Search for a command to run...
Five suspended sand concentration profile (C[z]) models, four existing formulations and one modified model, were evaluated against 119 laboratory tests from four large-scale datasets covering multiple cross-shore zones under regular and irregular breaking waves. Existing formulations were grouped into power-law and exponential models, producing mildly parabolic and quasi-linear profiles, respectively. The classical Rouse profile was found unsuitable near the bed and free surface, yielding unrealistic concentrations as z → 0 and z → d, and therefore recommended only when the local water depth greatly exceeds reference elevation. Power-law models derived from the Rouse formulation exhibited excessively steep profiles due to small mixing parameter values, leading to quasi-uniform concentrations throughout the water column. In contrast, measured profiles were consistently concave, with elevated near-bed concentrations and reduced values in the upper water column, resulting in substantial over-prediction by power-law models aloft. To address this limitation, an empirical correction factor was introduced to enhance the mixing parameter, producing a modified SR93-based formulation (L19 C[z]) that yields more realistic, smoothly varying profiles. The L19 model demonstrated superior agreement with observations across all cross-shore zones, with root-mean-square errors ranging from 0.03 to 2.3 kg m⁻³ (mean 0.56 kg m⁻³), outperforming existing models. Exponential formulations reproduced concentrations near the bed and surface but diverged in the mid-water column and were highly sensitive to reference concentration specification. Finally, the L19 model was validated within a process-based morphodynamic framework using field data from the Hasaki Oceanographic Research Station, achieving a Brier Skill Score of 0.72 under storm conditions.