Input was captured by note-takers at each table, and additional feedback forms were passed out and collected at the door. The forms turned out to be especially useful: they highlighted concrete steps researchers can take to make sure their products align with end-user priorities. What follows starts with that feedback, since it best captures how participants see the field, before moving into the scenario discussions.
To ensure the workshop met expectations to bridge gaps across research and wildland fire and fuels management sectors, participants were asked for their work affiliation and primary work focus based on broadscale categories. These selections are self-identified primary sector(s) of employment and role(s) in wildland fire and fuels management. Federal affiliations were the largest group that participated in the workshop, followed by academic, local, state, and private industry, respectively. Most participants had only one affiliation, with a few having multiple affiliations across sectors. Therefore, participants largely represented different sectors involved in wildland fire management. Participants were primarily from the United States, with a few international attendees. Self-identified work foci were grouped into four categories (research, tool development, fire operation planning, and incident response) as well as an “other” category with write-in details (Figure 1). Of 39 participants, 16 opted for other work foci in addition to, or instead of, the categories provided. Most write-ins converge on pre-fire planning and long-term solutions (e.g., training, fuels or land management, mitigation, resiliency). While most participants identified single affiliations, 49 per cent identified two to four primary work foci. Academic affiliations commonly identified two foci: research and tool development. Federal and local affiliations primarily identified a single work focus. Every state affiliated participant reported multiple job foci.
QUESTIONS
The main thread of the workshop was a series of questions for three weather forecast lead-time scenarios to gauge what is important to the wildland fire and fuels community three months before the fire season, one month before the fire season, and seven to 10 days before the fire season (Figure 2).
The fire season in the United States varies by region, but generally runs from spring to fall, with a peak in summer and early fall months. For each scenario, the questions discussed at each table were: what decisions participants have to make at that point; if weather plays a role in those decisions; what other factors such as staff, equipment, budget, or public engagement influence the decisions; which weather and climate factors matter most; whether participants and practitioners need precursor, current, or forecast information about those factors; what spatial and time scales matter for that information, both in span and resolution; what precision or confidence level participants need in that weather information; what tools and products participants currently use for weather information; and how well those tools succeed or fail to meet the needs just described.
RESULTS
Results were summarized using designated recorders notes from the discussions and through the feedback forms. These questions were then combined by fire weather timescale scenario.
THREE MONTHS OUT
Responses ranged widely on whether precision or confidence was more important for fire weather forecasts, with some participants even indicating that neither needed to be very high, likely in part because many indicated they had few or no reliable sources of information three months before fire season.
Lightning and dry thunderstorms were mentioned many times, and participants noted that no tools are currently available to predict these conditions further than a few days ahead of time.
Multiple respondents highlighted the importance of lead time that was multiple months before fire season. Having better information on conditions like precipitation, fuel moisture, wind, temperature, and lightning in advance would help fire managers at the local, regional, and national levels prioritize areas to pre-position resources.
Several respondents also noted that greater lead time on fire weather information could improve public education and increase the percentage of communities that would take warnings seriously.
In general, individuals who were directly responsible for wildland fire response in government, private industry, or at non-profit organizations were more likely than those who were affiliated with research to highlight a longer lead time as important.
ONE MONTH OUT
Most respondents expected to have more reliable information one month before the start of fire season than three months prior to the start. Participants indicated that the one-month mark is when training and preparation begin to transition into operations.
Fuel moisture and vegetation conditions were mentioned by 12 participants as being important for planning one month prior to fire season, as conditions are less likely to change significantly at that point than three months before. Participants were split on whether spatial scales should be regional to national or local, suggesting that tools capable of providing both would likely see better adoption.
SEVEN TO 10 DAYS OUT
The weather factors that were mentioned for the seven to 10 days before and into fire season scenario tended to be more immediate and included specific conditions like flash draught, lightning, probability of red flag warnings, wind shifts, wind drops, and the presence of drying winds. Three respondents specified that while one-month and three-month forecast tools were important for general, longer-scale weather patterns, seven-to-10-day forecasts are used to adjust to changing weather patterns. At this stage forecasts help fire managers and emergency response react effectively and efficiently to changing conditions.
