Landslide-tsunami cascade disasters, including subaerial, transitional, and submarine landslides, have claimed over 58,000 lives [1], underscoring their devastating impact. Rapid slope collapse displaces vast water volumes, generating powerful waves and widespread flooding that can breach protective barriers. A deeper understanding of these multiphase processes is essential for accurate disaster assessment and effective early warning systems. Herein, we introduced an enhanced weakly compressible Smoothed Particle Hydrodynamics (SPH) method for simulating the highly nonlinear dynamic processes of landslide-tsunami-induced high-density granular multiphase flows. In this system, the landslide body and ambient fluid are modelled as a multi-density, multi-viscosity continuum, employing a regularized viscous-inertial local rheology for predictive analysis. In particular, most particle models assume a lithostatic condition for fluid and sediment phases to predict pore-water pressure (e.g., [2]), neglecting the hydrodynamics of mixture flows. More advanced methods incorporate free-surface and interface detection but limit the ability to handle highly deformed interfacial [3]. While some using a modified equation of state (EOS), and the density of saturated sediments is prone to pressure oscillations [4]. To address this issue, we integrate the continuity equation of pore water with a dissipation formulation to improve the estimation of effective pressure within the solid skeleton. The proposed model is validated against both 2D and 3D submerged and dry sediment experiments, demonstrating strong agreement between simulations and experimental results. Overall, the qualitative and quantitative findings confirm the efficacy of the proposed method and highlight the critical role of pore pressure dissipation effects in submerged granular flows.