Investigating temporal and spatial variability of wildfire potential with the use of object-based image analysis of downscaled global climate models

Abstract

Prevailing natural and climatic conditions play a major role in determining wildfire potential. The aim of this study was to assess drought conditions in Lebanon based on climatic variability to gain a better understanding of the temporal and spatial changes in wildfire potential. The specific objectives were to: 1) estimate wildfire potential with the use of spatial climatic data, and 2) investigate temporal and spatial variability of wildfire potential. To meet these objectives we developed monthly Keetch‐Byram Drought Index (KBDI) maps for all Lebanon by employing monthly rainfall and temperature data downscaled from global climatic data, and employed an object‐based approach to analyze temporal and spatial variability of the monthly KBDI maps. Results are documented in spatially explicit maps of different degrees of wildfire potential including five classes based on KBDI values: very low, low, moderate, high, and very high wildfire potential on a monthly basis. In addition, the mapping allowed the identification of monthly wildfire spatial variability in function of altitude and landcover. This modelling effort provides a better understanding of potential wildfire temporal and spatial variability, information critical for developing effective fire prevention policies and adaptation strategies.