Understanding uncertainty in coastal modelling

Sandy beaches are among the most valuable but vulnerable environments on Earth. They continuously evolve driven by waves, sea level and sediment transport, and can be affected by storm or flooding events. Moreover, climate change is expected to intensify these processes. Numerical morphodynamic models have become essential tools for understanding coastal change and supporting future management. However, numerical modelling induces the incorporation of several uncertainties which can strongly influence the predictive capability of the models leading to different projections of coastal change.

This thesis evaluates the influence of two major sources of uncertainty: initial topobathymetric conditions and offshore hydrodynamic forcing. Using the process-based XBeach model and the reduced-complexity Q2Dmorfo model, uncertainties were investigated over short-, medium- and long-term scales coastal processes through applications based on Castelldefels Beach, Castell Beach (Palamós), and the Llobregat Delta, covering processes from individual storm events to century-scale coastal evolution.

The results demonstrate that detailed initial morphology can strongly influence storm- induced washover deposition, with certain morphological patterns increasing deposition by up to four times. For medium-term simulations, the choice of wave forcing source, particularly the accuracy of wave direction, proved to be critical for accurately reproducing the morphodynamic evolution of an embayed beach. Finally, long-term projections showed that model response is strongly controlled by the initial adaptation period. Once the model reaches a quasi-equilibrium state, the subsequent coastal evolution becomes primarily governed by the resulting topobathymetry, neglecting the influence of wave chronology.