Evaluation of geostationary satellite (COMS) based Priestley–Taylor evapotranspiration

Effective water use in the irrigated agriculture requires an accurate estimation of evapotranspiration (ET) to understand the interaction between land surface and the atmosphere. The operationally available polar orbit satellite datasets with low temporal resolutions have long been utilized for the estimation of ET from field to regional scale. However, geostationary satellites, which are continuously measuring several factors related to land surface and the atmosphere over large regional scales, have high temporal resolution compared to polar orbit satellites. Thus, in this study, we present a framework for estimating potential ET at three different temporal scales (instantaneous, 3-h and daily) using the Priestley–Taylor (ETPT) method with a new geostationary satellite, the Communication, Ocean, and Meteorological Satellite (COMS) dataset. The derived ETPT estimates were compared with ground based flux tower [Cheongmi (CFC) and Sulma (SMC)] measurements, with ETPT calculated from MODerate Resolution Imaging Spectroradiometer (MODIS) and with ETPT calculated from Global Land Data Assimilation System (GLDAS) datasets during the growing seasons of 2011. The GLDAS ETPT were significantly overestimated compared to the flux tower ETPT (bias of 147.92 and 169.69Wm−2 at 3-h time scale and bias of 2.64 and 3.37mmday−1 at daily time scale for SMC and CFC, respectively), while the COMS and MODIS ETPT were slightly underestimated (bias ranged from −15.31 to −55.65Wm−2 at instantaneous time scale and bias ranged from −0.04 to −1.03mmday−1 at daily time scale for SMC and CFC, respectively). Based on the results, the COMS estimated ETPT was slightly more accurate than MODIS ETPT in comparison with the flux tower ETPT, yielding the index of agreement (IOA) between ∼0.89 and 0.96 for all time scale.