Regional-scale maps of surface fluxes are being generated in realtime (daily) using the Atmospheric Land-EXchange Inverse (ALEXI) model at the University of Wisconsin, Cooperative Institute for Meteorological Satellite Studies (CIMSS). The ALEXI model is driven by realtime satellite-based inputs including GOES-derived land surface brightness temperature changes and AVHRR NDVI. Surface synoptic data, tropospheric soundings from regional-scale analyses and land cover characteristic information are used in conjunction with the satellite information to generate the flux components of the surface energy budget (sensible, latent, ground heat), along with net radiation. The key attribute of ALEXI is an estimation of near-surface air temperature from the combination of a Two-Source surface model and a simple planetary boundary layer model. The ALEXI model has proven itself by working well over a range of land surface and vegetation cover regimes. The ALEXI is designed to be insensitive to biases caused by incorrect atmospheric corrections and surface emissivity specifications, and accomodates off-nadir sensor view andgles.

A collaboration between CIMSS, the UW Dept. of Soil Science and the University of Oklahoma focuses on the validation of the ALEXI flux estimations by comparing and correcting them against in situ fluxes obtained at 5-minute frequencies from the 114 station Oklahoma Mesonet. The Mesonet data for mostly clear days during summer 2000 and 2001 across the Southern Great Plains region are being compared directly with the 5 km-averages obtained from ALEXI, as well as more indirectly when the 5 km ALEXI averages are disaggregated to 30 meter resolution-fluxes (of individual Mesonet stations) with the help of LandSat-7 thermal data.

Presented in the conference talk with be an overview of this research, our success with the validation and disaggregation components of this work, as well as with an alternate (statistical interpolation) approach to obtaining fluxes using the ALEXI model.