Author(s):
Eric L. Toman, Robyn S. Wilson, William Matt Jolly, Christine Olsen
Year Published:

Cataloging Information

Topic(s):
Fire Behavior
Simulation Modeling
Weather
Human Dimensions of Fire Management

NRFSN number: 23943
FRAMES RCS Number: 64962
Record updated: January 3, 2022

Fire weather tools, such as the National Fire Danger Rating System (NFDRS) and the Wildland Fire Decision Support System (WFDSS), have been developed to support wildland fire management decisions. However, little is known about how these tools are used in practice, the sensitivity of fire management decisions to fire weather variables, or the sensitivity of fire- weather tools to input errors. This project was designed to address these gaps in current knowledge. Objectives: This project sought to achieve four main objectives: 1) Consider how fire-weather models are used to support strategic and tactical decisions. 2) Consider the sensitivity of fire-weather tools to various sources of input error. 3) Assess the sensitivity of fire management decisions to fire-weather variables. 4) Consider implications of these results to improve support for fire management decisions. Methods: Project methods included: 1) Semi-structured interview with 27 full and part-time Incident Management Team (IMT) members, district rangers, fire management officers, regional fire management coordinators, fire management and fuels specialists, and others in the western and southern United States. 2) Leveraging concurrent work to explore the sensitivity analysis fire behavior modelling tools to explore how fire weather inputs affect model outputs. This phase identified critical fire weather inputs that heavily impact model predictions from fire modeling systems that are used in Decision Support Systems throughout the United States and this guided the development of fire management scenarios for subsequent phases of the project. While our original objectives included plans to complete a sensitivity analysis, the project was modified due to 1) a loss of the portion of project funds intended to support this analysis due to fire borrowing (funds were not returned to the project), and 2) another research team was commissioned to complete a similar analysis concurrently with our project during the same timeframe. 3) Web-based survey with federal fire managers (e.g., Fire Management Officers or FMOs) working for US Department of Agriculture Forest Service (USFS) including an embedded choice experiment designed to evaluate the influence of fire-weather and other key variables in a decision regarding whether to engage in direct or indirect attack. Key Results: Results from the interviews indicated that, in practice, models are being used differently than intended (e.g., to inform operational decisions), confidence in model outputs differs between managers and technical specialists and is influenced by both social and technical characteristics. Participants have lower confidence in precipitation and wind forecasts than humidity or general weather forecasts. In general, participants appeared to express a preference for indirect attack. Decisions to switch to direct attack was strongly influenced by the timing of the fire even (e.g., early in the fire season). Participants were more likely to decide to switch to indirect attack when the combination of variables suggested the potential for extreme fire behavior. Conclusions: Overall, these results emphasize the importance of designing decision support tools with the decision strategies used by managers in mind. Results also suggest the importance of both social and technical components in developing manager confidence in models. This emphasizes the importance of communication and relationship-building by technical specialists on IMTs. Moreover, our results suggest that wind and precipitation forecasts merit particular attention in further efforts to improve model accuracy and build confidence in existing models.

Citation

Toman, Eric L.; Wilson, Robyn S.; Jolly, W. Matthew; Olsen, Christine S. 2021. Fire modeling and social science analysis of fire managers’ use of fire weather data across the US - Final Report to the Joint Fire Science Program. JFSP Project No. 15-1-06-8. Columbus, OH: Ohio State University. 20 p.

Access this Document