In the present study, we developed a simulation model that could predict the growth of trees and erbaceousplants for a long time. The growth of plants is formulated as the material processing of the plants, photosynthesis, respiration and mortality. Empirically obtained allometry relationship is also added in the tree model. While, the herb model was formulated by sucession of plants. Combined with the physical model, it was applied to the observed results of the vegetation succession on the sediment bar of Arakawa River after the record-breaking flood in 2007. The sufficient agreement was obtained for the vegetation succession. The model enable us to predict the growth pattern of vegetation successfully.