Twelve out of 34 participants who responded to this scenario said that the seven-to-10-day forecast drives critical public communications, so the forecast needs to be precise, accurate, and easy to communicate. One participant even mentioned using the information for social media posts, which suggests a tool suitable for this stage may increase its visibility and use.
After the workshop discussion of the three time-frame scenarios was completed, the participants were shown a final list of three over-arching questions for group discussion and on paper feedback forms: whether it mattered if the tool used to answer the three-month questions was the same one used for the 10-day questions; how they evaluate a product or source; and what training or support they wanted for new tools or changes to existing ones.
Here are summary comments from this discussion.
On tool preference: More than half of the participants indicated they did not feel it mattered whether the model used to answer three-month and 10-day questions was the same. Many participants said having more information in one place made a tool more user friendly and increased the chances that their colleagues would use it. Several also indicated that if information was reliable, they did not care what model was used to answer questions. Several participants were split on whether the model mattered and answered that they use as many tools and data sources as they can to try to assess data convergence. Furthermore, several of participants said they verify the accuracy of information based on agreement among tools instead of building trust in a single resource. Respondents who provided more technical or specific data-related reasons for their points of view were more likely to suggest different models for each timescale was preferred. One respondent said three-month predictions focus on climatological information while 10-day forecasts focus on weather, which means different models will ensure both questions are answered with the same level of reliability.
On evaluating new products: Most respondents said they evaluate products and sources through comparison, either to published literature, existing tools they trust, or actual observed conditions.
This indicates new products and sources may have difficulty gaining traction but published papers and strong correlation to observed events can help build trust. Few participants said they evaluated a source based only on the opinions of their colleagues and scientists in the field, indicating responsibility for self-assessment when possible. A handful of respondents also said the user-friendliness of the tool was the most important metric for evaluating a product; these participants were less likely to cite the accuracy or performance of the product alongside useability. Nearly one-third of participants reported valuing non-technical features of a tool, which suggests that cosmetic features and usability contribute significantly to the adoptability of forecast products and tools.
On training and support: The responses to this question were the widest ranging, spanning “no idea” to “seven-day” training. Five participants indicated that a demonstration of the model would be helpful, either recorded live or laid out in a written document. Comparison was again mentioned as helpful, and numerous workshop participants said they would like the tool developers to give a snapshot of how their product compares to other currently available tools and to articulate the gap that they feel their model would fill. A few respondents mentioned that training on the “under the hood” workings of new models would be useful for helping them understand the limitations and biases. Multiple respondents also said having access to a user support line would be helpful to receive training in real time, but one participant noted that funding for this would likely be a challenge.
While responses varied widely, most participants indicated that an interactive element would be helpful, and it seemed that training that was directly applicable to individual use cases was most desired. For this reason, recorded training and written materials may be more effective when paired with live webinars, phone calls, or in-person training so that users can get answers to their individual questions.
Finally, a respondent provided this direct feedback: “in my opinion, one of the biggest challenges is sifting thought all the guidance (tools, data, etc.) [to] see what is useful and for what purpose. Make sure we are actually creating guidance that solves the questions we have.”
Whatever format is used to create training materials, multiple workshop participants indicated that unclear or verbose training material can negatively impact a tool’s adoption.
At the end of feedback forms, workshop participants were asked for any information they would find helpful three months, one month, and seven to 10 days before fire season. These sections were useful for highlighting critical gaps in fire weather prediction tools. Lightning was mentioned 34 times across the workshop participant feedback papers. One respondent wrote that “an analysis of future lightning events could be modeled looking at past years where major lightning has occurred” would be helpful, as he and his coworkers currently rely on local news for this information. His work is affiliated with prescribed burning at the federal level, and the local news does not have the level of scientific rigor or consistent accuracy necessary for fire weather forecasting. Another topic that was repeatedly referenced was moisture content of fuels and overall precipitation levels. Significant rain events can signal the end of fire season, and participants indicated that this information could impact how they manage fuels and prescribed burns, so accurate information and forecasts is important.
