【作者】 梁婕；

【导师】 曾光明； 郭生练；

【作者基本信息】 湖南大学 ， 环境工程， 2009， 博士

【摘要】 地下水污染系统中的溶质运移和污染风险研究是地下水资源管理的重要研究内容之一。然而,地下水环境系统中广泛存在的不确定因素影响着研究的精度,日益成为地下水资源管理的障碍。首先总结了地下水污染系统的主要特征,综述了地下水溶质运移模拟、地下水污染风险和地下水环境系统中不确定性理论的研究进展,在此基础上指出研究方法:第一,基于随机、区间、模糊数学方法的基本不确定理论;第二、基于随机建模的参数非均质性研究方法;第三,基于贝叶斯定理的不确定参数识别技术。研究工作和研究成果包括以下内容:(1)针对渗透系数的非均质性,采用高斯随机场模型和序贯高斯模拟生成渗透系数的空间分布;针对高斯随机场模型中参数的识别问题,采用基于贝叶斯定理的不确定性参数识别技术推断参数的后验分布;基于上述两种不确定因素,研究非均质含水层中溶质运移问题。研究发现,贝叶斯推断的参数后验分布并非对称分布,条件数据越多,参数后验分布的标准方差越小,参数分布的集中趋势越明显;渗透系数不确定性对地下水溶质运移的影响较大,地下水溶质运移的不确定程度随着时间的推移而增加,随着条件数据集数量的增加而减少。(2)针对非均质含水层中溶质运移的不确定问题,提出一种非均质含水层中溶质运移优化模拟的方法。该方法以渗透系数的测量值和溶质浓度的观测值为条件数据,用新型的变尺度混沌-遗传算法同时更新渗透系数和含水层中污染物浓度的空间分布。研究结果表明,渗透系数对参数的优化结果有重要作用;当渗透系数和溶质浓度同时参与优化过程时,渗透系数分布和溶质浓度分布的模拟精度都将大大提高。结合两类数据的优化过程有效平衡了地质参数和溶质浓度两种信息。(3)针对含水层中污染物监测浓度的变化引起的不确定性,以区间数表示污染物浓度,应用区间数理论和模糊属性识别方法,建立了基于区间数的地下水环境健康风险模糊综合评价模型。实例研究表明基于区间数的地下水环境健康风险模糊综合评价在一定程度上解决了评价过程中存在的不确定性问题,评价的结果更全面、合理地反映了地下水环境健康风险水平的真实情况。(4)针对美国环保局的健康风险评价模式中参数的不确定性,采用序贯指示模拟生成若干污染物浓度分布以反映浓度的不确定性,建立概率密度函数以反映其它参数(日饮水量、体重等)的不确定性,采用Monte Carlo模拟技术,建立了基于序贯指示模拟的水环境健康风险分析模型。实例研究表明,该模型可基本反映污染的环境风险,含水层中污染物浓度的不确定性是造成人类健康风险不确定的主要因素。(5)针对环境风险的主体和受体,提出环境风险等于含水层“脆弱性”与地下水污染造成人类健康“危害性”的逻辑乘积。同时考虑“脆弱性”的随机性和风险的模糊性,建立了基于随机模拟的地下水污染风险模糊评价模型。实例研究表明,该模型可以清晰的表达风险的等级,与确定性模型和单一方面的风险评价模型计算结果有显著差异。针对复杂的地下水环境系统,基于不确定理论,较为系统的剖析了地下水环境系统中的不确定因素对地下水溶质运移,建立了若干基于不确定性理论的地下水污染风险模型。一方面,有助于深刻揭示不确定因素对地下水溶质运移和污染风险的影响,在本质上加深对地下水环境系统不确定性的认识,具有以一定的理论意义;另一方面,以湖南省“两型社会”建设区长沙-株洲-湘潭地下水污染、长沙黄兴镇地下水锰污染为例证明,基于不确定理论的地下水溶质运移和污染风险研究,可以更加有效的反映复杂地下水环境系统的行为,更有效的指导地下水资源管理和污染防治工作,具有一定的实践意义。更多还原

