Integrated Stochastic Evaluation of Flood and Vegetation Dynamics in Riverine Landscapes

Abstract:

Areal expansion of trees on gravel beds and sand bars has been a serious problem for river management in Japan. From the viewpoints of ecological restoration and flood control, it would be necessary to accurately predict the vegetation dynamics for a long period of time. This presentation tries to evaluate both vegetation overgrowth tendency and flood protection safety in an integrated manner for several vegetated channels in Kako River, Japan. The predominant tree species in Kako River are willows and bamboos. The evaluation employs a stochastic process model, which has been developed for statistically evaluating flow and vegetation status in a river course through the Monte Carlo simulation. The model for vegetation dynamics includes the effects of tree growth, mortality by flood impacts, and infant tree invasion. Through the Monte Carlo simulation for several cross sections in Kako River, responses of the vegetated channels are stochastically evaluated in terms of the changes of discharge magnitude and channel geomorphology. The result shows that the river channels with high flood protection priority are extracted from the several channel sections with the corresponding vegetation status. The present investigation suggests that the stochastic analysis could be one of the powerful diagnostic methods for river management.