PSI - Issue 84

Luciano Pavesi et al. / Procedia Structural Integrity 84 (2026) 583–590

585

on bridges even when data availability is limited. The procedure distinguishes between river reaches with and without levees and applies different criteria to estimate the water level and hydraulic head. In embanked channels, the reference water level is defined as the lower of the two levee crest elevations increased by a fixed amount. In non-embanked reaches, it is derived from officially delineated flood hazard zones, corresponding to medium- and high-probability flooding scenarios (P2/P3 according to the PGRA - Piano di Gestione del Rischio Alluvioni) defined within national flood risk management plans, or from simplified surrogate criteria where such delineations are unavailable. While standardization ensures consistency at the national scale, it inevitably introduces uncertainty and potential bias in risk classification. A meaningful example is in reaches protected by levees, where simplified estimates of water depth can lead to an under/overestimation of overtopping conditions when fixed increments applied to levee elevations do not adequately represent actual flood dynamics, resulting in hazard classes lower than those that would emerge from physically based hydrologic-hydraulic analyses. Similarly, the assessment of erosion-related hazards, including both generalized and local scour, relies heavily on empirical formulations and simplified assumptions that do not capture more complex hydrodynamic processes at screening stages, further contributing to uncertainty in hydraulic risk classification. Officially delineated flood hazard maps constitute the reference framework for screening-level hydraulic assessments. At the national scale, such maps often represent a mosaic of studies developed independently across river basin districts, using different hydrological inputs, modelling approaches and spatial resolutions (de Moel et al., 2009). Such fragmentation complicates comparative risk assessments across regions. Moreover, in official flood hazard maps the secondary smaller tributaries are often poorly represented or entirely neglected. ANSFISA Guidelines explicitly acknowledge these limitations and recommend integrating screening-level assessments with more detailed hydraulic modelling where conditions are particularly critical. Their main strength lies in providing a standardized, simple and nationally homogeneous framework that enables consistent hydraulic risk screening across all the bridges in Italy. To fully exploit this potential within a multi-level assessment approach, screening procedures would benefit from being complemented by methodologies that explicitly account for the physics of flood processes while remaining standard, homogenized and simplified. In this context, recent advances in large scale flood modelling provide new opportunities to complement screening-level assessments with physically based yet scalable approaches. Among these approaches, the large-scale inundation model RESCUE (laRgE SCale inUndation model, Pavesi et al., 2022) represents a notable example, combining a geomorphology-based representation of the river network with simplified hydrological and hydraulic formulations to estimate water levels and inundation extents at large scale. By preserving computational efficiency while explicitly representing hydrologic-hydraulic processes along the full drainage network, RESCUE provides a suitable basis for evaluating inundation extent and characteristics where infrastructures such as bridges and levees interact with flood propagation. The use of a large-scale modeling framework makes it possible to go beyond purely geometric assumptions or fixed elevation offsets, while remaining consistent with the rapid and homogeneous nature of screening-level analyses. However, the potential of this approach is limited by the availability of detailed and spatially consistent information on these linear infrastructures, which are currently not explicitly represented in the framework. It is within this gap that the RESCUE_SAT project (Agreement n. 2025-2-HB.0), funded by the Italian Space Agency (ASI) under the “Innovation for Downstream Preparation for Science” (I4DP_SCIENCE) program, is positioned. RESCUE_SAT proposes an infrastructure-aware extension of the large-scale inundation model, based on the integration of high-resolution satellite Earth Observation data into RESCUE. The scope is bridging the gap between rapid screening-level assessments and more detailed hydraulic analyses to support present-day and future flood risk assessment, with particular emphasis on the representation of relevant infrastructures. 2. The RESCUE_SAT framework RESCUE_SAT builds upon the RESCUE model by introducing an explicit representation of infrastructures that locally control flood dynamics, as shown in Fig. 1. To this end, the project exploits multi-sensor satellite data to