|Albertson, John - UNIV OF VA, VA|
|Cahill, Anthony - TEXAS A&M UNIV|
Submitted to: International Association of Hydrological Science
Publication Type: Proceedings
Publication Acceptance Date: June 21, 2000
Publication Date: September 25, 2001
Interpretive Summary: Remotely sensed surface temperature from satellites has the potential of serving as a key boundary condition for models simulating plant water use. However, use of remotely sensed surface temperature in these models has not been successful in general because of discrepancies between model simulated versus remotely sensed surface temperatures. Field observations, collected during the 1999 Southern Great Plains Experiment over four field sites comprising the major land cover types for this region, are used to study this phenomenon. The results suggest that there are fundamental differences between simulated and remotely sensed surface temperatures which need to be accounted for before remote sensing surface temperature observations can be used to adjust simulation model predictions. Methods to account for model and observed differences in surface temperature could provide more reliable plant water use estimates for large areas using satellite data.
Technical Abstract: The relationship between modeled and radiometrically observed surface temperature is explored in this paper. Implied in many assimilation attempts is an equivalence in the definition of modeled and observed state variables. We contend and attempt to demonstrate here that this assumption is not of general validity. The analysis is conducted with a Soil-Vegetation-Atmosphere Transfer model and field data collected during the Southern Great Plains Experiment of 1999. It is shown from supporting equations that the modeled temperature evolves in a dynamic equilibrium with the fluxes where the radiative flux depends strictly on a radiometric temperature. The sensible and latent heat fluxes depend on an aerodynamic temperature resulting in a hybrid modeled temperature. It is shown that enforcing an equality between modeled and observed temperatures may result in a logical deterioration in model estimates of latent and sensible heat fluxes.