Validating surface energy balance fluxes derived from airborne remote sensing

Remote sensing-derived energy balance components were compared against measured eddy covariance energy balance terms using heat flux source area models to validate the airborne multispectral remote sensing procedure in the estimation of instantaneous and daily evapotranspiration rates.; A procedure was developed to generate raster layers of the footprint weights for weighting/integrating the different components of the energy balance model and obtain meaningful comparisons to similar energy balance terms measured at eddy covariance and/or Bowen ratio stations.; Soil heat flux and surface aerodynamic temperature models were studied in an effort to improve the remote sensing estimation of distributed evapotranspiration rates. Aerial and ground data were acquired over a riparian corridor (Salt Cedar, Tamarix grove), soybean and cornfields (rainfed crops) in different ecosystems.; The results confirmed that net radiation is well estimated with the remote sensing technique showing an estimation error of only -4.8 +/- 20.7 W m-2, (-0.5 +/- 3.6%). Linear and exponential soil heat flux models were found to correlate strongly to leaf area index and net radiation. The surface aerodynamic temperature term in the sensible heat flux equation was parameterized using surface radiometric temperature, air temperature, wind speed, and leaf area index. It is suggested that the surface aerodynamic temperature model be tested for a wide range of vegetation types, atmospheric stability conditions, surface heterogeneity, and ecosystems to assess the model limitations.; The flux source area footprint model "FSAM" integrated heat flux pixels that compared better to measured values and it is recommended as a standard procedure to compare airborne remote sensing-derived heat fluxes against measured fluxes by eddy covariance systems; when compared to the "FASOWG" footprint model and simple arithmetic averages.; Finally, the method that uses alfalfa reference daily evapotranspiration in extrapolating instantaneous remote sensing evapotranspiration estimates to daily evapotranspiration rates performed very better when compared to measured daily evapotranspiration rates by eddy covariance systems; estimation error of just -0.2 +/- 0.7 mm d-1. This finding, along with the parameterized soil heat flux, surface aerodynamic temperature, and the development of procedure to apply footprints to heat flux images, validates the remote sensing technique as a useful tool in the estimation of spatially distributed daily evapotranspiration.