Comparison of Four Different Energy Balance Models for Estimating Evapotranspiration in the Midwest United States

Abstract:

Availability of no-cost satellite images helped in development and utilization of remotely sensed images for water use estimation. Remotely sensed images are increasingly used for estimating evapotranspiration (ET) at different temporal and spatial scales. However, selecting any particular model from a plethora of energy balance models for estimating ET is challenging as each different model has its strengths and limitations. We compared four commonly used ET models, namely, Mapping EvapoTranspiration at high Resolution with Internalized Calibration (METRIC) model, Surface Energy Balance Algorithm for Land (SEBAL) model, Surface Energy Balance System (SEBS) model, and Operational Simplified Surface Energy Balance (SSEBop) model using Landsat images for estimating ET in the Midwest United States. We validated our model results using three AmeriFlux cropland sites at Mead, Nebraska. Our results showed that the METRIC and the SSEBop model worked very well at these sites with a root mean square error (RMSE) of less than 1 mm/day and an R² of 0.96 (N=24). The mean bias error (MBE) was less than 10% for both the METRIC and the SSEBop models. In contrast, the SEBAL and the SEBS models have relatively higher RMSE (> 1.7 mm/day) and MBE (> 27%). However, all four models captured the spatial and temporal variation of ET reasonably well (R² > 0.80). We found that the model simplification of the SSEBop for operational capability was not at the expense of model accuracy. Since the SSEBop model is relatively less data intensive and independent of user/automatic selection of anchor (hot/dry and cold/wet) pixels, it is more user friendly and operationally efficient. The SSEBop model can be reliably used for estimating water use using Landsat and MODIS images at daily, weekly, monthly, or annual time scale even in data scarce regions for sustainable use of limited water resources.