Data-Enabled Correlation Analysis between Wildfire and Climate using GIS

Abstract

This paper presents a study of examining the statistical correlation between wildfire and weather by mining historical spatial and temporal wildfire and climate data. Large wildfires have been recently becoming more frequent, intense and destructive in the West of United States. The occurrence of wildfires can be determined by many human and natural factors, such as the availability of fuels, physical settings, and weather conditions, among which weather is of great interest and importance for wildfire forecasting. The availability of landscape fire data sets and weather data sets now enables the analysis of correlation between wildfire and weather which indicates the possibility of wildfire for given weather conditions in one region. This paper investigates the relation between wildfire and drought conditions in California and visualize the results using geographic information system (GIS) computing technology. Our data analysis findings show a high correlation between the normalized number of wildfires per forest unit area and drought severity, illustrating the potential of forecasting wildfire using weather data.