83rd Annual

Thursday, 13 February 2003
The Atmosphere Land-EXchange Inverse (ALEXI) model: Regional-scale flux validations, climatologies and available soil water derived from remote sensing inputs
John R. Mecikalski, CIMSS/Univ. of Wisconsin, Madison, WI; and M. C. Anderson, R. D. Torn, J. M. Norman, G. R. Diak, S. J. Richardson, and J. A. Brotzge
Poster PDF (273.2 kB)
Continental scale maps of real-time daily surface energy fluxes are being generated using the Atmospheric and-Exchange Inverse (ALEXI) model at the University of Wisconsin. This model uses GOES-derived surface brightness temperature changes, AVHRR-derived land cover properties and synoptic weather data. Sensible, latent and ground heat flux components along with net radiation are calculated on the 5-10 km scale from these inputs. A collaboration between the CIMSS, the UW Dept. of Soil Science and the University of Oklahoma is focusing on the validation of these ALEXI flux estimates by comparing them to in situ fluxes obtained from the 114 station Oklahoma Mesonet. Using LandSat-7 derived NDVI and thermal data, we are able to disaggregate these ALEXI fluxes to 30 meter resolution to get a better estimate of the true station surface fluxes using the model.

Direct validation of ALEXI to the approximately 1500 Mesonet flux sites over 27 days during 2000 and 2001 has been performed on 5 km scales with very promising results. In addition, the disaggregation validation focuses on four days from 2000 and 2001, each over a different geographical area of Oklahoma. We chose to validate the 30 meter fluxes against the 10 OASIS "SuperSites" that measure all four fluxes directly. Plots of surface fluxes over a 5 km ALEXI box are very heterogeneous, yet the average of these 30 m fluxes is consistent with the ALEXI 5 km flux. The 30 meter scale fluxes compare more favorably to the mesonet station fluxes then the 5 km scale fluxes. Errors in our disaggregation methods can be attributed to flux footprints that were advected by the mean wind over these areas, and when the LandSat average radiometric temperature over a GOES pixel and the GOES radiometric temperature do not match. Future disaggregation work will focus on using 250 meter MODIS derived NDVI and thermal data as inputs rather then LandSat.

In addition to the ALEXI model validation, our presentation will also highlight flux climatologies formed from the daily ALEXI inputs, as well as ALEXI-derived root zone and surface available soil water. Other recent improvements to the ALEXI program will also be discussed.

Supplementary URL: