MODELLING NEARSHORE WAVES AND OVERTOPPING USING A STOCHASTIC TYPHOON SET

Abstract

Typhoons and the storm surges and large waves generated by them have caused severe coastal damage in Japan such as during typhoon Vera in 1959. This event struck the Ise Bay region of Japan and caused wide spread devastation where coastal sea walls were destroyed and storm surges caused significant flooding destroying thousands of buildings and contaminating drinking water. This event caused approximately $1.9 billion worth of damage in today’s money (RMS report). Whilst Japan is well defended against surge and waves caused by typhoons some areas of defence could be liable to overtopping by very large waves acting on increased sea level due to surge and wave-setup. Breaching by powerful wave action may also occur exacerbating the situation and leaving low lying land susceptible to flooding. For example, during typhoon Vera earthen levees were either overtopped or breached and only defences constructed since 1953 along one side of Ise Bay after a previous typhoon could withstand the increased water levels and force of the waves (JWF, 2005). In this respect defences are constructed to design levels based on previous experience including a small tolerance level and may not cope with a storm that occurs in the tail end of the severity distribution. Were the 1959 Vera event to occur today, damage due to waves and storm surge could add up to $6 billion to total value of loss to residential properties including the wind loss despite improved flood defences (RMS report). It is important to try and quantify the possibility of flooding by all probable typhoons now and in the future. To estimate the potential loss due to certain events stochastic sets of typhoons can be statistically generated to create thousands of scenarios to drive surge and wave models and predict the magnitude of the surge and wave height at the coast (e.g. Yasuda et al., 2011). Modelling waves is very computationally expensive. Therefore, to create an event set based on thousands of typhoons a simplification of the modeling approach for waves at the coastal defences is required before simulating any subsequent overtopping. In this work a modeling system is developed to propagate the most severe typhoons (in terms of wave generation) in a stochastic set through a wave model to calculate any subsequent overtopping that is computationally viable and has a high degree of flexibility to explore uncertainties and investigate a range of scenarios.