Forecasting Models for the Ten-day Streamflow of Kao-Ping River

The nonlinearity and uncertainty of the runoffs process are such that estimating or predicting required hydrologic data is often tremendous difficult. Consequently, this study applies three forecasting models to predict the ten-day streamflow resulting from rainfalls on the Kao-Ping river watershed. The three techniques employed in establishing the models include: time series, Grey system theory, and the adaptive fuzzy neural network model. Through the research it is anticipated a more appropriate hydrologic data series forecasting model for the Kao-Ping river watershed can be identified.