Developing Satellite-derived Estimates of Surface Moisture Status

Abstract

Recent research has shown that the combination of spectral vegetation indices with thermal infrared observations may provide an effective method for parameterizing surface processes at large spatial scales. In this paper, we explore the remotely sensed surface temperature (Ts)/normalized difference vegetation index (NDVI) relationship regarding a) influence of biome type on the slope of Ts/NDVI, b) automating the definition of the relationship so that the surface moisture status can he compared with Ts/NDVI at continental scales. The analysis was carded out using 1) NOAA Advanced Very High Resolution Radiometer (AVHRR) data over a 300-km ? 300-km area in western Montana under various land-use practices (grass, crops, and forests), 2) Earth Resources Observations Systems Data Center continental United States biweekly composite AVHRR data. A strong negative relationship was observed between NDVI and Ts over all biome types. The similarity of the Ts/NDVI relationships over different biomes indicated that fraction of vegetation cover has strong influence on the spatial variability of Ts. A substantial change in the Ts/NDVI relationship was observed over forests between wet and dry days. In comparison, no change was observed over irrigated crops. Results from the automated approach agreed well with those using manual selection. At continental scales, the slope of Ts/NDVI is strongly correlated to crop-moisture index values indicating that Ts/NDVI relation is sensitive to surface moisture conditions. Upon further development, this relationship may be useful for parameterizing surface moisture conditions in climate models, decomposition studies, and fire weather monitoring.