The Research On Medium-long Term Hydrological Forecast At XiangJiang River Basin

Medium-long term hydrological forecast plays an important role in flood control and drought relief, reservoir operation, water resources planning and management, etc. At present medium-long term hydrological forecast is a field falling behind relatively in hydrology, still in the stage being exploring and developing. The Xiangjiang River is the largest river in Dongting Lake water system. It has frequently been threatened by flood risks for a long time. In addition, the accuracy of the current medium-long term hydrological forecast remains low, which is difficult to effectively guide the production practice. Therefore, the study of medium-long term hydrological forecast of the Xiangjiang River Basin is of great significance for flood control and drought relief, as well as storage of water in the Dongting Lake.In this paper, the predictors such as rainfall and atmospheric circulation which affect the hydrological phenomenon in earlier stage at xiangtan station ,which is selected as the representative station of the Xiangjiang River, by combining cause-and-effect and statistical analysis with fuzzy analysis. Based on this, the medium-long term hydrological forecasting model of fuzzy pattern recognition , artificial neural network (ANN) and multivariate mixed regression(MMR) are set up to forecast the annual maximum peak discharge and monthly discharge. A suitable, simple, practical and higher forecast accuracy forecasting model of xiangtan station, is desired to establish.Main results and conclusions are as follows: Based on fuzzy pattern recognition, artificial neural network model and multivariate mixed regression, the qualified rates of annual maximum peak discharge and monthly discharge of xiangjiang river basin are more than 80 percent, the reasonable and high reliability of the Medium-long term hydrological forecast outcome in xiangjiang xiangtan station are authenticated mutually. From the forecasts and error results of annual maximum peak discharge and monthly discharge, it is obviously that the artificial neural network models has the smallest error and the highest forecast accuracy in the special Year of large and small discharge. Satisfactory results can be gotten when the ANN model simulating error is about 10%～11% for reference. Multiple correlation coefficient is taken as an important parameter to appraising model forecasting. When Multiple correlation coefficient no less than 0.75,the model works out satisfactorily. In a word, the Research results have important reference value to flood prevention and drought control of Xiangjiang River and the storage of Dongting Lake.