Comparison between Post-Fire Analysis and Pre-Fire Risk Assessment According to Various Geospatial Data

Wildfires in forest ecosystems exert substantial ecological, economic, and social impacts. The effectiveness of fire management hinges on precise pre-fire risk assessments to inform mitigation efforts. This study aimed to investigate the relationship between predictions from pre-fire risk assessments and outcomes observed through post-fire burn severity analyses. In this study, forest fire risk was assessed through the Fuzzy Analytical Hierarchy Process (FAHP), in which fire-oriented factors were used as input. The degree of burn was determined by the Random Forest method using 11,519 training points and 400 test points on Sentinel-2 satellite images under three different classes. According to the results obtained from 266 selected test points located within the forest, all primary factors put forth increased high burn severity. Climate, in particular, emerged as the most significant factor, accounting for 52% of the overall impact. However, in cases of high fire severity, climate proved to be the most effective risk factor, accounting for 67%. This was followed by topography with 50% accuracy at a high fire intensity. In the risk assessment based on the FAHP method, climate was assigned the highest weight value among the other factors (32.2%), followed by topography (27%). To evaluate the results more comprehensively, both visually and statistically, two regions with different stand canopy characteristics were selected within the study area. While high burn severity had the highest accuracy in the Case 1 area, moderate burn severity had the highest in the Case 2 area. During the days of the fire, the direction of spreading was obtained from the MODIS images. In this way, the fire severity was also interpreted depending on the direction of fire progression. Through an analysis of various case studies and literature, this research underlines both the inherent strengths and limitations of predicting forest fire behavior-based pre-fire risk assessments. Furthermore, it emphasizes the necessity of continuous improvement to increase the success of forest fire management.