Intercomparison of very high-resolution surface soil moisture products over Catalonia (Spain)

• DOI: 10.1016/j.rse.2024.114225
• Publication Date: 2024-05-28T09:50:04Z
• Authors: Nadia Ouaadi; Lionel Jarlan; Michel Le Page; Mehrez Zribi; Giovani Paolini; Bouchra Ait Hssaine; Maria Jose Escori...; Pascal Fanise; Olivier Merlin; Nicolas Baghdadi; Aaron Boone
• Keywords: 550; [SDE.IE]Environmental Sciences/Environmental Engineering; Remote sensing; Surface soil moisture; L-band; [SDE]Environmental Sciences; [SDU.STU.HY] Sciences of the Universe [physics]/Earth Sciences/Hydrology; Very high-resolution; [SDE.IE] Environmental Sciences/Environmental Engineering; [SDU.STU.HY]Sciences of the Universe [physics]/Earth Sciences/Hydrology; C-band; Evaluation; Semiarid region; Semiarid

The surface soil moisture (SSM) is a key variable for monitoring hydrological, meteorological and agricultural processes. It can be estimated from active passive microwave remote sensing data. While coarse-resolution SSM products (> 1 km) have already been evaluated large range of ecosystems, such assessments lack very high-spatial-resolution products, although they are increasingly available thanks to high-resolution radar data or disaggregation methods applied coarse-scale products. Within this context, the aim current study carry out, first time, an intercomparison high-spatial resolution using in situ database collected 33 fields located Ebro basin (Spain) that were cultivated with different crops irrigated techniques. Three considered: (i) SSMTheia at field scale derived Sentinel-1 Sentinel-2 machine learning algorithm; ii) SSMρ 50-m both backscattering coefficient interferometric coherence based on inversion simple radiative transfer model; iii) SSMSMAP20m 20-m obtained by disaggregating SMAP Sentinel-3 statistical metrics computed whole show two outperform disaggregated approach product exhibits better than product. This mainly attributed inability retrieve >0.3 m3/m3. correlation coefficients >0.4 (up 0.8) 72%, 40% 27% SSMρ, SSMSMAP20m, respectively. Similarly, 80% had RMSE values between 0.06 m3/m3 0.1 against 36% SSMSMAP20m. In addition, time series analysis showed was able detect large-scale wetting events as rainfall impacted pixel while irrigation not detected, because used land temperature related hydric status surface. results perform reasonably well cereals and, lesser extent, annuals, drastic drop observed tree crops. Finally, spatial pattern over area also depicted comparison airborne GLORI GNSS-R (Global Navigation Satellite System Reflectometry) maps. highlights limitations provides insights improving scheduling scale.