Abstract:

Objective: Based on the burn probability model, this study aims to uncover the spatial patterns of seasonal forest fire risks, and optimize the allocation of forest fire suppression resources, so as to provide theoretical support and practical guidance for improving the firefighting resource management in northern forests and enhancing forest fire prevention and control capabilities. Method: The Daxing’anling Mountains, an important forest area in China, was targeted. An enhanced burn probability model was employed to simulate the spatiotemporal dynamics of burn probabilities across the three distinct fire-risk seasons: spring, summer, and autumn for Daxing’anling forest region. Extensive iterative simulations were used to quantify seasonal variations in burn probabilities (BP), fire spread rates (ROS), and fire intensities (FI), and thereby assess the initial attack success rate. Two simulation scenarios were developed for comparative analysis: a baseline scenario reflecting current firefighting resource allocations and an optimized scenario that adjusted the placement and amount of temporary ground firefighting bases according to BP distribution. Resource reallocation strategies were designed to relocate temporary firefighting bases from areas of low BP areas to high BP areas. The resource allocation was strengthened in deciduous coniferous forests in the eastern and southern regions during spring, and the resource allocation was strengthened in mixed forest zones in autumn. One-way repeated measures ANOVA was used to examine seasonal differences in BP, ROS, and FI, and paired-sample t-tests was used to validate the efficacy of pre- and post-adjustment outcomes. Key evaluation metrics encompassed response time, initial attack success rate, and burn probability, providing a comprehensive framework for assessing the impact of resource optimization on fire management effectiveness. Result: There were significant seasonal variations in forest fire risks in Daxing’anling region, with the burnt area in spring accounting for over 80% of the annual burnt area. The combination of dry fuel conditions (fine fuel moisture code, FFMC > 85) and elevated fire weather index (FWI) during spring significantly amplified fire spread, whereas precipitation in summer and autumn exerted a strong suppressive effect on fire risks. The model simulations revealed that the average BP in spring was 0.002 6, with 94.6% of the area exhibiting potential for combustion. During this period, the ROS peaked at 6.3 m·min^?1, and FI reached its highest annual value of 4 504.6 kW·m^?1. In contrast, summer showed a substantial decline in BP to an average of 0.000 9, which is 73.6% lower than that in spring, accompanied by decreases in ROS to 2.6 m·min^?1 and FI to 1864.1 kW·m^?1. Autumn experienced the lowest fire risk, with only 20.3% of the area at risk of burning, with an average BP of 0.000 1, and further reductions in ROS to 2.4 m·min^?1 and FI to 1 512.7 kW·m^?1. Following the dynamic reallocation of temporary ground firefighting bases informed by BP simulation outcomes, the number of temporary firefighting bases in spring, summer, and autumn was reduced by 14.4%, 27.1%, and 33.3%, respectively. Despite these reductions, the response times continued to meet the critical threshold of <1.5 hours, while the initial attack success rate was sustained within the range of 77.2%?93.5%. The optimization of temporary ground firefighting resource allocation resulted in no statistically significant changes in burn probability (P = 0.84) or fire behavior (P = 0.91), underscoring the efficacy of the adaptive management strategy. Conclusion: This study can quantify the spatial differentiation patterns of seasonal forest fire risks in Daxing’anling and proposes a dynamic adjustment approach for optimizing the allocation of limited ground firefighting resources based on the BP simulations. The approach has successfully achieved the dual objective of reducing the number of temporary firefighting bases by 14%?33% while maintaining established firefighting goals.

Key words: burn probability model, wildfire behavior simulation, wildfire management, fire behavior, Daxing’anling