Abstract
We have developed and implemented a real-time dam inflow prediction system based on a distributed rainfallrunoff model. The system predicts dam inflow for 6 hours every 10 minutes, receiving predicted rainfall data and calculating hillslope runoff and river flow on channel network. River flow is routed with the lumped kinematic wave model, of which state variables are updated by the Kalman filter with observed inflow data of the dams.Additionally, to improve the accuracy of prediction for practical use, we estimate effective precipitation using a runoff coefficient which varies with cumulative rainfall, and introduce addition of base flow discharge considering influence of snow melt.The system has been already installed into the dam administration offices on the Kurobe-River basin and is being used for practical purposes.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
