Evaluation of surface flux maps over the southern Great Plains in IHOP

Investigations of land-atmosphere interactions over heterogeneous terrain require estimates of the spatial variability of surface fluxes, including sensible and latent heat. Direct measurements of fluxes over relevant spatial and temporal scales, tens to hundreds of kilometers and multiple hours, is not possible. Remote sensing of the land surface, reanalysis of precipitation data, and numerical models of soil-vegetation-atmosphere interactions provide tools for creating such estimates, but evaluation of these products is challenging. We have used data from the 2002 International H2O Project (IHOP) to evaluate two advanced methods for estimating surface fluxes over large regions. One approach, the Atmospheric Land-Exchange Inversion (ALEXI), diagnoses fluxes using a combination of meteorological observations, satellite radiometric temperature, a two-stream land-surface parameterization, and a simple model of boundary layer growth. The High Resolution Land Data Assimilation System (HRLDAS) is full land surface model (LSM) that assimilates radar precipitation data to calculate soil moisture and land surface fluxes. We have compared both of these products during the IHOP study period, and used extensive ground-truth data (nine surface flux towers and three 60 km flux aircraft tracks) to assist in diagnosing differences between these products. The comparison encompasses nearly 300x300 km2 of the southern Great Plains. Strengths and weaknesses of both HRLDAS and ALEXI are described. General agreement among the models and observations is encouraging, and suggests that studies of land-atmosphere interactions over mesoscale domains can be constrained with surface flux estimates based on remote sensing and data assimilation products.