Geographical analysis of the factors influencing pyrocumulonimbus and their regional differences over temperate southeast Australia

The frequency and intensity of pyrocumulonimbus (pyroCb) have escalated in the context of global warming, making a robust understanding of the historical and future patterns of pyroCb occurrence increasingly critical. Previous studies have focused on describing or simulating the conditions for pyroCb development, lacking statistical analysis of the drivers and occurrence patterns of pyroCb at larger scales. In this study, atmospheric conditions, surface fire weather variables, topography and vegetation type previously associated with pyroCb were extracted from 1980 to 2020 over temperate southeastern Australia. Utilizing both global logistic regression and geographically weighted regression (GWR) models, the study tested a selection of predictors previously associated with pyroCb, ranked their relative importance and assessed regional differences. This study reveals significant regional variations in pyroCb drivers, demonstrating considerable spatial heterogeneity in the influence of different variables on pyroCb occurrence. Overall, atmospheric variables provide the best explanation of pyroCb occurrence patterns, followed by topographic variables, with the continuous Haines (C-Haines) index serving as the most important factors influencing pyroCb occurrence. These findings deepen our understanding of the factors influencing pyroCb occurrence, providing valuable insights for future risk assessment and strategic disaster management.