Wildfire Ignition in the Forests of Southeast China: Identifying Drivers and Spatial Distribution to Predict Wildfire Likelihood

Futao Guo,Zhangwen Su,Guangyu Wang,Long Sun,Fangfang Lin,Aiqin Liu

DOI: https://doi.org/10.1016/j.apgeog.2015.11.014

IF: 4.732

2015-01-01

Applied Geography

Abstract:Understanding the spatial distribution and driving factors of forest fire facilitates local forest fire management planning and optimization of resource allocation for fire prevention geographically. In this study, we analyzed the spatial pattern and drivers of forest fire in Fujian province, southeastern China, during 2000–2008 using Ripley's K-function and logistic regression (LR) model. The likelihood of fire occurrence was mapped based on the resultant model. The data regarding fire ignitions, weather conditions, vegetation, topography, infrastructure, and socioeconomic factors were extracted from ArcGIS environment. The study revealed that fire ignition was mainly clustered in space due to the comprehensive influence of different factors. Elevation, daily precipitation, and daily relative humidity were negatively associated with fire ignitions, whereas distance to settlement, population density, and per capita gross domestic product (GDP) impacted fire occurrence positively. The spatial distribution of fire occurrence likelihood was highly variable in Fujian: high fire likelihood was prevalent in the northern and southeastern parts of Fujian, whereas it was relatively low in the western province. Fire risk may be underestimated in some areas of Fujian according to the spatial patterns of the model residual, which should be paid more attention to in the forest fire management practice.