Anagrafe della ricerca

The management of flood risk scenarios is a major open issue with strong societal and economic impacts, especially in the context of climate change adaptation. Roads are frequently affected by floods, while they are essential for emergency operations. More specifically: (i) floods (flash-floods in particular) are hard to accurately and timely predict; (ii) most loss of lives because of floods occur on roads, particularly near river crossings; (iii) the era of big data lacks an effective integrated flood-road emergency management system for scenarios that, by definition, are rare. To overcome these difficulties, the general objective of the research is the development of a new multidisciplinary framework that will combine scientific advancements in the capability of (i) predicting flood and flash-flood events, (ii) identifying critical river-road intersections, (iii) assessing the impact of road closures on the network traffic with the goal of assessing drivers’ risks in adverse meteo conditions. The application objectives are the design and preliminary validation of an innovative risk analysis tool for road managers and users, based on predetermined ranking-based hydrological assessment of river crossings in the road network and communication track of meteorological scenarios fed by the ready (real-time) identification of the rarity of the rainfall patterns. The starting point is the use of Quantitative Precipitation Forecast from high-resolution operational meteorological models for the estimation of the exceedance probability of predetermined rainfall thresholds, defined through off-line scenario simulations of possible dangerous conditions in road-river crossings with new integrated hydrologic-hydraulic modeling. A meso-scale (province area) transport model will be implemented to understand the disruption effects of the flood on the road network, and identify safe itineraries. This will result in the capability of issuing informative flood alerts targeted at specific portions of the road network. A first hindcast approach will assess how the system would behave in the case of replication of past extreme rainfall events; the preliminary demonstration in true real-time will follow. The development of a dedicated geospatial data collector/sharing platform will ensure integration of scenarios and modelling components to efficiently feed the real-time application. The application of the new methodologies on two different study areas (one for the setup and one for replication) will demonstrate the capability of delivering also tangible operational products,