Sensitivity analysis of a morphodynamic modeling system applied to a Portuguese tidal inlet

Abstract

Coastal area morphodynamic models are prone to severe errors arising from various sources, including: (1) the use of empirical sediment transport formulae (Pinto et al., 2006); (2) the reliability of the data used to feed the models; (3) the use of simplifying physical assumptions; (4) the error propagation between the various modules. Since these errors can be significant, the usefulness and credibility of morphodynamic simulations require a thorough understanding of their uncertainty. This paper addresses the practical implications of these input parameters on the development of a tidal inlet. A sensitivity analysis is performed through the application of the morphodynamic modeling system MORSYS2D (Fortunato and Oliveira, 2004) to the Óbidos lagoon (Oliveira et al., 2005), a small but very rapidly evolving coastal system located in western Portugal. The influences of: (1) sediment characteristics; (2) the choice of the forcing tide and (3) the sediment transport formula are analysed, namely through the inlet cross-section evolution and the ebb-delta development. The choice of the forcing tide appears important, since the use of a real tide, rather than a representative tide, induces: (1) faster morphological changes; (2) 15-day cyclic evolutions (spring-neap tidal cycle); (3) larger ebb-deltas and inlet cross-sections. Sediment grain size and empirical transport formulae rather affect the rapidity of morphological changes, since equilibrium is reached after 3 months of simulations and final inlet morphologies are noticeably comparable. Nevertheless, the good agreement of the results after three months of simulation demonstrates the reliability of the modelling system on the time scales of months.