FLOWS aims to determine how and when  Earth Observation data and derived products can best support flood crisis management at three time steps: flood crisis, aftermath and reconstruction. The research will be based on the experience of flood crisis managers (e.g. Belgian emergency services, local/regional/federal administrations, crisis centres) and the retrospective analysis of the acquisition, processing, and use of Earth Observation data from recent floods in Germany and Belgium.  The research hypothesis is that the gaps in spatial information needed by flood crisis managers could be met by the currently available Earth Observation data and technologies. 

We will innovate in the field of flood crisis management by integrating multi-sensor remote sensing, thus optimizing both acquisition speed and coverage, tuning Vision transformers (ViT) to capture flood extent, damage, and victim location as fast as possible with these data, mapping population dynamics in flood-affected areas based on social media and mobile phone network data, making the best use of systematically acquired satellite and aerial EO dataset to delineate and identify flood affected critical infrastructures, producing probabilistic hot spot maps of flood-affected areas based on heterogeneous data sources, and using participatory approach to focus research on social benefits for flood crisis managers. The needs in geoinformation of the flood managers at the three-time steps mentioned above will be identified through participatory discussions using problem trees. They will be used as compass to ensure relevance between research results and field needs. The expected outcome is that multiple geoinformation needs of flood crisis managers will be met by the products developed in the frame of this project. These results would pave the way to future operational developments helping to save lives during floodings, bring support to flood-affected populations and design more flood-resilient landscapes.