Abstract:
The exposure to wave overtopping is growing worldwide which forces coastal communities to
adopt methodologies for anticipating the risks associated with it. In areas with shallow foreshores,
like those created by extended ebb-tidal deltas, at the entrance of estuaries or harbours for
instance, infragravity waves play an important, if not dominant, role. Therefore, hydrodynamic and
topographic data collected on the downdrift side of the entrance to Figueira da Foz harbour, along
the wave exposed western coast of Portugal, were used to calibrate a local XBbeach 2DH-surfbeat
model to (1) investigate the role of infragravity waves and (2) the ability of a phase average model
to account for the main drivers of coastal overtopping in similar locations. The local model was
forced on its open boundary by water levels and 2D wave spectra dynamically downscaled using
an operational model workflow developed for providing near real-time forecasts. The local
dissipation of short-waves in the surf zone was calibrated based on the hydrodynamic data. This
data, collected under a moderate swell, was also used to ensure the model’s resolution and
numerical scheme were correctly setup to reproduce the energy and shape of the infragravity
wave’s frequency spectrum. Lastly, the model’s option to use an unconventional breaking criterion
was found useful to improve the model’s ability to reproduce an overtopping event. For this event,
which was observed and surveyed during slightly energetic waves combined with high tides,
results from the surfbeat mode were compared to results from the non-hydrostatic phaseresolving
mode of XBeach (applied with a resolution four times thinner in both horizontal
directions). In both cases, the modelled overtopping extents were similar and matched the data.
However, it was found that the wave-induced setup was much larger in the surfbeat model.
Furthermore, the extra energy brought in by accounting a fraction of the instantaneous wave
height into the equation of the wave breaking criterion allowed the water to overtop the dune
crest in similar proportion as in the phase-resolving case. So, the finely tuned surfbeat model was
run for scenarios of a storm surge with a return period of ~70 years, combined with present day
sea level and projections for 2050 and 2100. Like in the calibration runs, in these three scenarios
the wave spectra for the chosen 2014’s Hercules storm were dynamically downscaled. Again, the
inundation maps produced with this methodology were compared to those created with the
phase-resolving version of XBeach. It transpired that, for those scenarios, the wave-induced setup
and the runup of the infragravity waves were the dominant drivers of overtopping along the
waterfront in the shadow of the large ebb-tidal delta from the harbour’s entrance. Our study
therefore suggests that with minimal observation data it was possible to calibrate the phaseaveraged
version of XBeach to reproduce and map overtopping. Moreover, for similar coastal zones, where wave-setup and infragravity waves dominate, the inundation maps may be more
accurate that those produced with its phase-resolving counterpart, and this at a lower
computational cost.
