Retrieval of real-time live fuel moisture content using MODIS measurements

Abstract

Live fuel moisture content (LFMC) is one of the most important fuel properties and a critical parameter for wildland fire danger rating estimation and fire behavior analysis. Direct ground measurement of live fuel moisture content has disadvantages of high cost and limited spatial distribution extent. This paper presents an algorithm to retrieve live fuel moisture content from multiple bands of MODIS measurements. We analyzed the physical relationship between surface reflectance and live fuel moisture content using simulated MODIS measurements of diverse leaf samples, derived approximate inversion models, and proposed a semi-physical approach for live fuel moisture retrieval employing multiple MODIS bands. Using simulated MODIS measurements, the correlation coefficients between the true LFMC and estimated LFMC with our inversion models are 0.7738, 0.8397, 0.9560 and 0.9576 respectively. For validation, we tested our inversion method with woody live fuel moisture measurements at fire weather stations in Georgia. The correlation coefficients between measured LFMC and estimated LFMC with our inversion models are 0.5727, 0.6522, 0.7551, and 0.7737 respectively. Both model simulation and station measurements demonstrated advantages of our approach in accuracy. Our study suggests the potential for near real-time applications of live fuel moisture.