Hydrological Effects with Impact of Human Activities Based on Data Mining

With the increasing impact of human activities on the watershed hydrologic cycle, the precision of the existing flood forecasting models is declined. And the changes of watershed runoff regulation result in many difficulties in flood forecasting and reservoir operations, which directly affects the exertion of flood control and benefit of reservoir. So, it is necessary to quantitatively analyze the influence of flood forecasting and reservoir operations by human activities in order to offer reformative ways to improve corresponding level. Firstly the change-point of the impact of human activities on watershed hydrologic cycle is gained with the analysis of watershed underlaying surface changes and hydraulic projects running conditions, and it is validated by analysis on rainfall runoff trend. Then hydrological effects with human activities are stressed based on data mining method. And finally the reservoir storage scales inference model is established based on decision tree with Fengman reservoir watershed for example. It is verified that the precision of the inference model is comparative high, the corrected effect of the flood volume forecasting results using the established model is good comparatively and the model is applied to practically supervise flood forecasting and reservoir operation.