Evaluating longshore sediment transport: A comparison between empirical formulas and XBeach 2DH numerical model

Longshore sediment transport (LST) is a crucial process shaping coastal environments. As multiple site-specific factors (e.g. wave height, coastal topography) influence the LST, an accurate estimation of LST rate represents a scientific challenge. This research evaluates the predictive capacity of three widely used empirical formulas (CERC, 1984; Kamphuis, 1991; Van Rijn, 2014) and the XBeach 2DH numerical model for estimating the LST rate by comparing the obtained results with field LST data. We chose three coastal sites in Malta Island (Għadira Bay, GB) and northern Sicily (Cefalù, CF, and Campofelice di Roccella, CR) based on different sediment grain size, coastal type (open or embayed) and morpho-bathymetry. For each site, we analysed wave parameters, grain size of the beach and seabed sediments, coastal morphology, and marine vegetation distribution. Furthermore, we used field measurements to calibrate the numerical model's sediment transport and morphological parameters. The calibrated numerical model provides greater accuracy in LST rate estimation than the empirical formulas. The latter overestimates the LST rates by factors ranging from 435 to 7885, whereas the numerical model overestimates by factors of 1.8 and 1.9 at the CF site and GB, respectively, and underestimates by a factor of 0.5 at the CR site. The good performance of the numerical model is due to its consideration of site-specific factors. The parameter values for the model calibration can be used successfully in embayed fine/coarse sandy beaches. Moreover, the numerical model, tested so far only for sandy beaches, also works well on gravelly beaches.