Uncertainty of wave runup prediction on coral reef-fringed coasts using SWASH model

Predicting wave runup on reef-fringed coasts using phase-resolving modelings suffers from uncertainty induced by wave randomness, spectral-width determination, and reef-surface-roughness reproduction. A multi-layer non-hydrostatic model SWASH was first validated against two physical experiments and then applied to simulate runup over coral reefs. Simulations with different wave sequences for the same wave condition indicated that infragravity swash (R-IGW) variability was the main contributor to the extreme runup (R-2%) uncertainty. R-IGW was positively correlated to the off-reef incident-wave groupiness factor and the average number of waves per wave group, suggesting that prior knowledge of incident group structure could help in reducing the uncertainty of runup prediction. In ensembles with 100 simulations of different random wave sequences, the ratio of R-2% with 5% exceedance level to R-2% with 95% exceedance level decreased with increasing number of spectral components and approached to 1.15-1.3 with over 600 components. Runup uncertainty induced by randomly phasing did not significantly vary with spectral widths and reef-surface roughness. Runup uncertainty due to arbitrarily specifying reef-surface roughness was much more significant than the uncertainty induced by wave randomness and spectral width, suggesting that accurate reproduction of reef surface roughness has a high priority in physical and numerical modelings.

Automated summary

The study shows that predicting wave runup on reef-fringed coasts is mainly affected by infragravity swash variability. Prior knowledge of the incident wave group structure can help reduce this uncertainty. Increasing the number of spectral components can decrease the uncertainty in wave runup height.