High spatial resolution thermal infrared data from airborne acquisition reveal inconsistencies in evapotranspiration and crop water stress retrieval from different models

Auteurs

Corbari C., Hu T., Paciolla N., Schlerf M., Mallick K., Ronellenfitsch F.K., Ceppi A., Crisafulli V., Bossung C., Feki M., Llorens R., Skokovic D., Al Bitar A., Sobrino J., Mancini M.

Référence

Remote Sensing Applications: Society and Environment, vol. 38, art. no. 101563, 2025

Description

As agriculture is the largest consumer of water worldwide, water use efficiency and the impacts of water stress on crops are of critical concerns. This study investigates the diurnal and spatial variability of evapotranspiration and crop water stress by integrating very high spatial resolution (1–4 m) airborne thermal infrared (TIR) data with visible and near infrared data from Planet satellites (3.7 m). These datasets were incorporated into three surface energy balance models with contrasting structures: the numerical FESTresidual, the analytical STIC, and the semi-empirical S-SEBI models. The analysis focused on an agricultural area in central Italy, comprising several orchards and vegetable fields. An intensive airborne survey campaign using a hyperspectral TIR camera was conducted in July 2022, with three overpasses per day. Planet data were used to derive vegetation biophysical parameters. The three models were evaluated against measured energy fluxes from an eddy-covariance station, showing inconsistent results throughout the day, with the greatest disparities occurring at midday, and across various land cover types. Additionally, the impacts of irrigation system, timing, and volumes on airborne land surface temperature (LST) and modeled latent heat (LE) were analyzed, showing significant differences. Finally, the crop water stress index effectively distinguished between fully irrigated fields and non-irrigated areas, providing valuable insights for water stress management in agriculture.

Lien

doi:10.1016/j.rsase.2025.101563