SEI+II (GEO): An IT-Enabled, Hierarchical and Patch Dynamic Environment for Modeling Complex, 3D Groundwater Systems Across Multiple Spatial Scales

尽管近年来计算能力急剧增长，但由于许多棘手的计算和概念挑战，我们对复杂的3D地下水系统进行建模的能力仍然受到严重限制，包括：机器瓶颈；算法瓶颈；还有数据同化问题。本研究采用跨学科方法系统地解决了大规模地下水建模中的这些基本困难。该研究特别利用了层次理论，开发了一个多层次和斑块动态的多尺度复杂地下水系统建模框架；一个新的建模范式解决了一个重要的计算机科学问题——分层建模的实际实现的障碍。具体来说，新的、it支持的、分层的补丁动态框架将：(a)允许建模复杂的地下水系统，而不必解决大型的、病态的矩阵系统，从而消除或大大减轻臭名昭著的“维度诅咒”和相关的三维地下水建模的计算瓶颈；(b)提供一个系统的“尺度阶梯”，将多个尺度的数据和模型联系起来，并吸收来自不同来源的信息；(c)提供了分层计算和可视化的动态集成，使用户不必与每个子尺度建模补丁进行交互，并消除了相关的人力瓶颈。提出的分层建模环境可能会潜在地改变地下水水文学计算建模的方式，并大大提高我们理解地下系统、过程和规模相互作用的能力，并开发基于模型的综合水资源管理和环境清理工具。

Despite the recent dramatic growth of computational capability, our ability to model complex, 3D groundwater systems is still severely limited because of a number of tough computational and conceptual challenges including: the machine bottleneck; the algorithmic bottleneck; and the data assimilation problem. This research addresses systematically these fundamental difficulties in large-scale groundwater modeling using an interdisciplinary approach. In particular, the research takes advantage of hierarchy theory and develops: a hierarchical and patch dynamic framework for modeling complex groundwater systems across multiple scales; and a new modeling paradigm that resolves a significant computer science problem - a barrier to the practical implementation of hierarchical modeling. Specifically, the new, IT-enabled, and hierarchical patch dynamic framework will: (a) allow modeling complex groundwater systems without having to solve large, ill-conditioned matrix systems and thus eliminate or substantially alleviate the infamous "curse of dimensionality" and the associated computational bottlenecks in 3D groundwater modeling; (b) provide a systematic "scaling ladder" to link data and models across multiple scales and assimilate information from disparate sources; and (c) provide on the fly integration of hierarchical computations and visualizations, free users from having to interact with each subscale modeling patches, and eliminate the associated human bottlenecks. The proposed hierarchical modeling environment may potentially change the way computational modeling is done in groundwater hydrology and substantially improve our ability to understand subsurface systems, processes, and scale interactions and to develop model-based tools for integrated water resources management and environmental cleanup.