Participants were able to request continued communication from the project team regarding future data availability, prototype forecast products, and training opportunities. Most participants requested follow-ups for all updates regarding fire weather data, forecasts, and training. Those affiliated with state agencies were less interested in data and forecast prototypes, possibly due to having multiple work-foci and therefore less capacity to use these items. Academics conversely were less interested in webinars and demonstrations, as their primary focus involves research and not operations.
FINDINGS
The workshop highlights the benefits of open communication between fire weather research and wildland fire and fuels management communities. Key points are:
- Continued engagement benefits researchers and end-users, and should begin during project planning
- There are overwhelming amounts of data and numbers of tools, but also an overall lack of key information needed to transfer those products to use in the field
- Fire weather forecast reliability and precision thresholds for decision making change depending on lead-time, as management objectives and impacts are different
- Fire weather data need to be easily interpretable for public communication and education in addition to management decision making
- Adoption of new products and tools relies heavily on directly applicable use cases, comparison to existing products, and training opportunities
- There is a need to combine different forecast products across timescales into a single resource for easy cross-comparison and ease of use
- It is common for the wildland fire weather community to have multiple work foci and therefore different and specialized forecasting needs
- Prediction of lightning and information on current and predicted fuel moisture are key missing factors at all time scales.
The authors would like to acknowledge all participants in the workshop. These insights are a direct result of their engagement, feedback, and efforts to improve communication in the fire weather community.
This work is supported by the National Oceanic and Atmospheric Administration Office (NOAA) of Oceanic and Atmospheric Research (OAR) Weather Program Office (WPO) Innovations Fire Weather Competition, Award #NA23OAR4590383.
Samantha Kramer is applied atmospheric scientist with expertise in wildfire smoke, aerosol pollution, climatology, and sub-seasonal prediction. Kramer leads cross-sector research focused on improving fire weather data accessibility and evaluating forecast products for decision-maker use and provides scientific analysis and technical support for federal and private litigation. Kramer’s work emphasizes collaboration across agencies and institutions to translate complex science into meaningful outcomes for public health, policy, and environmental planning. Kramer is the founder and principal scientist at the Atmospheric Interdisciplinary Research Institute, a small women-owned independent research group.
Brian Potter is a research meteorologist with the US Forest Service, studying fire-atmosphere interactions for the US Forest Service since 1994. Potter serves on the Interagency Wildfire Air Quality Program’s Air Resource Advisor Teaching Cadre and the Pacific Northwest Region’s Teaching Cadre for Advanced Fire Behavior Calculations. Potter was a member of the USDA Forest Service National Prescribed Fire Program Review team. His past research has focused on atmospheric processes associated with wildfire growth, including the Hot-Dry-Windy Index and Fires of Unusual Size. Potter’s current research focuses on weather needs for safe and successful prescribed burning, serving as the research group leader for the Central Washington Pilot Project.
Nate Benson is the senior fire ecology and fuels specialist at Spheros Environmental. Benson’s 38-year career with the National Park Service spanned field operations at Glacier, Yellowstone, Great Smoky Mountains, and Everglades national parks, focusing on prescribed fire and wildfire management. Benson served at the National Interagency Fire Center as the NPS national fire ecologist and program lead for wildland fire science and ecology. Benson was instrumental in developing the Monitoring Trends in Burn Severity (MTBS) Program and FEAT/ FIREMON Integrated software application; he also chaired the Joint Fire Science Program (JFSP) governing board.
Crystal McClure is an atmospheric scientist and wildfire expert with more than 15 years of experience in scientific research and project management. McClure is adept at analyzing complex data and developing innovative solutions to address environmental challenges. McClure is a principal scientist at the Atmospheric Interdisciplinary Research Institute.
ShihMing Huang is the smoke management specialist for the U.S. Wildland Fire Service headquarters, addressing wildland fire smoke issues for the Department of the Interior since 2024. Previously at Sonoma Technology, Huang developed and conducted projects and research that advanced the understanding of wildfire and prescribed fire smoke emissions and their impacts on air quality and public health for 16 years.
Megan Schill is a scientific and technical writer with more than 15 years of experience communicating complex environmental and scientific information for technical, regulatory, and public audiences. Schill’s work has focused on the synthesis and presentation of multidisciplinary research, requiring expertise in technical editing, scientific communication, and the translation of complex data into clear, precise language.
