Recent coastal storm impacts emphasized the need of proper coastal risk assessment to propose adequate risk reduction plans. Coastal managers should be able to predict and compare the effect of Disaster Risk Reduction (DRR) measures under current and future scenarios. The analyses should integrate multi-hazard assessments at the receptor scale and should be as flexible as to be applied at different morphologic and socio-economic settings. The EU Risc-Kit project (www.risckit.eu) provided tools in support to coastal management to be applied at different scales throughout the disaster management cycle (Van Dongeren et al., 2017). In this work, the local scale Hotspot tool (Jäger et al., 2017) was applied at two Mediterranean case studies to test and compare DRR measures under current and future scenarios, considering flooding and erosion hazards. The tool implements the Source-Pathway-Receptor-Consequences (SPRC) concept into a Bayesian Network (BN)(Figure 1). The hazard component is calculated through a process-based model chain. The last step of the chain consists of a 2DH XBeach model providing hydro-morphodynamic results. Hazards are translated into consequences for the exposed receptors through vulnerability relations. Then, results are integrated into the BN linking storm characteristics with expected impacts through conditional probabilities. The approach was repeated for a large number of forcing conditions, in current and future conditions, with and without the implementation of DRR measures affecting hazard, vulnerability or exposure. The BN was used to explore and compare results in an integrated manner. The tool was applied at two urbanized touristic sandy beaches, in Spain (Tordera Delta, Maresme-La Selva) and Italy (Lido degli Estensi-Spina, Comacchio). DRR measures, such as artificial dunes, nourishments, managed retreat and nonstructural measures were tested in both current and future conditions. At both case studies: (i) the results of the current conditions appeared to be consistent with the knowledge of the area; (ii) the method was able to provide quantitative information on the variations of the level of risk for receptors under the projected future conditions; (iii) the DRR measures evidenced mainly positive effects in terms of risk reduction, with some exceptions, in both current and future conditions. Strategic alternatives (i.e., single or set of measures) were compared to select optimal combinations of DRR. The main limitations of this study were related to the numerical approximations of the model chain, the adopted vulnerability relations, the design of the measures and the future conditions. Finally, the higher is the number of considered storm conditions, the more complete is the integrated assessment. However, this may lead to a massive effort in terms of computational time. Despite these limitations the scenario comparisons were effective. This work highlighted the advantages of using the Risc-Kit Hotspot tool for DRR testing in current and future conditions. Future applications will be implemented in Scandinavia in the framework of the EU ANYWHERE project (www.anywhere-h2020.eu).
Bayesian Network Approach for Climate Change and DRR Scenarios' Testing - Pilot Cases from Italy and Spain
Enrico Duo
;Paolo Ciavola;
2017
Abstract
Recent coastal storm impacts emphasized the need of proper coastal risk assessment to propose adequate risk reduction plans. Coastal managers should be able to predict and compare the effect of Disaster Risk Reduction (DRR) measures under current and future scenarios. The analyses should integrate multi-hazard assessments at the receptor scale and should be as flexible as to be applied at different morphologic and socio-economic settings. The EU Risc-Kit project (www.risckit.eu) provided tools in support to coastal management to be applied at different scales throughout the disaster management cycle (Van Dongeren et al., 2017). In this work, the local scale Hotspot tool (Jäger et al., 2017) was applied at two Mediterranean case studies to test and compare DRR measures under current and future scenarios, considering flooding and erosion hazards. The tool implements the Source-Pathway-Receptor-Consequences (SPRC) concept into a Bayesian Network (BN)(Figure 1). The hazard component is calculated through a process-based model chain. The last step of the chain consists of a 2DH XBeach model providing hydro-morphodynamic results. Hazards are translated into consequences for the exposed receptors through vulnerability relations. Then, results are integrated into the BN linking storm characteristics with expected impacts through conditional probabilities. The approach was repeated for a large number of forcing conditions, in current and future conditions, with and without the implementation of DRR measures affecting hazard, vulnerability or exposure. The BN was used to explore and compare results in an integrated manner. The tool was applied at two urbanized touristic sandy beaches, in Spain (Tordera Delta, Maresme-La Selva) and Italy (Lido degli Estensi-Spina, Comacchio). DRR measures, such as artificial dunes, nourishments, managed retreat and nonstructural measures were tested in both current and future conditions. At both case studies: (i) the results of the current conditions appeared to be consistent with the knowledge of the area; (ii) the method was able to provide quantitative information on the variations of the level of risk for receptors under the projected future conditions; (iii) the DRR measures evidenced mainly positive effects in terms of risk reduction, with some exceptions, in both current and future conditions. Strategic alternatives (i.e., single or set of measures) were compared to select optimal combinations of DRR. The main limitations of this study were related to the numerical approximations of the model chain, the adopted vulnerability relations, the design of the measures and the future conditions. Finally, the higher is the number of considered storm conditions, the more complete is the integrated assessment. However, this may lead to a massive effort in terms of computational time. Despite these limitations the scenario comparisons were effective. This work highlighted the advantages of using the Risc-Kit Hotspot tool for DRR testing in current and future conditions. Future applications will be implemented in Scandinavia in the framework of the EU ANYWHERE project (www.anywhere-h2020.eu).I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
