首页> 外文学位 >Development and evaluation of local grid refinement methods for forward and inverse groundwater models.
【24h】

Development and evaluation of local grid refinement methods for forward and inverse groundwater models.

机译:开发和评估用于正向和反向地下水模型的局部网格细化方法。

获取原文
获取原文并翻译 | 示例

摘要

This work develops a new method of local grid refinement for block-centered finite-difference groundwater models using shared nodes, implemented in conjunction with MODFLOW-2000. Unlike one-way coupled telescopic mesh refinement methods that are commonly used for groundwater models, the shared node method developed here contains a feedback from the embedded grid that influences the outer grid. In contrast to directly embedded approaches, which require the solution of irregular matrix equations because of the irregular finite-difference stencils at the grid interface, the shared node method uses an iterative coupling between the grids and maintains a regular matrix structure. This allows exploitation of efficient matrix solvers based upon standard finite-difference stencils.; The accuracy and computational efficiency of the new local grid refinement method is put into context by comparisons to several other grid refinement methods including variably spaced grids and one-way coupled methods. Test cases in homogenous and heterogeneous, two and three dimensions, and linear and nonlinear flow regimes are investigated. This constitutes the most thorough comparison of these methods for groundwater flow problems known to the author. Results indicate that the shared node method is a compromise between the accuracy of the computationally intensive but inflexible variably spaced grids and the efficient CPU time of the less accurate but flexible one-way coupled methods.; This research also investigates the performance of several grid refinement methods during inverse groundwater modeling. Results indicate that the one-way coupled methods perform poorly because they lack a feedback to relate parameter sensitivity at both grid scales. In contrast, the feedback present in the shared node method produced good results in all cases tested.; Local grid refinement was evaluated as a tool to address the scale issues of representing the features of surface water-groundwater interactions. Results of this investigation indicate that (1) highly refined grids are necessary to represent the small scale hydraulic features of stream-aquifer interactions, and (2) the stream-aquifer connection is grid-scale dependent. The shared node method was able to represent the stream-aquifer interaction with reasonable accuracy (0.6–3.9% error), while a one-way coupled method resulted in much larger errors (6–20%).*; *This dissertation is multimedia (contains text and other applications not available in printed format). The accompanying CD requires the following system requirements: Adobe Acrobat; Microsoft Office.
机译:这项工作结合了MODFLOW-2000,为使用共享节点的以块为中心的有限差分地下水模型开发了一种新的局部网格细化方法。与通常用于地下水模型的单向耦合伸缩网格细化方法不同,此处开发的共享节点方法包含来自嵌入式网格的反馈,该反馈会影响外部网格。与直接嵌入式方法相反,由于网格接口处存在不规则的有限差分模板,因此需要求解不规则矩阵方程,而共享节点方法则使用网格之间的迭代耦合并保持规则的矩阵结构。这允许基于标准的有限差分模板开发有效的矩阵求解器。通过与其他几种网格细化方法(包括可变间距的网格和单向耦合方法)进行比较,将新的局部网格细化方法的准确性和计算效率置于上下文中。研究了均质和异质,二维和三维以及线性和非线性流态下的测试案例。这构成了作者已知的针对地下水流动问题的这些方法的最彻底的比较。结果表明,共享节点方法是计算密集但不灵活的可变间隔网格的精度与不太准确但灵活的单向耦合方法的有效CPU时间之间的折衷。这项研究还研究了地下水反演过程中几种网格细化方法的性能。结果表明,单向耦合方法的性能较差,因为它们在两个网格尺度上都缺乏与参数敏感性相关的反馈。相反,共享节点方法中存在的反馈在所有测试情况下均产生了良好的结果。评估了局部网格细化作为解决代表地表水-地下水相互作用特征的规模问题的工具。这项研究的结果表明(1)高度精细的网格对于表示流-水层相互作用的小规模水力特征是必要的,并且(2)流-水层的连接是网格规模的依赖。共享节点方法能够以合理的准确度(0.6–3.9%的误差)表示流-水层相互作用,而单向耦合方法会导致更大的误差(6–20%)。 *本论文是多媒体的(包含文本和其他应用程序无法以打印格式显示)。随附的CD需要满足以下系统要求:Adobe Acrobat;微软办公软件。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号