LONG-TERM DAILY FIELD-SCALE EVAPOTRANSPIRATION ESTIMATION USING MULTI-SATELLITE DATA FUSION IN AN INTENSIVELY DRAINED AGRICULTURAL AREA IN SOUTH DAKOTA, USA

Subsurface tile drainage is a widely used agricultural practice in Midwestern USA to remove excess water to improve crop yield. Research shows an increasing trend of baseflow over the last 60 years and may be due to this artificial drainage activity. The influence of tile drainage on streamflow has been studied using in situ measurements and hydrological models. Evapotranspiration (ET) is an important component of hydrologic cycle. However, the impact of tile drainage on ET, or on other components of the water budget, is not well documented. In this study, we applied an energy balance based multi-sensor (GOES, MODIS and Landsat) data fusion method to estimate daily 30 m ET over an intensively drained area in South Dakota, USA to assess the model performance and further explore the spatial and temporal ET patterns through 2004-2013 and their relationship to drain installation.