Spectral Refraction Modeling of Waves Around the Steep Reef at Palau

Abstract

Western Pacific islands, such as Palau, are susceptible to wave-driven inundation due to low land elevations and limited space for coastal retreat. Assessing the impacts of extreme wave events on these low-lying islands requires regional numerical models that account for remotely generated waves from all directions and that resolve the complex bathymetry surrounding the islands. Such models often are computationally expensive and impractical for operational forecasts. In this study, eight years of wave observations from 10 sites around the island chain of Palau, at depths ranging from 9.4 to 17.6 m, are used to characterize the spatial variability of the island group's wave climate and to test the ability of a spectral refraction model to downscale regional wave hindcasts and reproduce the observed incident conditions. The spectral refraction model demonstrates significant skill predicting reef edge significant wave heights (mean $r^2$ = 0.77, mean $RMSE$ = 0.21 m) at the locations of our island scale wave gauge array and outperforms a regional 7 km resolution hindcast product based upon the WaveWatch III community model (Smith et al., 2021, https://doi.org/10.1002/gdj3.104). Furthermore, the spectral refraction model accurately downscales wave predictions during Typhoon Lan demonstrating its potential for operational forecasting and place-based damage and risk assessments during extreme events. When used with deep-water wind-wave models as input, spectral refraction methods can significantly improve reef edge incident wave boundary conditions necessary for shallow water reef modeling with low computational cost.