Development Of Stochastic Methods For Hydrological Process Analysis And Water Quality Management

Water resource of China is scarce,which has been suffered severe pollution.The water crisis of China has been an important factor that restricts the development of national economy and improvement of people's living standards.To ensure the sustainable utilization of water resources and the quality of w ater body,advanced hydrological/water quality process analysis techniques and efficient stochastic analysis methods are desired for the water resources and watershed environment management.The water resources and watershed environment system is complex and giant,including many subsystems,such as hydrological processes,sediment transport,pollutant transportation and transformation,water quality management,effluent trading,etc.Each subsystem are related;various uncertainties exit in the subsystems.Predicting the hydrological/water quality processes as well as planning the watershed environmental system based on consideration of the uncertainties and complexities are the key issues of the giant system.Therefore,stochastic methods have been developed for hydrological process analysis as well as water quality management(effluent trading planning),which are based on the recognizaiton of uncertainties in water resources and watershed environment system.The interactions among multiple hydrological processes have been disclosed.Simulation-optimization modeling systems are developed to analyze the optimal effluent trading scheme under multiple uncertainties and system risks.A series of studies have been conducted:(1)A Markov-Chain-MonteCarlo-based multilevel-factorial-analysis(MCMC-MFA)method has been developed for assessing parameter uncertainty in hydrological model.The developed MCMCMFA method can not only assess parameter uncertainty considering prior information,model structure,observations and watershed characteristics(e.g.topography,land use,and meteorology),but also investigate the individual and interactive effects of multiple parameters on model outputs through measuring the specific variations of hydrological responses,such that the quantitative information for identifying significant factors and their interactions can be provided.The MCMC-MFA is applied in cold and arid Kaidu watershed.The results can help specify the significant parameters and the relationship among these uncertain parameters,such that hydrological model's capability for simulating/predicting water resources can be enhanced.(2)A Bayesian framework has been proposed for the evaluation of uncertainties in input data and parameters of hydrological model.The advanced Bayesian framework can help facilitate the exploration of variation of model parameters due to input data errors,as well as propagation from uncertainties in data and parameters to model outputs in s now-melting and non-melting periods.Model performances under multiple periods and calibration cases can be obtained,such that the question of how to enhance model's capability for simulating/predicting water resources in different seasons for a snowmelt-precipitation-driven watershed can be answered.(3)A robust simulationoptimization modeling system(RSOMS)has been developed for supporting agricultural nonpoint source(NPS)effluent trading planning.A case study is conducted for mitigating agricultural NPS pollution with an effluent trading program in Xiangxi watershed.RSOMS couples hydrological simulation with Soil and Water Assessment Tool(SWAT)as well as an inexact two-stage robust optimization(ITRO)model.Simulation model is used to deal with spatial and temporal variations of hydrologic elements within the watershed and provide random inputs for the optimization process based on physical mechanisms of runoff yield and routing in the watershed.Optimization model specializes in handling uncertainties expressed as probability density functions(PDFs)and interval values and capturing the notion of system risk.According to the obtained results,different robustness enforcement levels would lead to varied optimal effluent trading scheme and system benefit.The manager should balance the tradeoff between the agricultural benefit and system risk in identifying desired trading schemes for Xiangxi watershed.Besides,comparied with non-trading scheme,trading program can successfully mitigate a gricultural NPS pollution even with an increased system benefit.(4)An inexact simulation-based stochastic optimization method(ISSOM)has been developed for identifying effluent trading strategies of agricultural nonpoint sources.The ISSOM has advantages in(i)analyzing NPS contaminant transport behaviors with uncertain parameters related to nutrient yields being handled,(ii)reflecting uncertainties in trading system expressed as fuzzy,stochastic and interval formats within a multi-farmland,multi-period and multi-nutrient generation level context,(iii)analyze the flexibility in the constraints and fuzziness in the objective(which are denoted as ‘fuzzy constraints' and a ‘fuzzy goal',respectively),(iv)analyzing the effect of decision maker's preferences toward risk on system benefit and satisfaction degree.(5)A Bayesian estimation-based simulation-optimization modeling approach(BESMA)has been developed for identifying effluent trading strategies.BESMA incorporates nutrient fate modeling with SWAT,Bayesian estimation,and probabilistic–possibilistic interval programming with fuzzy random coefficients(PPI-FRC)within a general framework.Based on the water quality protocols provided by SWAT,posterior distributions of parameters can be analyzed through Bayesian estimation;stochastic characteristic of nutrient loading can be investigated which provides the inputs for the decision making.PPI-FRC can address multiple uncertainties in the form of intervals with fuzzy random boundaries and the associated system risk through incorporating the concept of possibility and necessity measures.Multiple stochastic system analysis methods have been developed for hyrological process analysis and water quality management.On one hand,hyrological process and the related uncertainty are analyzed in Kaidu watershed located in cold and arid region.Hydrological parameter uncertainty is analyzed;the contritution of the related hydrological processes to the water balance as well as the correlations among the hydrological processes are disclosed.Besides,the interactions between input data error and model parameter uncertainty are investigated.The model performances under the effect of different uncertainty sources are evaluated.On the other hand,multiple simulation-optimization modeling systems are developed for effluent trading planning in Xiangxi watershed.The modeling systems incorporate inexact simulation for nutrient movement and fate as well as effluent trading model.Inexact water quality simulation provides physical mechanisms of runoff yield and routing as well as nutrient transportation and transformation for the decision making in effluent trading.The inexact simulation can also deal with spatial and temporal variations of hydrological and water quality elements within the watershed as well as model parameter uncertainty.The simulation process provides reliable data support and science bases for water quality management.Describing the trading market and traders' activities,effluent trading model can simulate the process of achieving the optimal configuration of discharge permits through the market mechanism and free consultation.Then the optimal effluent trading scheme can be obtained.The developed analysis method for hydrological modeling uncertainties can help enhance model's capability for simulating/predicting water resources.The proposed simulationoptimization modeling systems can help establish and improve the effluent trading mechanism,provding important scientific basis for water resources and watershed environment management in China.