Physical and Chemical Heterogeneity of Aquifers: A Quantitative Geological Investigation

9526795戴维斯今天，水文地质学家面临的主要挑战之一是预测地下的溶质运动。人们普遍认为，含水层渗透性的空间变化是污染物羽流演化的主要控制因素之一。此外，最近的研究表明，含水层化学性质的空间变化也可能影响溶质的运移。地统计学方法是量化地下可变性和相关污染物迁移的一种非常有前途的方法。地统计或随机模型用空间相关的随机场来表示含水层的属性。近年来，文献中提出了各种各样的统计模型来表示渗透率的空间变化。其中一些模型还被应用于化学性质的空间变化。一般而言，每个模型都基于一种不同的方法，用于数学表示或计算生成空间相关的随机变量，并且每个模型在外观上都表现出总体差异。这项研究调查了各种统计模型如何与沉积的地质过程相关，以及如何利用地质信息来选择最合适的模型并将其参数化。前人的工作结果表明，河流边界面的沉积学特征与形态和空间相关性的地质统计特征之间存在着可量化的关系。河流边界面在本质上是分层的，已经观察到，一般来说，边界面的阶数越高，表面分隔具有不同平均渗透率的区域的可能性就越高。不同平均渗透率的区域依次控制着非均质性统计模型的形态和空间相关性结构。关于含水层化学性质的空间变异，很少有关于变异空间模式的现场数据。这项研究将获得有关化学非均质性的性质及其与沉积地质过程的关系的重要数据。研究的主要目的是：1)经验研究渗透率相关结构与层状边界面空间分布的关系；2)研究化学性质的空间变化与地质特征的关系；3)开发一种用于模拟物化非均质性的沉积环境沉积学信息的编码算法。第一阶段将调查新英格兰砾石坑中暴露的沉积物的渗透性和化学性质(矿物学、颗粒涂层、表面积、孔隙度和阳离子交换能力)的空间变化。野外工作的结果将被用于开发一种模拟含水层非均质性的算法。预计该算法将使用指示器形式主义和贝叶斯统计分别处理不同尺度的异质性。每个地质环境都是独一无二的，任何特定的露头研究都不会直接转移到其他地点。然而，如果物化非均质性能够与一般地质性质和沉积学特征相关联，那么这项研究的结果将使不同地质背景之间的可携性更强。

9526795 Davis One of the major challenges facing hydrogeologists today is the prediction of solute movement in the subsurface. It has been generally recognized that the spatial variation of aquifer permeability is one of the dominant controls on the evolution of a contaminant plume. In addition, recent studies suggest that the spatial variation in aquifer chemical properties may also influence solute movement. One very promising approach to quantifying subsurface variability and the associated contaminant transport is the geostatistical approach. Geostatistical, or stochastic, models represent aquifer properties with spatially correlated random field. In recent years, a wide variety of statistical models have been proposed in the literature to represent the spatial variation of permeability. Some of these models have also been applied to the spatial variation of chemical properties. In general, each model is based on a different method for mathematically representing, or computationally generating, spatially correlated random variables and each exhibits overall differences in appearance. This research investigates how the various statistical models relate to the geological processes of deposition and how geologic information can be used to select and parameterize the most appropriate model. Results from previous work indicate that a quantifiable relationship exists between the sedimentological features of fluvial bounding surfaces and the geostatistical properties of modality and spatial correlation. Fluvial bounding surfaces are hierarchical in nature and it has been observed that, in general, the higher the order of the bounding surface, the higher the probability that the surface separates regions with different mean permeability. The regions of different mean permeability in turn are the dominant control on modality and spatial correlation structure of the statistical models of heterogeneity. With regard to the spatial variation of aquifer chemical properties, very little fi eld data on the spatial patterns of variability. This research will obtain important data on the nature of chemical heterogeneity and its relationship to the geological processes of deposition. The primary objectives of the proposed research are to: 1) empirically investigate the relationship between the correlation structure of permeability and the spatial distribution of hierarchical bounding surfaces; 2) investigate the relationship between the spatial variation of chemical properties and geologic features; and 3) develop an algorithm for coding sedimentological information of depositional environment for the simulation of the physical and chemical heterogeneity. The first phase will investigate the spatial variation of both permeability and chemical properties (mineralogy, grain coatings, surface area, porosity, and cation-exchange capacity) of deposits exposed in New England gravel pits. The results of the field work will be used to develop an algorithm of simulating aquifer heterogeneity. It is anticipated that the algorithm will treat different scales of heterogeneity separately using the indicator formalism and Bayesian statistics. Each geologic setting is unique and any particular outcrop study will not be directly transportable to other sites. However, if the physical and chemical heterogeneity can be related to general geologic properties and sedimentological features, the results of this research will enable much more portability between different geologic settings.