Combining a Two-Source Patch Model with Satellite Data to Monitor Daily Evapotranspiration at a Regional Scale

In this work, we present a micro-meteorological approach for estimating surface energy fluxes that can be operationally used together with satellite images to monitor surface energy fluxes at a regional scale. In particular we will focus on the retrieval of daily evapotranspiration. The feasibility of the model is explored at a local scale using data collected over a maize crop in Beltsville, Maryland, USA, and a boreal forest in Sodankyla, Finland. Comparison of the results with ground measurements shows errors between plusmn15 and plusmn50 W m^-2 for the retrieval of net radiation, soil heat flux, and sensible and latent heat fluxes in both sites. A methodology to apply the model to Landsat imagery is presented. Finally, we show its application to three different Landsat scenes covering the whole Basilicata region (Southern Italy). Maps of daily evapotranspiration are generated. Results are compared with ground measurements. Accuracy close to plusmn30 W m^-2 is obtained.