🏛️ Participation in EGU26

11/05/2026

Last week I presented in Vienna the talk "Trade-offs between data-driven and process-based approaches for root-zone soil moisture retrieval in a Mediterranean vineyard" (EGU26-21629), where we compare the effectiveness of physical and machine learning models for soil moisture monitoring in Terra Alta (Spain). The results obtained with the team from the Ebro Observatory and isardSAT highlight the superiority of XGBoost models, which achieve high precision and prove to be very robust even in the absence of precipitation data, as the surface moisture signal already contains the key information about infiltrations. This research, linked to the DTE Hydrology Next (ESA) and Life eCOadapt50 (EU) projects, will allow us to improve operational monitoring and irrigation recommendations in non-instrumented plots using satellite data. You can check the presentation here.