Research On Reservoir Water Supply Operation In Tarim River Basin Based On Uncertainty Of Runoff Prediction

Water is one of the fundamental requirements for human survival and a necessary requirement for social development.Due to the acceleration of urbanization,population growth,and uneven distribution of water resources in time and space.The contradiction is becoming increasingly tense between supply and demand of water resources,and the situation of water shortage is becoming increasingly severe.One of the important ways to alleviate the contradiction between supply and demand of water resources is to carry out water supply operation for cascaded reservoirs in the watershed or cross watershed.Runoff forecasting is an input condition for future water supply operation of reservoirs.The accuracy of runoff forecasting directly affects the effectiveness of water supply operation decisions.Therefore,the uncertainty of runoff forecasting has become one of the factors affecting the water supply operation of reservoir groups.Exploring the risk laws of reservoir groups operation with forecasting runoff uncertainty is of great significance for future reservoir group operation.The risk issues faced by mountainous and plain reservoirs operating under uncertain forecasting runoff conditions in the main stream of the Tarim River Basin have been paid attention to.This paper aims at the risk problems in the operation of reservoirs in the mountainous area of the source stream of the Tarim River basin and reservoirs in the plain of the main stream under the uncertainty of runoff forecast.Runoff prediction is carried out based on the artificial neural network(LSTM)model with long-term and short-term memory.The probability distribution of forecast error is fitted.According to the forecast error distribution,the Monte Carlo method is used to simulate the forecast runoff set in dry years.The predicted runoff set serves as an input to the risk operation model for drought years.Three mountainous reservoirs DBX,STEA,and TWLW in the source stream of the Tarim River,as well as six plain reservoirs JRLK,DZ,QM,PM,KEQG,and TLM in the main stream,were considered.Through long-term certain operation,the theoretical optimal process was obtainedfor water supply risk operation of basin reservoir groups,providing a risk calculation basis and boundary constraints for the operation of water supply risks in the basin.Under the uncertainty of runoff prediction,a risk operation model with the minimum average water supply risk has been established in various sectors and industries within the watershed.The risk operation law has been analyzed by solving the model.And risk prevention and control measures have been proposed.The main research results obtained are as follows:(1)The runoff forecast of the three source flows of the Tarim River is carried out by using Long Short-Term Memory artificial neural network,and the monthly runoff forecast error is obtained.The probability distribution of the monthly runoff forecast error of the three sources of the Tarim River is fitted,and the normal distribution is the best probability distribution of the monthly monthly runoff forecast error of the three sources.According to the fitted distribution of the monthly runoff forecast error,the runoff forecast process of 1000 groups forecast under 75%frequency is simulated with the Monte Carlo method in dry year.The simulated predicted runoff can serve as input for risk operation model.(2)Using the randomly simulated predicted runoff as the input for risk operation model of water supply in the basin,the water supply risk is calculated based on the water supply volume of the optimized operation in dry year and risk definition.With the objective function of minimizing the average water supply risk in the basin,a risk operation model for the Tahe River Basin is established.The model is solved using the GA algorithm,and the results are analyzed in Tarim River.(3)Taking the predicted flow simulated by prediction error as input and the optimized regulation of water supply in the basin for a long series of dry years as the basis for risk calculation,with the minimum average water supply risk in the basin as the objective function,a risk operation model for dry years in the Tahe River Basin was established.The risk model was also solved using the GA algorithm,and the results of risk management were obtained during dry years.The law of risk was analyzed in the Tarim River Basin during dry years.(4)The uncertainty of runoff forecast brings risks to the water supply in the basin.Through the joint operation of reservoirs in the mountainous area of the source stream of the Tarim River and reservoirs in the main stream plain,the risks of industrial and domestic water supply in the basin can be completely reduced,but the risks of agricultural irrigation water supply and ecological water supply outside can’t be completely reduced.Moreover,the risk of agricultural water supply in each section of the source river is higher than that in the main stream,and the risk of ecological water supply outside the river in each section of the main stream is significant.(5)The possibility of reducing water supply risks in the Tarim River Basin was analyzed.from two aspects:reducing runoff uncertainty and regulating water resources in the basin.The conclusion is drawn that improving the accuracy of runoff prediction and the ability of watershed water resource regulation can effectively reduce the risk of water supply.