【Abstract】 Solute transport and contamination risk assessment in groundwater is one of the important research fields of groundwater resource management. However, there are a wide range of uncertainties affecting the accuracy of the research. And these uncertainties are increasingly becoming an obstacle to the management of groundwater resources.This dissertation first summarized the main features of the groundwater contamination environemntal system, reviewed the research progresses of solute transport, environmental risks, and uncertainty theory and their applications to the groundwater environmental system. Based on that, it was pointed out that there were three types of uncertain theory that used in the disseration. The first type is the basic uncertainty theory, such as stochastic mathsmatic methodology, interval mathsmatic methodolody and fuzzy mathsmatic metholodgy; the second is the heterogenity of the parameters and their stochastic modeling techniques; and the third is the uncertain parameter identification techniques based on Bayesian theory. The main works and novel results include the following elements. (1) Aiming at the heterogenity of the hydraulic conductivity, Gaussian Random Space (RSF) function was used to discribe the distribution of hydraulic conductivity, and aiming at the parameter identification problem, the uncertain parameter identification techniques based on Bayesian theory was applied to infere the posterior distributions of parameters in RSF. The solute transport in the heterogenious aquifer was studied under the two kinds of uncertainties. The results showed that the posterior distributions of parameters were dissymmetrical. When more conditional data were incorporated in the analysis, the standard deviation of parameters was smaller, the centralized trend was more obvious, and the optimal parameters were closer to the hypothetic ones. Hydraulic conductivity had great effects on solute transports in groundwater. Taking the effect of hydraulic conductivity uncertianty into account, the uncertainty for solute transport increased with the increasing of time and decreased with increasing of the number of conditional data. (2) Aiming at uncertainty associatied with solute transport in heterogeneous formations, a coupled inverse modeling metholodgy about optimal solute transport was developed. Conditioning to the measured hydraulic conductivity and observed concentration data, the hydraulic conductivity field and concentration distribution were regenerated at the same time using a new type of algorithm-Mutative Scale Chaos Genetic algorithm (MSCGA). The results showed that the description of hydraulic conductivity and solute concentration filed was improved when both kinds of data were combined in the coupled inverse modeling system. The coupled inverse modeling provided a trade-off between geological setting and solute transport. (3) Aiming at uncertainty associated with different mintoring data, interval number was used to respersent contaminants concentration. Based on the application of interval number theory and fuzzy attributed recognization technique, an intergrated fuzzy model for groundwater environmental health risk assessment was established. The results of case study showed that the hydraulic conductivity was important for the results of parameter identification. The model could resolve the uncertainties in reality to some extent. And compared with groundwater environmental health risk assessment model under certainty, the results of the established model were more comprehensive and reasonable in reflection of the situation of environmental health risks associated with groundwater contamination. (4) Aiming at uncertainties of the parameters in the health risk assessment model developed by the United States Environmental Protection Agency (USEPA), a spatial stochastic risk analysis model was developed. In the model, the concentration of contaminants was simulated with sequential indicator simulation, and the other parameters were sampled from their probability density functions. The results of case study showed that the developed model could basically reflect the human health risk of groundwater pollution. The contaminants concentration was the main factors of uncertainty of the resulted risk. (5) Aiming at the subjects and receptors, the environmental risk was considered to be the“vulnerability”of the aquifer multiplied by“hazard”caused by groundwater contamination logically. Taking the stochastic characteristic of“vulnerability”and fuzzy characteristic of risk into account, an intergrated stochastic-fuzzy modeling approach for risk assessment of groundwater contamination was developed. The results of case study showed that the developed approach offered a unique tool for systematically quantifying various uncertainties and clearly expressing the level of risk.Due to the complexity, unique transport rules and various internal reactions of groundwater environmental system, this disseration analyzed the effects of uncertainties of solute transport in groundwater and developed several models assciated with contaminated risk of groundwater under uncertainties. On one hand, this disseration is helpful to reveal the character of uncertainties in solute transport and contamination risk of groundwater system, and to better understand the uncertainties in groundwater environmental system, which was of theoretical significance in certain degree. On the other hand, the results of typical case studies, such as Chang-Zhu-Tan region for“Resource-saving and Environment-friendly Society”construction and manganese contamiantion in groundwater of Huangxing Town, proved that solute transport in groundwater and contamination risk anlysis under uncertainties could effectively reflect the complex behavior of groundwater system and provide scientific support to the management of groundwater resource and control of groundwater contamiantion.更多还原