High resolution evapotranspiration from UAV multispectral thermal imagery: Validation and comparison with EC, Landsat, and fused S2-MODIS HSEB ET

Abstract

Accurate evapotranspiration (ET) estimation is crucial for optimizing irrigation and managing water resources at the field scale. This study investigates the potential of unmanned aerial vehicles (UAVs) equipped with the MicaSense Altum sensor for high-resolution ET mapping using the Hybrid Single Source Energy Balance (HSEB) model. We focused on a 4.5 ha sprinkle-irrigated potato field at the American University of Beirut Agricultural Research and Education Center (AREC) in Lebanon’s Bekaa Valley. Eleven UAV flights were conducted throughout the growing season, synchronized with Landsat 8 and 9, and MODIS LST overpasses. HSEB ET from the Altum sensor was compared against EC data from a flux tower setup, and a comparative analysis was performed with HSEB ET from Landsat 8, Landsat 9, and Sentinel-2 (with MODIS LST). HSEB ET from the UAV exhibited very close agreement (3 % lower) with EC data, with a low RMSE of 0.60 mm/day. Notably, UAV-derived land surface temperature (LST) was on average 3 % higher than infrared radiometer LST. In contrast, comparisons of UAV LST with Landsat and S2MOD LST data revealed significant overestimations of LST (43 % and 24 %, respectively). Consequently, HSEB ET from Landsat and S2MOD were lower than EC ET by 17 % and 6 %, respectively. The strong agreement between UAV-HSEB and EC data underscores the potential of UAV thermal data for accurate irrigation management in heterogeneous fields using the HSEB model. While limitations exist regarding coverage area and cost, the detailed information obtained from UAVs can be highly valuable for optimizing irrigation practices and improving water use efficiency at sub-field scales.