Proposal Technique of Parameter Range that made Optimal Performance in Coastal Modeling using Process-based Model

Abstract

The process-based XBeach model has many empirical parameters because of an inadequate understanding of sediment transport and hydrodynamics on the nearshore. Thus, it is essential to calibrate parameters for use in various studies, research, and wave conditions. Therefore, parameter calibration is necessary for improving prediction performance. Generally, trial-and-error is universally used. However, this method is passive and restricted to comprehensive and various parameter ranges. In this research, the GLUE (Generalized Likelihood Uncertainty Estimation) technique was used to estimate the optimal parameter range of 3 parameters gamma, facua, gamma2) using morphological observed data collected in Maengbang beach during the 4 typhoons from September to October 2019. The prediction performance and optimal parameter range were evaluated using BSS (Brier Skill Score) along with the sensitivity, baseline profiles, and likelihood density of the BSS value in the GLUE. Accordingly, the optimal combinations of parameters were derived when parameter facua was below 0.15 and simulated well the moving shape, from crescentic bars to alongshore uniform bars, in the surf zone of Maengbang after storm events. However, the accretion and erosion patterns nearby in the surf zone and coastline remain challenges in this numerical model.