AI4Drought: AI for SCIENCE - Multi-Hazards, Compounds and


There is a clear and urgent need to deepen our understanding of the occurrence and the cascading effects of droughts. In this proposal, we consider seasonal climate predictions in the Iberian peninsula (Spain and Portugal) that are projected to move towards a drier climate (Vicente-Serrano et al. 2014, Donelly et al. 2017). This case study represents an ideal opportunity for testing our ability to improve drought-related phenomena predictions and to measure their impacts. AI4Drought presents a comprehensive approach combining AI techniques, dynamical SPS and multiple EO-products that are appropriate for studying societal impacts (e.g., soil moisture, NDVI, lake levels or burned area) to improve our prediction capabilities and enrich our understanding of the causes, evolution and consequences of droughts at seasonal time-scales.