Modelling forest fire danger in Lebanon with the combined use of socio-economic and biophysical variables in object-based image analysis

Abstract

Like many other countries in the Mediterranean, the occurrence and spread of forest fires in Lebanon are related to human activities. More specifically, landcover and land use changes (e.g., conversions of lands, abandonment of land and accumulation of fuel) driven by socio-economic changes occurring in the country have increased the probability of occurrence and spread especially in the Wildland-Urban Interface. The aim of this work was to model the influence of both socio-economic and biophysical variables on fire occurrence in Lebanon. The specific objectives were to 1) analyze socio-economic and biophysical drivers of forest fires, and 2) use object-based image analysis to derive a spatially explicit probability of fire occurrence across the country. Forward stepwise binary logistic regression analysis of 24 socio-economic and biophysical variables was used to predict wildfire occurrence. Spearman correlation analysis was conducted in order to eliminate multi-collinearity between selected variables. Eighty percent of the total number of administrative units was randomly selected for use in the development of the modelling, while the remaining 20% of units were used for testing and validating the final model. Object-based image analysis was used to map the spatial distribution of fire occurrence by modelling socio-economic and biophysical drivers including land-cover and land-use changes. The final map showed 5 different fire danger classes ranging from very low to very high. The quality of the classification results was evaluated and underand overestimations errors of fire occurrence were mapped. The accuracy of the fire occurrence mapping model was approximately 85% when tested on the validation data set. The probabilistic spatial output of the fire threat model was considered satisfactory given the challenges of using multi-source data in an object-based image analysis approach. Results suggest increasing the resolution of socio-economic data would improve mod.