Risk mitigation policies (like dike rising) are essential to address increasing coastal flood risks due to global warming. Furthermore, the optimal level of risk mitigation policy should be determined by public preferences for risk reduction. However, it is difficult to reveal public preferences for coastal flood risk reduction because projections of coastal flood risks inevitably involve uncertainty. This study aims to estimate household preference for coastal flood reduction under ambiguity and multiple projections of coastal flood risks. By coupling storm surge inundation simulations and stated preference experiments with decision models, we estimate the expected loss reduction, risk premium, and ambiguity premium for coastal flood risk mitigation policies. Results of the study show that ignoring the ambiguity premium causes significant undervaluation of coastal flood risk mitigation, and the ambiguity premium stems from households' over-concern about the worst projection, which may lead to an over-allocation of resources to prevent inundation damage caused from the worst-case flood before a disaster. The study concludes that a risk mitigation policy combining public insurance for the worst projection and pre-disaster prevention measures can be effective and efficient.

To assess inundation risks and ambiguity in coastal areas, we proposed a framework: (1) conduct a simulation of typhoon generations for 200 years using a global stochastic tropical cyclone model (GSTCM); (2) the total number of typhoons in this study generated over half a million in the Western Pacific Ocean, to understand the uncertainty of typhoon storm surge inundation, also to avoid unnecessary inundation simulations, the significant four typhoon ensembles (each ensemble including 25 typhoon cases) are selected (after fulfilling the conditions), and the storm surges of Osaka Bay are simulated by a full-coupled surge-wave-tide coupled model (SuWAT); (3) predict the inundation depth due to storm surges using the inundation simulation model; (4) repeat step (1) to (3), get 25 projections of the inundation risk for each dike level: current level and rising by 0.5m, 1.0m, 1.5m, and 2.0m (25×5); (5) the average and worst projections of the inundation risks are specified by each zip-code in the web-based survey; (6) estimate households' preferences by asking them to choose whether to buy hypothetical insurance to cover all losses from coastal flooding by presenting the average and worst scenarios of the inundation risks to their houses; (7) by using the choice experiment data, a decision model is applied for estimating risk premiums and ambiguity premiums; (8) analyze the geographical distribution of risk premiums and ambiguity premiums by geographic information system (GIS).

Typhoon data sets for this research are available through Mori et al. (2019).

Inundation simulation model codes are available at: https://github.com/HEMLab/hipims. Please refer to Liang (2010) for further information on this model.