The impact of land surface heterogeneity on the accuracy and utility of spaceborne soil moisture retrievals

The sharp contrast between the fine spatial scales at which surface soil moisture fields exhibit heterogeneity and the coarse resolution (>50 km) at which current microwave antennae technology allows for observation of soil moisture from space poses a challenge for efforts to develop, and effectively articulate the need for, a spaceborne observational capacity for soil moisture. Using predictions from the TOPMODEL-based Land-Atmosphere Transfer Scheme (TOPLATS) and data from the 1997 Southern Great Plains Field Experiment (SGP'97) this dissertation examines the impact of land surface heterogeneity, and the inability to adequately represent it, on efforts to retrieve accurate measurements of surface soil moisture from space and use these measurements to make large-scale surface energy flux predictions within the land surface component of a weather prediction model.; Chapter 3 examines the first issue by quantifying the degree to which land surface heterogeneity, in concert with nonlinearities in the soil moisture retrieval process, is capable of inducing retrieval errors in coarse-scale soil moisture products derived from the Advanced Microwave Scanning Radiometer (AMSR-E). Even if large-scale representations of soil moisture are free from error, nonlinear relationships between soil moisture and various land surface fluxes may make coarse-scale representations of soil moisture inadequate for the prediction of coarse-scale energy fluxes. Chapters 4 and 5 examine this possibility and describe the impact of aggregating soil moisture information up to grid scales consistent with the resolution expected in next generation passive microwave sensors. While the impact of soil moisture spatial aggregation appears large, Chapter 6 demonstrates the feasibility of developing effective closure strategies for nonresolved soil moisture variability. One such strategy, a downscaling closure strategy based on an assumed power-law relationship between soil moisture statistical second moments and scale, is shown to be capable of significantly improving large-scale TOPLATS predictions of energy fluxes derived from coarse-scale soil moisture data.; Results suggests that, despite the potentially large impact of nonlinearities in microwave emission and land surface processes, coarse spatial resolutions will not necessarily prevent spaceborne soil moisture retrievals from aiding weather prediction within the United States southern Great Plains.