Methods and models to assess the effectiveness of hydraulic risk mitigation measures

Floods are among the most dangerous natural disasters, affecting thousands of people worldwide every year. Their frequency and intensity are expected to increase due to climate change and ongoing urbanization, making the identification of effective mitigation strategies increasingly important. However, flooding includes different types of events, such as pluvial, fluvial, and coastal floods, which differ in their governing physical processes and in the current level of scientific understanding. Mitigation strategies must therefore be developed with reference to the specific type of flooding considered. This thesis focuses on fluvial and pluvial flooding from different perspectives. For fluvial flooding, the study investigates simplified methodologies for the preliminary assessment of mitigation measures, avoiding the need for computationally expensive hydrodynamic models. The proposed approach was tested on two case studies in Sardinia and yielded promising results. For pluvial flooding, the research initially focused on two urban basins in New York City, represented using high-resolution geospatial data and a 2D shallowwater- equations (SWE) model. Three real flood events were simulated to compare different approaches for representing the influence of the sewer system when detailed drainage data are unavailable. Model performance was evaluated using street-level water-depth observations collected through the FloodNet sensor network. The results showed that the proposed approach, based on representing sewer outflow at inlets through weir/orifice equation, outperformed the other simplified methods considered. The validated modelling framework was then extended to assess the effectiveness of Nature-Based Solutions for pluvial flood mitigation. Green roofs and permeable pavements were included in the model and tested, both individually and in combination, under design storms with different return periods and coverage scenarios. The results indicate that, under the adopted simplified schematization, these measures can effectively reduce both flood depth and flood extent, particularly for events with low to medium return periods (10–25 years). Finally, the thesis proposes a new building representation for 2D SWE models applied to urban flooding. The approach is designed to include, in a simplified way, the influence of water storage within buildings and basements. Water is allowed to enter buildings through openings, while basements are represented through a storage-based formulation. The method was first evaluated on a simplified test case and then applied to an urban basin in London under a design storm. Despite its simplified formulation, the proposed approach produced realistic results, with magnitudes consistent with those reported for real flood events.