The effect of thermal image quality on the estimation of Crop Evapotranspiration

Understanding the surface energy balance is essential for studying land-atmosphere interactions and their impact on weather, climate, and hydrology. Accurate estimation of sensible and latent heat fluxes is critical for applications like hydrological modelling and climate studies, but traditional methods like eddy covariance are limited in spatial coverage. Remote sensing technologies, particularly models like the Two-Source Energy Balance (TSEB), address these limitations by partitioning energy fluxes between soil and vegetation using spatially distributed observations such as surface temperature and vegetation indices. Advances in TSEB include refined resistance networks for modelling soil-canopy interactions and improved disaggregation of surface temperatures into soil and canopy components, with iterative algorithms enhancing flux partitioning. Challenges remain in accounting for vegetation clumping and accurate modelling in water-limited ecosystems. In this study, the potential of three thermal data providers, Ecostress and Landsat from NASA and Sentinel-3 from ESA, in estimating evapotranspiration using TSEB was assessed. Other data, like meteorological, is the same for both simulations. We want to see how the quality of the thermal data, resolution and accuracy, affects the result of TSEB. This study is necessary to determine the minimum requirements of a thermal imagery dataset, suitable for this use-case. The final aim is to improve water productivity and improve yield by early detection of water stress in crops, before it becomes visible.