Key points are not available for this paper at this time.
In May 2023, the region of Emilia-Romagna, Italy, experienced an unprecedented hydrological event when 350 million cubic meters of rain fell over 36 hours, leading to widespread flooding and landslides. This disaster, affecting 100 municipalities, was compounded by antecedent drought conditions that had decreased the soil's water absorption capacity. Earth observation (EO) data became critical, providing emergency services with the means to assess and manage the catastrophe and facilitate post-event damage evaluation. The SaferPlaces platform, supported by the ESA InCubed programme, played a pivotal role in disaster response. It provided the Civil Protection of Emilia-Romagna with high-resolution flood water and depth maps, crucial for decision-making in the aftermath of the floods. This cloud-based platform integrates satellite data, climatic records, and AI algorithms to generate global flood forecasts. Leveraging AI, SaferPlaces processed terrain data alongside inundated area information, combining in situ measurements with satellite data from Copernicus Sentinel-2, CosmoSky-Med, Planet, and SPOT. This was further enriched with local data from municipalities and the Emilia-Romagna Civil Protection, enhancing urban flood area accuracy. Detailed maps illustrating flood extent in the severely hit municipalities of Faenza, Cesena, Forl, and Conselice were generated. These contained vital data on water depth and volume, forming the basis for a preliminary Flood Damage Assessment. These assessments were crucial for authorities to estimate economic losses swiftly. The suite of tailored algorithms within SaferPlaces, extracts flood water masks from satellite imagery. This module, accessible on-demand through a user-friendly interface, requires few parameters from users to accurately delineate flooded areas and contribute to the Global Flood Monitoring system. The main workflow of algorithm includes the GFM procedure for baseline flood extent retrieval, the Hydraflood method for flood mask extraction via GEE, and the CommSNAP pipeline for processing commercial data. The final output is a flood mask for the area and event of interest, which can also feed into the GFI model to identify flood-prone areas. This study underscores the essential role of integrated EO and AI technologies in managing hydrological disasters. The SaferPlaces platform's capacity to synthesize multi-source data and provide actionable intelligence marks a milestone in the power of interdisciplinary approaches in enhancing disaster resilience and preparedness.
Mazzoli et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: