Wildland fire risk quantification for prescribed fire management in the Great Plains

Abstract

Wildfire activity in the Great Plains has increased over the past three decades. As a fire-dependent ecosystem, the region relies on prescribed fire to sustain the grassland structure, improve forage quality, support biodiversity, and reduce wildfire risk posed by antecedent fuel accumulation. Despite its ecological importance, prescribed fire application remains constrained by safety concerns and lack of tools for accurate fire risk assessment. Existing fire danger indices often focus on forests or have coarse spatial resolution, limiting their use for operational-scale grassland fire planning. This study quantified and modeled wildland fire risk in the Great Plains using historical local weather, fine fuel, and wildfire data from 1992 to 2020. A composite fire risk framework was developed, defining wildland fire risk as the product of seasonal and daily fire risks. The seasonal fire risk was estimated as the ratio of smoothed multi-year average daily burned area to the total land area and was modeled using multi-year average daily grass curing and multi-year average daily air temperature changes as predictors representing fuel phenology and climatic conditions, respectively. The seasonal fire risk model explained 97% of the variability in seasonal fire risk. Seasonal fire risk in the southern Flint Hills counties showed a bimodal pattern, with a primary peak in March and a secondary peak in October, covering 0.01% of the land area. In spring, seasonal fire risk decreases rapidly from its March peak of 0.034% of the land area. This decline coincides with the traditional April period for intensive prescribed burning, highlighting the importance of seasonal fire risk assessment. The daily fire risk was represented as the ratio of wildfire daily burned area to the smoothed multi-year average wildfire daily burned area, and was modeled using maximum sustained wind speed and dead fuel moisture as predictors. The daily fire risk was estimated as the ratio of wildfire daily burned area to the smoothed multi-year average wildfire daily burned area, and was modeled using maximum sustained wind speed and dead fuel moisture as predictors. The model explained 15% of the variability in daily fire risk. Analysis of wildland fire risk during the spring burn window (March 8 - May 8) over 15 years in the Flint Hills reveals that the quantified risk ranged from 0.005% to 0.088%, representing an approximately 18-fold difference in burning potential. Eleven days were found within the defined low-risk level (≤ 0.01%), during which optimal prescribed burning can be safely conducted, and 28 days were within the marginal range (0.01% - 0.18%), during which prescribed burning requires utmost caution. Spring’s peak seasonal risk (0.034%) was three times more than the fall seasonal risk peak (0.01%), which lies within the defined safe range of wildland fire risk. This pattern indicates that the fall season offers opportunities to implement prescribed burns at lower daily fire risk. The consistent distribution of fire size, frequency, and fuel characteristics across the Flint Hills offers an opportunity to generalize the framework prospectively across the Great Plains. These results provide a data-driven basis for selecting optimal prescribed burn days that consider fuel conditions and daily fire-weather variability. This approach establishes a foundation for continued efforts to develop localized decision-support tools for prescribed fire planning and wildfire management in the Great Plains.