Development Of Stepwise Cluster And Stochastic Analysis Methods For Water Resource Management In Watershed

Climate change can lead to the alterations in re gional hydrological processes;which potentially results in risks of flooding and drought.Increased water demand with economic and social development,and varied water resources due to climatic and anthropogenic changes have resulted in conflict-laden issues of water resources allocation.Correspondingly,water resources management has become a challenging task due to climate change uncertainties as well as water scarcity.Therefore,the stepwise cluster analysis method and stochastic analysis method have been developed in this paper,for hydrological simulation and water resource management under uncertainty.The methods are applied in an arid-cold region in order to forecast climate change scenarios and assessment of its impacts on hydrological processes,to help managers implement mitigating policies,and to make strategic investments in infrastructure for future water resources management.In detail,(1)A hybrid factorial stepwise-cluster analysis(HFSA)method is developed for modeling hydrological processes.The HFSA method employs cluster tree to represent the complex nonlinear relationship between inputs(predictors)and outputs(predictands)in hydrological processes.A real case of streamflow simulation at the Kaidu River basin is applied to demonstrating the efficiency of the HFSA method.Results disclose that the HFSA method can formulate a SCA-based hydrological modeling system for streamflow simulation with a satisfactory fitting.In addition,the variability and peak value of streamflow in the Kaidu River basin can be effectively captured by the SCA-based hydrological modeling system.Results from 26 factorial experiments indicate that not only the minimum-temperature and precipitation are key drivers on the system performance,but also the interaction between precipitation and minimum-temperature,precipitation and relative-humidity pose significant impacts on the streamflow.(2)A stepwise-clustered downscaling model(SCDM)is advanced for transferring atmospheric simulation outputs to acquire high-resolution climate projections at a large-scale watershed system.SCDM can operate different temporal resolutions of atmospheric variables with continuous and discrete complexities.SCDM coupling with hydrological model is applied to evaluating clima te change impacts on hydrology of the Kaidu watershed in northwestern C hina.The daily and monthly series of large-scale atmospheric simulation outputs for the Kaidu watershed are exacted from the ensemble of GCMs during past(1961-1990),recent(2006-2011)and future(2015-2040)periods.Results reveal that(i)SCDM is capable of downscaling climate projections for different stations,and can help understand the spatial heterogeneity of climate change;(ii)the performance of SCDM is more acceptable for te mperature than precipitation;(iii)increase trends of Tmin and Tmax(minimum and maximum temperatures)from recent to future are projected.Besides,results from multiple downscaled climate change projections are used for driving a daily climate-streamflow hydrological model.Results disclose that the streamflow would increase because temperature change will cause more glacier melt in future.(3)An integrated multi-GCM-based stochastic weather generator and stepwise cluster analysis(MGCM-SWG-SCA)method is developed,through incorporating multiple global climate models(MGCM),stochastic weather generator(SWG),and stepwise-clustered hydrological model(SCHM)within a general framework.MGCM-SWG-SCA can investigate uncertainties of projected climate changes as well as create watershed-scale climate projections from large-scale variables.It can also assess climate change impact on hydrological processes and capture nonlinear relationship between input variables and outputs in watershed systems.MGCM-SWG-SCA is then applied to the Kaidu watershed with cold-arid characteristics in the Xinjiang Uyghur Autonomous Region of northwest C hina,for demonstrating its efficiency.Results reveal that the variability of streamflow is mainly affected by(i)temperature change during spring,(ii)precipitation change during winter,and(iii)both temperature and precipitation changes in summer and autumn.Results also disclose that:(i)the projected minimum and maximum temperatures and precipitation from MGCM change with seasons in different ways;(ii)various climate change projections can reproduce the seasonal variability of watershed-scale climate series;(iii)SCHM can simulate daily streamflow with a satisfactory degree,and a significant increasing trend of streamflow is indicated from future(2015-2035)to validation(2006-2011)periods;(iv)the streamflow can vary under different climate change projections.The findings can be explained that,for the Kaidu watershed located in the cold-arid region,glacier melt is mainly related to temperature changes and precipitation changes can directly cause the variability of streamflow.(4)An inexact joint probabilistic programming(IJPP)approach is developed for risk assessment and uncertainty reflection in water resources management systems.IJPP can dominate random parameters in the model's left-and right-hand sides of constraints and interval parameters in the objective function.It can also help examine the risk of violating joint probabilistic constraints,which allows an increased robustness in controlling system risk in the optimization process.Moreover,it can facilitate analyses of various policy scenarios that are associated with different levels of economic consequences when the promised targets are violated within a multistage context.The IJPP method is then applied to a case study of planning water resources allocation within a multi-reservoir and multi-period context.Solutions of system benefit,economic penalty,water shortage,and water-allocation pattern vary with different risks of violating water-demand targets from multiple completing users.Results also demonstrate that different users possess different water-guarantee ratios and different water-allocation priorities.Therefore,there is a tradeoff between economic objective,water-shortage and water-allocation patterns and constraint-violation risk.The results can be used to help water resources managers to identify desired system designs against water shortage and for risk control,and to determine which of these designs can most efficiently accomplish optimizing the system objective under uncertainty(5)An integrated multistage stochastic simulation-optimization(IMSSO)method is developed for assessing climate change impacts on water resources.In the IMSSO method,uncertainties presented as both interval numbers and probability distributions can be reflected.Moreover,IMSSO is applicable for permitting in-depth analyses of various policy scenarios that are associated with different levels of economic consequences when the promised water-allocation targets are violated.The results also reveal that the system uncertainties would significantly affect water resources allocation pattern(including target and shortage).It can be discovered that from t he 2016 to 2070,the drought conditions would be vulnerable,with large amount of water shortage.Besides,the more the inflow amount decreases,the higher estimated water shortage rates are.Correspondingly,water deficit in such an arid region would occur under climate change impacts,which desires effective water management measures during decision-making processes.The developed stepwise cluster analysis simulation methods are integrated with large-scale GCMs and the statistical downscaling method,for capturing nonlinear hydrological characteristics,for predicting of climate change scenarios,and for assessing climate change impacts on hydrological processes and water allocation patterns.Besides,in terms of water resource management model,the solutions of inexact optimization method can support the adjustment of the existing plans and policies,and facilitation of dynamic analysis for water allocation plans.It can not only provide scenario analysis for government departments and decision-makers,but also could offer forward-looking policies of water systems,and gradually resolve problems which decision maker may encounter in the future.