<p>The river flow, sediment transport, bed deformation and contaminant transport are strongly coupled under the impacts of climate change and anthropogenic activities. Two-dimensional (2D) mathematical models are practical tools for predicting the morphodynamic processes and contaminant transport while their computational efficiency is a key impediment to simulations featuring large scales of time and space, especially when refined meshes are required for key spot such as hydraulic structures. Therefore, a computationally efficient and fully coupled 2D model for sediment-borne contaminant transport is established under the framework of Finite Volume Method on unstructured meshes in this paper. The computing strategies of local time stepping, parallel computing and high performance clusters are employed, which can speed up the computation and reduce the run time drastically. The model is tested against three experimental cases and one idealized test case, where numerical solutions are in general agreement with the measured data and analytical solution.</p>