Estimation of high spatiotemporal resolution actual evapotranspiration by combining the SWH model with the METRIC model

Abstract

Accurately quantifying the daily evapotranspiration of the region is one of the essential steps for studying the water cycle and regulating water resources. Nowadays, the combination of hydrologic model and remote sensing technology is considered to be an advanced means to estimate evapotranspiration and hydrological processes on a regional scale. However, continuous daily evapotranspiration cannot be simulated based on remote sensing image products. In this study, we developed a new model, SWH − METRIC model, by combining the SWH model with the Mapping Evapotranspiration at High Resolution with Internalized Calibration (METRIC) model, to simulate continuous daily actual evapotranspiration (ET) with high spatiotemporal resolution. The new model solved the problem of ET discontinuity calculated by remote sensing images and hydrological models. Regarding the results, the simulation of the SWH − METRIC model had a high consistency with the eddy covariance measurements, with an average R2 of 0.70 and a root mean square error of 0.67 mm day−1, and was superior to the SWH model. Finally, The SWH-METRIC model was used to estimate continuous ET in the Loess Plateau in northern Shaanxi (LPNS), China. Meanwhile, considering the interaction between the climate and the rugged terrain of the LPNS, the ANUSPLIN procedure was used to spatially interpolate the input meteorological factors of the model, and the temporal resolution of the simulated ET was 1 day and the spatial resolution was 1 km. The SWH-METRIC model can improve the simulation accuracy and spatiotemporal resolution of ET and contribute to study the response mechanism of climate change to ecosystem. It can provide a useful reference for the fine quantification and management of agricultural water consumption.