Differences in assigning probabilities to coastal inundation hazard estimators: event versus response approaches

Abstract

Coastal flood risk assessment requires a reliable estimation of the frequency of inundation hazards, that is, characterising the hazard magnitude and assigning a probability of occurrence. In this work we analyse the uncertainty introduced in the assessment associated to the method to assign the probability of occurrence to coastal flood hazards. To this end we have compared the use of two general methods, the response and the event approaches. Different procedures are used to characterise coastal inundation hazards depending on the analysis scale and data availability. Thus, a range of possibilities has been analysed, from simple estimators such as run-up to modelled flood-prone areas. The analysis has been performed for all wave and water level conditions around the Spanish coast. The results show that the differences between the methods are location-dependent, and thus, determined by the exposure to wave and water level conditions. When using the event approach, the run-up or total water level (with good correlation between waves and surge) distributions reasonably approximate those of the response approach with low associated uncertainty. When the assessment aims to output overtopping discharges or inundation maps, observed differences suggest that the event approach would produce misleading conclusions in inundation-related coastal management and decision-making.