Evaluation of satellite remote sensing-based crop evapotranspiration models over a semi-arid irrigated agricultural farm

Abstract

Abstract The measurement of evapotranspiration (ET) is essential in maintaining the energy and water balance in agricultural ecosystems, and it plays a vital role in the hydrological cycle. Precision irrigation water management requires accurate spatiotemporal coverage of crop ET across the farm. Fortunately, with the availability of multi-temporal high-resolution satellite datasets and remote sensing-based surface energy balance (SEB) models, near-real-time estimation of ET is now possible. A recent study evaluated and compared the performance of several SEB models, including the Surface Energy Balance Algorithm for Land (SEBAL), Surface Energy Balance Index (SEBI), Surface Energy Balance System (SEBS), Simplified Surface Energy Balance (SSEB), Simplified-Surface Energy Balance Index (SSEBI), and Two Source Energy Balance (TSEB) models over semi-arid irrigated farms in India. The study used 24 Landsat images captured during the post-monsoon seasons of 2021-22 and 2022-23. The statistical evaluation revealed that SEBAL had the best overall performance (r = 0.91, MBE= -0.48 mm d− 1, MAE = 0.42 mm d− 1 and RMSE = 0.51 mm d− 1), followed by SSEB, TSEB, SSEBI, SEBI, and SEBS, respectively. While SEBAL, SSEB, S-SEBI, and TSEB models performed similarly, SEBI and SEBS consistently underestimated ET over the season. The spatiotemporal map was also used to evaluate the model's performance, and it could accurately differentiate between ET over less water-intensive pulses and water-intensive wheat fields on the farm. Despite discrepancies among the SEB models, SEBAL can still be an operational tool for mapping ET with high accuracy and sufficient variation across pixels, making it an ideal option for incorporating into irrigation scheduling over semi-arid farms.