Towards a global forecast for wave-induced flooding on coral reef-lined coastlines

Many tropical coral reef-lined shores are low-lying, with elevations less than five meters above mean sea level. Due to their low elevation, these coasts are particularly vulnerable to marine flooding as a result of climate-change-driven sea level rise, coral reef decay and changes in (storm) wave climate. Enabling communities to continue living in these areas into even the near future will require improved resilience to marine flooding. As part of a range of resilience-increasing measures, accurate early warning systems (EWS) are vital for providing timely information of potential flood events to local communities. Unfortunately, current global flood predictions have limited practical value on reef-lined coasts due to their neglect of the complex hydrodynamics involved in wave-induced flooding, and their reliance on poor-quality, or non-existent topographic and bathymetric data.

In this study we attempt to overcome these limitations by designing an EWS methodology based on a Bayesian-based predictor of total water level (TWL; including tides, surge and wave run-up), that is informed by satellite-derived estimates of the local reef geometry and world-wide predictions of offshore water levels and wave conditions. Since poor quality island topographic data (where elevation errors in the order of tens of centimeters or greater are common) cannot be used to reliably predict flooding pathways, we use a statistical approach to compute the potential for flooding. In this approach, we use hindcast values of deep-water conditions to develop long time series of hindcast TWL at the coast. We then apply a comparative approach, in which TWL predictions are equated to historical events and return periods, to provide more meaningful information to local communities.

In this work we present the validation of the EWS methodology using wave transformation and wave run-up data collected over a one month period on an atoll island in the Republic of the Marshall Islands. We also identify the components in the EWS framework (i.e., reef geometry detection, offshore wave and water level predictions, wave transformation and run-up computation) that are the main sources of prediction uncertainty; these components require further research to improve flood predictions in operational EWS.