Comparing surface energy flux models using ASTER imagery over Oklahoma

Surface energy flux estimates over central Oklahoma, during the summer of 2001, are derived in two different ways from thermal infrared and visible-near infrared ASTER (Advanced Spaceborne Thermal Emission and Reflection radiometer) 90 m resolution observations. In one approach, surface flux estimates are computed using a two-source energy balance (TSEB) model, which distinguishes between soil and vegetation flux components. The benefit of TSEB is an improved surface representation over sparsely vegetated terrain, as compared to one-layer models. In the other approach, surface flux estimates are computed from the Surface Energy Balance Algorithm for Land (SEBAL) model, which computes meteorological variables using information contained within the spatial variability of convective fluxes. A major benefit of the SEBAL model is that estimates can be obtained solely from remote sensing observations. Both models compare reasonably well with surface flux measurements, but their relative sensitivities and biases are different for the same input data. Sources of these modeling sensitivities and biases are discussed.