Developing a soil water index-based Priestley–Taylor algorithm for estimating evapotranspiration over East Asia and Australia

Abstract

Evapotranspiration (ET) is an important component of hydrological processes. Reliable estimates of ET variation are of vital importance for natural hazard adaptation and water resource management. In order to improve the accuracy of ET estimation, this study first developed a soil water index (SWI)-based Priestley–Taylor algorithm (SWI–PT) based on the enhanced vegetation index (EVI), SWI, net radiation, and temperature. The algorithm was then compared with a modified satellite–based Priestley-Taylor ET model (MS–PT). After examining the performance of the two models at 10 flux tower sites in different land cover types over East Asia and Australia, the validation results showed that the daily estimates from the SWI–PT model were closer to observations than those of the MS–PT model in each land cover type. The average correlation coefficient of the SWI–PT model was 0.81, compared with 0.66 in the original MS–PT model. The average value of the root mean square error (RMSE) decreased from 36.46 W/m2 to 23.37 W/m2 in the SWI–PT model, which used different variables of soil moisture and vegetation indices to capture soil evaporation and vegetative transpiration, respectively. Overall, the results showed that the newly developed SWI–PT model captured ET more accurately than the MS–PT model. In comparison with the NDVI in the MS–PT model with high values, the low seasonal EVI values indicated that vegetative transpiration accounts for a smaller part of ET in the SWI–PT model. The ranges of fraction of soil moisture (fsm) in MS–PT remained unchanged in each season, while a clear seasonal change of the SWI was found in SWI–PT which indicated that the temperature change-based fsm could not capture soil water conditions well compared with the SWI. The estimated ET from the MS–PT model was most sensitive (to the normalized difference vegetation index (NDVI) in forests) to net radiation (Rn) in grassland and cropland. The estimated ET from the SWI–PT model was most sensitive to Rn, followed by SWI, air temperature (Ta), and the EVI in each land cover type.