Predictive Simulation And Optimal Allocation Of Surface Water Resources In Reservoir Basins Under Climate Change

It is crucial to improve the comprehensive management of water resources with limited total water resources,to cope with the shortage of regional water resources and low water resource utilization efficiency.As a component of water resources management,water conservancy engineering measure plays a critical role in the optimal allocation of surface water resources.Among them,the reservoir,as a widely used water conservancy engineering measure,can effectively alleviate seasonal water deficiency problems and reduce the risk of water shortage through the collection,storage,and allocation of surface water resources.Reservoirs play a critical role in the supply and regulation of surface water resources.Surface water resources as a significant source of water supply are essential for promoting socioeconomic development and protecting ecological systems.However,climate change has a major influence on surface water resources(e.g.available water resources in reservoir basins)by affecting meteorological elements,which have become increasingly severe in recent years.Such consequences have greatly pressured the prediction of available water resources in reservoirs and have even caused a risk of water shortage.Predictive simulations and optimal allocation of water resources in reservoir basins are becoming increasingly important to tackle the risk of water shortages and the challenges of climate change,especially in reservoir basins.Surface water resource management in reservoir basins is extremely complex and includes multiple objectives and uncertainties.Therefore,the effective treatment of systematic uncertainties and complexities under climate change has become an important and challenging issue in surface water resource research in reservoir basins.Advanced models must be proposed to support predictive simulation and optimal allocation of water resources,which are urgently required to ensure regional water supply security.The innovation of this research is that it proposes a model for predictive simulation and optimal allocation of surface water resources in reservoir basins under climate change.SWAT,interval parameter programming,multi-objective programming,and climate change scenarios were combined to:(1)predict the utilizable amount of surface water resources under climate change based on the identified uncertainty of hydrological model prediction simulation,(2)address system uncertainties expressed as discrete intervals,and reflect the multiple objectives of water resource management in reservoir basins.The proposed method was applied to the Xinfengjiang(XFJ)reservoir basin in South China.The multiple results obtained under different scenarios can provide policymakers with a series of water resource management alternatives for adaptation to climate change.Thus,the proposed model provides a more reliable method and decision-making support for water resource management in reservoir basins.Specific research contents and main results are as follows:(1)Collect data such as digital elevation model,land use map,soil map,and daily climate data in the XFJ reservoir basin to construct the SWAT model to simulate basin hydrological processes.Calibrate the parameters of SWAT model based on long-term monthly observed data of runoff at the Shuntian and Yuecheng hydrological stations.The Nash–Sutcliffe efficiency(NSE)index and determination coefficient(R~2)were used to evaluate the applicability of the SWAT model.The results show that the calibration and validation results for the monthly runoff at Yuecheng and Shuntian stations meet the model accuracy requirements of R~2>0.6,NSE>0.5.The SWAT model could simulate runoff in the XFJ reservoir basin and can therefore be used for subsequent research of optimal allocation of water resources.(2)Carry out parameter uncertainty analysis of the SWAT model by SUFI-2(Sequential Uncertainty Fitting Process version 2)technique in SWAT-CUP(SWAT calibration and uncertainty procedure).Uncertainty degree can be identified using two factors:the P-factor and R-factor.Based on the goal-value of sensitivity parameter groups with different values in the SWAT-CUP uncertain analysis results,several parameter groups suitable for XFJ reservoir basin were selected to modify the SWAT model parameters.Collect the daily precipitation,minimum temperature,and maximum temperature in 2025 under RCP(Representative Concentration Pathway)2.6,4.5,and 8.5,respectively,in the Hadley Center Global Environment Model.The available water resources in reservoir basins could be predicted by inputting meteorological data into the calibrated SWAT model.The interval characteristics of surface water resources in the reservoir basin could be analyzed by taking the maximum and minimum monthly runoff of the reservoir basin under each RCP scenario.The results show that the P-Factor of Shuntian and Yuecheng station is 22%and 19%,respectively.The R-Factor is 0.01 and 0.01,respectively.Thus,the SWAT model has uncertainty associated with model parameters.It is necessary to select several parameter groups applicable to XFJ reservoir basin to analyze the interval uncertainty of surface water resources in the basin.Based on 16 groups of parameters applicable to XFJ reservoir basin,the average annual runoff in the forecast year(i.e.,2025)from the reservoir outlet under the RCP 2.6,4.5,and 8.5 scenarios would be[235.2,350.7],[302.0,407.0],and[274.6,393.6]m~3/s,respectively.(3)The water resource allocation processes in reservoir basins are highly complicated.Many parameters,such as reservoir storage,power generation efficiency,and economic cost coefficients,cannot be assigned deterministic values and are presented as interval numbers instead.Concurrently,reservoirs typically have multiple service functions that affect and restrict each other.Therefore,interval parameter programming was introduced into the multi-objective programming framework,developing an interval multi-objective programming model for optimal allocation of regional surface water resources in reservoir basins.The decision variables of the model included the water supply quantity,water quantity for power generation,and surplus water quantity from the reservoir in different months.The primary objectives were to maximize power generation,maximize the economic benefit of water supply,and minimize the surplus water quantity.These objectives were limited by a series of constraints,including water balance,water quantity for power generation,water quantity released from the reservoir,total power generation output,reservoir storage,and non-negativity constraints.The result shows that the optimal water quantity for power generation in the forecast year,2025,would be the largest at～60%of all water resources.Specifically,the optimization results predicted that the water quantity for power generation from April-September 2025 accounted for approximately 75%of all allocated water resources.The total power generation of the XFJ reservoir in 2025 will be[1.20×10~9,2.00×10~9]k Wh.The optimal water quantity for water supply would be the same,(i.e.,[5.48×10~8,5.50×10~8]m~3),with a proportion of～35%.Comparatively,the optimal surplus water released from reservoirs would be the lowest,not exceeding 5%of all water resources.In addition,climate change and reservoir initial storage would greatly affect surplus water quantity but not power generation and water supply quantities.In general,the model shows good applicability in predictive simulation and optimal allocation of surface water resources in reservoir basins under climate change.Multiple complexities and uncertainties in water resource prediction and management systems can be addessed.The multiple results obtained under different scenarios can provide policymakers with a series of water resource management alternatives for adaptation to climate change.Thus,the proposed model provides a more reliable method and decision-making support for water resource management in reservoir basins.