| Water resources are an important element of ecological protection and high-quality development in the Yellow River Basin,and with the rapid economic and social development of the Yellow River Basin,the contradiction between supply and demand of water resources in the basin is increasingly prominent,which seriously restricts the sustainable development of the basin.Water resources are the most important basic resources to maintain the natural environment and human survival and development.Due to the impact of climate change and human activities,the temporal and spatial distribution law of water resources in the Yellow River Basin has changed.Reservoir operation is an important way to realize the temporal and spatial re-distribution of water resources and solve the contradiction between supply and demand of water resources and is an extremely complex system engineering.However,environmental change not only changes the water inflow process,but also changes the water demand process,which is bound to lead to the change of reservoir operation model.Therefore,it is urgent to carry out the research on the adaptive operation of cascade reservoirs in the Yellow River Basin under the changing environment,so as to provide important theoretical support for promoting the ecological protection and high-quality development of the Yellow River Basin.This paper takes Longyangxia,Liujiaxia,Wanjiazhai and Xiaolangdi cascade reservoirs in the main stream of the Yellow River as the control object.By fully considering the spatial heterogeneity of meteorological and underlying surface conditions,the temporal and spatial evolution characteristics of meteorological and hydrological elements in the basin are investigated from the perspective of watershed.Eight conceptual lumped hydrological models are constructed to explore the impact of model inputs,parameters and structural uncertainties on water resources assessment.Meanwhile,several methods to reduce the uncertainty of runoff prediction are proposed,and their adaptability among different sub basins of the Yellow River Basin is clarified.Then,the temporal and spatial evolution laws of future water inflow(supply side)and agricultural water demand(water demand side)of important sections in the Yellow River Basin are clarified.On this basis,the joint operation model of cascade reservoirs in the Yellow River Basin is established,and the adaptive operation of cascade reservoirs in the Yellow River basin under changing environment is studied.Finally,the influence and propagation law of multiple uncertainties on reservoir operation mode and related benefits are analyzed from the perspective of history and future.The main research contents and results are as follows:(1)Based on the data of areal precipitation,potential evaporation in seven sub-basins and the annual runoff for important control sections of the Yellow River Basin from 1970 to 2016,the temporal and spatial variation characteristics of various climate elements are statistically analyzed.The results show that the annual precipitation and potential evaporation of the seven sub-basins of the Yellow River Basin are characterized by a "zigzag" change in time scale,among them,the annual precipitation of the downstream decreases with a climate slope of 4.70 mm/10a,and the annual potential evaporation of the upstream increases with a climate slope of about 6.0 mm/10a,the middle reaches decrease with a climate slope of about 4.70 mm/10a,meanwhile,the mutation years of Tangnaihai,Xiaochuan,Lanzhou,Toudaoguai,Xiaolangdi,Huayuankou and Lijin are 1989,1987,1987,1986,1992,1986 and 1988,respectively.(2)Based on the results of Morris sensitivity analysis,eight conceptual lumped hydrological models are constructed in seven sub-basins of the Yellow River Basin,and the adaptability of the models is evaluated and then the interaction between remote sensing precipitation products and model structure uncertainty is revealed,and two methods to reduce the uncertainty of hydrological simulation are proposed.The results show that the characteristics of watershed climate and underlying surface change the sensitivity of parameters,but the model parameters are transplantable in spatial scale and the simulation accuracy of eight conceptual lumped hydrological models constructed in seven sub-basins is better(KGE>0.50,R2>0.80 and RMSE<30%).At the same time,GLDAS,GPM and TRMM remote sensing precipitation products have good adaptability in the upstream,and the influence of hydrological model structure uncertainty on runoff simulation is dominant,accounting for 70%,and the interaction between remote sensing precipitation products and hydrological model structure uncertainty is weak,only 10%,and the interaction gradually increases from upstream to downstream.In addition,both the Bayesian Model Average(BMA)and Ensemble Kalman Filter(EnKF)algorithms can effectively improve the accuracy of runoff prediction,and the effect is better in downstream sub-basins.(3)Two distributed models(statistical downscaling SDSM model and AquaCrop crop growth model)are constructed to predict the future water inflow and agricultural water demand in the Yellow River Basin,respectively,and the statistical analysis method is used to reveal the temporal and spatial variation characteristics of water demand from both historical and future perspectives.The results show that the simulated and measured values of precipitation and temperature under ten climate models(BNU,CN5,CM60,CAN,GF2G,GF2M,PCM,MI5,MIEC and MRI)downscaling by SDSM model have high correlation,and the values are greater than 0.70,and the temperature is better than precipitation.In addition,on the time scale,the historical and future agricultural water demand of the basin shows a weak decreasing trend,and it is well consistent with the crop grain yield in space,on the spatial scale,the agricultural water demand in the west of the basin is small,while that in the north and east of the basin is large.(4)On the basis of obtaining the water inflow and the comprehensive water demand of the Yellow River Basin,the optimal operation model of cascade reservoirs in the Yellow River Basin is constructed and solved by particle swarm optimization to reveals the impact of environmental changes from both historical and future perspectives on the benefits of power generation and water supply of cascade reservoirs.The results show that the climate model uncertainty has little impact on the inflow in dry period,but has a greater impact in wet period,and different climate models have different effects on the inflow of different reservoirs.Meanwhile,in the historical period,BMA method has effectively reduced the number of water shortage periods and water shortage of Lanzhou and Huayuankou sections,and improved the guarantee rate of water supply and the multi-year average power generation of cascade reservoirs.In the future period,the uncertainty of climate model has a great impact on the benefits of power generation and water supply of cascade reservoirs,followed by the uncertainty of hydrological model,and the uncertainty of climate change scenario.(5)From the perspective of history and future,the adaptive operation model of cascade reservoirs in the Yellow River Basin is constructed,and the multi-stage and multi-model uncertainty decomposition method is used to quantitatively evaluate the impact and propagation law of multiple uncertainties on reservoir operation mode and related benefits.The results show that in the coupling system of“meteorological-hydrological-cascade reservoir group",the contribution of remote sensing precipitation product and hydrological model uncertainty to cascade power generation is 40%and 60%respectively,and the contribution of GPM in dry period is large and TRMM is small,and the IHACRES hydrological model has the greatest impact on the cascade power generation of the basin,but different hydrological models have different effects on the water supply benefits of Lanzhou and Huayuankou sections.In the coupling system of "climate model-climate change scenario-hydrological model-cascade reservoirs",the contributions of climate model,climate change scenario and hydrological model uncertainty to the power generation of cascade reservoirs are 45%,35%and 20%respectively,but the contribution of hydrological model to the water supply efficiency of Lanzhou and Huayuankou sections is as high as 60%,meanwhile,the multi-stage uncertainty transfer law is greatly affected by seasonality,and the interaction in dry period is weak,and the interaction in wet period is strong. |
PDF Full Text Request