Assessing interannual variability of evapotranspiration at the catchment scale using satellite‐based evapotranspiration data sets

Abstract

The catchment water‐energy balance at the interannual scale remains a challenge for bridging the important gap in our knowledge of the hydrologic cycle. This study investigates interannual evapotranspiration (ET) variability and water‐energy balance at over 547 catchments across the contiguous United States in different climate zones. The investigation is based on ET data estimated from satellite images and surface daily meteorological data during the period of 1983–2006 provided by the University of Montana. We find that the interannual relationship between annual potential ET to annual precipitation (PET/P) and ET/P, the energy, and water factors defined with the Budyko curve framework can be captured by a linear function, with an average goodness of fit of 0.928 over all the catchments. The linear relationship is validated by another independent remote sensing ET product generated by the University of Washington and is demonstrated to be rational using regression diagnostics. Additionally, a comparison study using a water balance method and the Budyko curve is undertaken to examine the linear relationship. In addition to water supply and energy supply, the primary controls on evapotranspiration, soil water storage, response of vegetation to climate variability, and human interference are also major factors of the interannual relationship between PET/P and ET/P.