Medium And Long Term Prediction Model Of Natural Runoff In The Middle And Lower Reachesof The Yellow River

The hydrological forecast is based on the existing information to predict qualitatively or quantitatively the hydrological state in the future period.How to improve prediction accuracy with limited data is a concern for hydrological workers.The medium and long term forecast of runoff is one of the important kinds of medium and long term forecast for hydrology,which plays a great role in flood control,drought resistance,exploitation and utilization of water resources,ensuring the safe production of industry and agriculture,and developing the significance of water conservancy facilities.There are many models and methods for middle long term runoff forecasting at present.However,since the applicability of the model is different,it is more practical to select the model according to the characteristics of the basin and the features of runoff,and to carry out the middle and long term forecast of runoff.Taking the river runoff of several typical hydrological stations in the middle and lower reaches of the Yellow River as the research object,this paper comprehensively analyzes the traditional mature forecasting methods and the current not yet mature prediction methods,and summarizes the application characteristics and advantages and disadvantages of each model in different time scales,in order to find a better prediction model as valuable reference provided to improve the medium and long term forecast for middle and lower reaches of the Yellow River.Five representative forecasting methods are studied and applied to the study on monthly runoff,annual runoff,and periodical forecast of multi-year runoff in the typical hydrological stations of the middle and lower reaches of the Yellow River.The main contents and results are as follows:In the study of monthly runoff forecast in the middle and lower reaches of the Yellow River,the time series analysis method is used to deal with the seasonal variation of non stationary sequence.The average monthly runoff data of the Yellow River HuaYuanKou hydrologic station of 1975-2012 is selected,and the Eviews software is used to convert the obtained sequence into a stationary time series after the seasonal difference and establish the analysis model.Similarly,the fitting model is established by a variety of time series methods.After analysis and comparison,it is concluded that AR(2)model has the best fitting effect,but the accuracy of the forecast period still needs to be improved due to the poor smoothness of monthly runoff series and the tolerableness of precision of the model fitting period.In the study of annual runoff forecast in the middle and lower reaches of the Yellow River,the characteristics of the series of "small sample,poor information and uncertainty" are mainly studied by the grey system theory,and the annual runoff data of the the Yellow River Huayuankou hydrological station of 1965-2012 and Shuokou hydrological station of 1960-2007 is selected to establish the differential equation,and the results are measured and tested.It shows that the prediction effect is not ideal.After comprehensive analysis,it is found that the random volatility of the original sequence is too large,then the grey Markov model of Markov chain and GM(1,1)model is used to model the annual runoff of the Yellow River Huayuankou hydrological station of 1965-2012.After prediction,the relative error is basically controlled at 10%,and the prediction accuracy is relatively high.In the study of the annual runoff periodic prediction in the middle and lower reaches of the Yellow River,selecting the annual runoff of 1951-2012 of the Yellow River Huayuankou hydrological station,we obtain the number series of dead water years according to The Standard of Hydrological Information Forecast and gain the tolerance period of 13 years in the dry year of this area based on the commensurable network diagram.Therefore,it is preliminarily judged that the next dry water season will be in 2018 or 2019.At the same time,we also use the characteristics of "60 years"circulation cycle to draw the conclusion that the next dry season of the Yellow River Huayuankou will appear in 2019.