Estimation of live fuel moisture and soil moisture using satellite remote sensing

Vegetation and soil are primary terrestrial components of the Earth's surface. Estimation of moisture content in vegetation and soil using remote sensing measurements is very important for fuel management, fire danger assessment, fire behavior analysis, drought monitoring, and climate and weather analyses. Research has been conducted for several years to retrieve live fuel moisture and soil moisture values using either microwave or optical/IR remote sensing measurements. Most existing approaches rely on the identification of empirical relations between fuel/soil moisture and spectral indices. For live fuel moisture retrieval, I investigated the inverse problems of the leaf model PROSPECT, and identified the requirements for leaf parameter inversion using MODIS measurements. I proposed a semi-physical approach for live fuel moisture content estimation through approximate inversion of PROSPECT. Both model simulation and ground validation demonstrated the advantages of my approach over previous approaches using spectral indices. For soil moisture retrieval, I used MODIS optical/IR measurements by applying the universal triangle method with SCAN measurements for Mississippi state. Ground validation demonstrated a good match between MODIS retrieved soil moisture and SCAN measurements.; I also investigated the estimation of real-time fuel moisture and soil moisture using MODIS Direct Broadcasting measurements, identified critical components and proposed a systematic approach towards near real-time applications for estimating soil moisture and fuel moisture retrieval.; In addition, future studies are also discussed, especially the correction of the BRDF effects in real-time remote sensing measurements, development of soil moisture and fuel moisture algorithm for the next generation sensors, and the cross-sensor data continuity problem.