How well do high-resolution surface soil moisture products capture irrigation signals?

Irrigation profoundly alters near-surface hydrological processes, yet its representation and detectability in satellite-based surface soil moisture (SSM) products remain insufficiently understood. While SSM observations are increasingly used in irrigation-related studies and applications over managed agricultural landscapes, most existing evaluations focus on natural or rainfed conditions, leaving a critical gap in anthropogenically influenced environments. 

We assess the performance of six high-resolution (1 km) SSM products, including BEC SMOS L4, UFZ-Sentinel-1, RT1, CGLS, NSIDC SMAP, and a newly developed downscaled ESA CCI product. The analysis focuses on three major European irrigation hotspots: the Ebro Valley (Spain), the Po Valley (Italy), and the Thessaloniki region (Greece). 

The evaluation is structured around three complementary analyses. First, spatial and temporal consistency is examined by comparing SSM distributions over irrigated and rainfed areas using global irrigation maps, and by assessing temporal dynamics against district-scale irrigation records. Second, satellite SSM products are benchmarked against model-based ERA5-Land estimates that do not explicitly represent irrigation, in order to analyze anomalies and identify potential human-induced soil moisture signals. Third, physical consistency is assessed by examining the relationship between SSM and land surface temperature (LST), as irrigation is expected to induce surface cooling through increased evapotranspiration. 

The analysis highlights marked differences between products in their ability to detect irrigation-related SSM signals and provides a basis for their evaluation and use in irrigated, human-modified environments.