Western Fujian has high forest coverage rate and fire frequency. However, the studies regarding to the spatial distribution and prediction model of forest fire in Fujian are relatively rare. Fire dataset from 2000 to 2010, combining with topography, climate, vegetation and population density factors were used to analyze the spatial distribution and driving factors, furthermore, to establish fire prediction model based on Ripley′s K-function and logistic regression model. The results showed that forest fires in the west of Fujian were clustered in space during 2000 to 2010. In addition, elevation, daily average temperature, daily average relative humidity, daily precipitation, fractional vegetation cover, density of population, etc were identified as main driving factors for forest fire in the western Fujian. The model prediction accuracy is close to 70% with the significantly independent variables. According to forest fire risk probability and fire danger class, westhern Fujian forest fires occurred mainly concentrated in the northwestern and central regions.
model predictionwestern Fujianforest firelogistic regression
国家科技期刊平台
NSTL
维普
万方数据
