RELATIONSHIP BETWEEN SATELLITE MICROWAVE RADIOMETRIC DATA, ANTECEDENT PRECIPITATION INDEX, AND REGIONAL SOIL-MOISTURE

Satellite microwave brightness temperatures (T(B)'s) have been shown, in previous studies for semi-arid environments, to correlate well with the antecedent precipitation index (API), a soil moisture indicator. The current study, using the Special Sensor Microwave/imager (SSM/I), continued this work for parts of the U.S. Corn and Wheat Belts, which included areas with a more humid climate, a denser natural vegetation cover, and a different mix of agricultural crop types. Four years (1987-1990) of SSM/I data at 19 and 37 GHz, daily precipitation and temperature data from weather stations, and API calculated from the weather data were processed, geo-referenced, and averaged to 3/4-degrees latitude-longitude grid quadrants. Correlation results between T(B) at 19 GHz and API were highly dependent on geographical location. Correlation coefficients (r values) ranged from -0.6 to -0.85 for the semi-arid parts of the study area and from -0-3 to -0.7 for the more humid and more densely vegetated parts. R values were also higher for the very dry and very wet years (-O.5 to -0.85) than for the 'normal' year (-0.3 to -0.65). Similar to previous results, the Microwave Polarization Difference Index (MPDI), based on the 37 GHz data, was found to correspond to variations in vegetation cover. The MPDI was used to develop a linear regression model to estimate API from T(B). Correlation between estimated and calculated APIs was also geographically and time dependent. Comparison of API with some field soil moisture measurements showed a similar trend, which provided some degree of confidence in using API as an indicator of soil moisture.