This paper presents the generation of unstructured triangular grids based on the sediment data obtained from the erosion and sedimentation map of the Dez Dam watershed, covering an area of 17,320 km². The proposed grid generation algorithm starts the refinement process with a coarse initial grid. The grid refinement is an iterative process that simultaneously employs both point insertion and cell subdivision methods. Edge-based and cell-based connectivity matrices have been employed to accelerate the retrieval of grid information. An advantage of using these matrices is the minimal number of computations required to extract information for each grid element (vertex, edge, and cell). For instance, the database can identify the cells adjacent to an edge using only two computational operations. The improvement of the final grid quality is another aspect addressed in the proposed grid generation algorithm. The novelty of this paper lies in the use of erosion and sedimentation data for the adaptive distribution of grid vertices. In this study, erosion distribution data obtained from the PSIAC model were utilized to define edge sources. From a performance standpoint, the proposed algorithm produces unstructured grids in less than one second. More than 70% of the grid cells exhibit shapes close to equilateral triangles.