GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 2.1 GNGTS 2024 might be particularly susceptible to tsunamis but are not easily discernible in broader regional models. Hence, developing a local hazard model is an essential step in accurately forecasting the potential impact of tsunamis and developing effective mitigation and response strategies. But turning offshore into high-resolution onshore PTHAs, comprehensively capturing inundation hazard and uncertainty, while resolving spatial scales relevant to risk management on the order of 5-10 m, is a challenging task (Lorito et al., 2015; Lynett et al., 2017; Sepúlveda et al., 2019; Volpe et al., 2019; Gibbons et al., 2020; Tonini et al., 2021; Davies et al., 2022). In this study we introduce an enhanced method for conducting a local Probabilistic Tsunami Hazard Assessment (SPTHA) based on a regional SPTHA without the need of HPC (High Performance Computing) resources (Fig. 1,a). The method aims to reduce the computational effort required for a local tsunami assessment, updating and simplifying some previous approaches (Lorito et al., 2015, Volpe et al., 2019). The procedure is tested in the region of Catania, Sicily, south of Italy, and applied to the Ravenna harbour, situated in the Northern Adriatic Sea, Italy (Fig. 1,b). The method The developed method allows to refine the regional SPTHA by identifying the most significant tsunami sources that impact the local hazard. The resulting procedure simplifies some previous workflows (Lorito et al., 2015, Volpe et al., 2019) for quantifying local SPTHA and represents a useful tool that can be potentially applied wherever there is regional hazard information available, ultimately leading to improved accuracy in the assessment process. The first and innovative step of our approach involves the application of an “importance” sampling technique that adopts regional hazard disaggregation as weighting information. A source refinement of the scenarios closest to the target is then applied to the new subset, enhancing the characterization of local sources, thereby improving hazard modelling by capturing natural variability (aleatory uncertainty) and reducing epistemic uncertainty. Offshore tsunami simulations are conducted on the retrieved scenarios and the water height profile over a series of points close to the target area; together with the coseismic field information, it represents the feature for a subsequent filtering operation to further reduce the number of high-resolution tsunami simulations required for the local hazard definition. The workflow of the approach is reported in Fig. 1. and consists of four main steps: source preselection; source refinement; cluster analysis; local hazard quantification.