Search

Accurate estimation of stored water by Land Surface Process (LSP) models is crucial to the prediction of continental weather and near-term climate by General Circulation Models (GCMs). This dissertation represents an important step toward assimilating the satellite radiometric observations to improve the soil moisture estimates. It consists of forward modeling of terrain brightnesses through a biophysically-based Land Surface Process/Radiobrightness (LSP/R) model, and correlating ground-based brightnesses with those from the Special Sensor Microwave/Imager (SSM/I). The LSP/R model was modified and calibrated for representative terrain in the Great Plains during summertime, when the surface processes are dominant and strongly coupled. The calibration used data from two collaborative field investigations, the fourth and the fifth Radiobrightness Energy Balance Experiments (REBEX-4 and REBEX-5). REBEX-4 was a collaboration with the Atmospheric Environment Service (AES), Canada, at the USGS EROS Data Center near Sioux Falls, SD. During the experiment, we observed microwave emission and concurrent micro-meteorological parameters at a bare soil and a nearby grass site from June-September in 1996. REBEX-5 was the University of Michigan's contribution to an extensive field investigation, Southern Great Plains Hydrology Experiment (SGP'97), conducted in north central Oklahoma from June 18--July 17 in 1997. During REBEX-5, we measured brightnesses of senescent winter wheat and after harvest wheat-stubble. In general, the calibrated LSP model predicted realistic surface processes that compared well with the field observations. The model predictions were most sensitive to shortwave albedo of the terrain and soil thermal and hydraulic conductivities. The Radiobrightness module captured the mean diurnal variations in brightnesses. The H-pol terrain brightnesses at 19 GHz were more sensitive to soil moisture and roughness than the V-pol brightnesses. The comparison of the EASE-Grided SSM/I and the ground-observed brightnesses at 19 and 37 GHz demonstrated that the satellite brightnesses were sensitive to terrain moisture and temperatures at both frequencies. The brightnesses at 19 GHz were unaffected by scattering and can potentially be assimilated to improve soil moisture estimates, particularly for regions with low vegetation biomass (2).