Forest fire risk change assessment model based on the improved G1-CRITIC method

Forest fires present a significant threat to both ecosystems and socio-economic systems, making accurate risk assessment essential for effective prevention and control. However, conventional assessment approaches are often limited by subjective bias, incomplete indicator frameworks, or a reliance on static analysis, which hinders a comprehensive understanding of the multidimensional drivers and spatiotemporal evolution of fire risk. To address these gaps, this study proposes an integrated forest fire risk assessment model that incorporates natural, social, and economic dimensions. An improved G1–CRITIC combined weighting method is applied to determine indicator weights, effectively balancing expert judgment with objective data metrics. The model was implemented using multi-source data from 2013 to 2022–including remote sensing imagery, government statistics, and field survey data–in Huangu Town, Ankang City, Shaanxi Province. The results show that the model reliably assesses regional forest fire risk levels and successfully delineates their spatiotemporal dynamics. Correlation analyses further confirm the robustness of the model. The study reveals that forest fire risk is co-driven by natural and socio-economic systems: natural factors predominantly influence short-term fluctuations, whereas socio-economic factors shape long-term trends. By overcoming several limitations of traditional methods, the proposed model enables a dynamic and multi-factor driven analysis of forest fire risk. These findings offer a scientific foundation for phased, zoned, and graded precision management of forest fire prevention, as well as for the optimized allocation of firefighting resources.