<p>This study investigates the morphodynamic evolution of gravel beach profiles under regular wave conditions through a combination of physical experiments and process-based numerical model XBeach-G. The experiments were conducted in a wave flume with two different beach face slopes (1:5, 1:7) and two coarse sediment types (<i>D</i>₅₀ = 2.6 mm and <i>D</i>₅₀ = 7.1 mm). Hydrodynamic and morphological data, including free surface elevations, flow velocities, pore water pressures, and beach profile changes, were collected. The focus of this study on beach profile evolution lies in applying the Generalized Likelihood Uncertainty Estimation (GLUE) method, as proposed in XBeach-G, to improve the model's accuracy in predicting beach deformations. The results shows that the seepage effect significantly influences wave height attenuation, velocity envelope and its asymmetry, nonlinear dynamic characteristics, plunging breaker formation and propagation, and the seepage velocity of the top sediment layer. These factors collectively play a crucial role in the spatiotemporal deformation of beach profiles. In addition, this study evaluates the applicability of the SHINGLE model in predicting profile deformation. The results indicate that the model exhibits certain deviations under specific conditions. Furthermore, the impact of the Iribarren number on berm development and the critical threshold for classifying beach profile types are examined. The MLRA method is employed to rank and assess key factors influencing berm development. Furthermore, we demonstrate that the GLUE method improves model calibration by systematically optimizing key empirical parameters, leading to more accurate predictions of berm formation and erosion patterns. These findings provide new insights into the hydrodynamic and morphodynamic processes governing gravel beaches and contribute to refining numerical modeling approaches for coastal management applications.</p>