Scale issues in land-atmosphere interactions: implications for remote sensing of the surface energy balance

The interaction between the land surface and the atmosphere is discussed focusing on the recent literature. Of primary importance is the issue of scale and the variation across different levels of pixel resolution for using remotely sensed data in the determination of the different components of the surface energy balance (SEB). This leads into a discussion of various methods of aggregating local scale input parameters in order to calculate larger scale surface fluxes. As an example application, airborne and satellite data are coupled to a soil-vegetation-atmosphere-transfer (SVAT) model and applied to the Southern Great Plains 1997 (SGP97) Hydrology Experiment. A physically based aggregation scheme is suggested for radiometric temperature by decomposing the temperature into vegetation and soil components and then aggregating the radiance field based on the percent cover of vegetation. Fluxes derived from the airborne data agree well with those measured with eddy covariance stations. Satellite data agrees poorly with the surface and airborne estimates. However, after aggregating the airborne data to the resolution of the satellite, the agreement is better. This suggests that the satellite data is useful at estimating the fluxes at the 1 km(2) scale, while the comparison between the 1 km(2) satellite and the surface data may be problematic at such scales. (C) 2003 Elsevier Science B.V. All fights reserved.