Testing Remotely Sensed Evapotranspiration Estimates Using Airborne and Ground Measurements
Cressida L. Savige, Andrew N. French, Andrew W. Western, Jeffrey P. Walker, Mohamad Abuzar, Jorg M. Hacker, and Jetse D. Kalma
Evapotranspiration (ET) is a key hydrologic flux that is rarely monitored. Remotely sensed measurements of land surface characteristics, including surface temperature, in combination with surface energy balance models provide an opportunity to improve the frequency and accuracy of ET estimates. Here we compare results from two ET estimation models with ground and airborne measurements of latent and sensible heat fluxes from an irrigation district in South-Eastern Australia during January 2003 (mid-summer). The irrigation district contains a mixture of wet and dry patches at the scale of hundreds of metres. Ground data collected during the field campaign include eddy covariance measurements of latent and sensible heat flux at two sites, a variety of associated meteorological measurements and detailed soil moisture and vegetation (NDVI) patterns about these sites. The airborne data include eddy covariance measurements of latent and sensible heat flux, line measurements of surface temperature, and detailed (1m resolution) NDVI images over the field sites. Landsat ETM+ and ASTER images are available during the field campaign. ET estimates from these images are made with a two-source model and the Surface Energy Balance Algorithm for Land (SEBAL). Results of comparisons of these with the ground and airborne data will be presented and the models will be intercompared. Some issues with the spatial averaging techniques for the airborne flux data will be discussed.