U.S. Geological Survey Upper Midwest Water Science Center, 8505 Research Way, Middleton, WI, 53562, USA.
U.S. Geological Survey Upper Midwest Water Science Center, Michigan, 5840 Enterprise Drive, Lansing, MI, 48911, USA.
Ground Water. 2021 Sep;59(5):761-771. doi: 10.1111/gwat.13095. Epub 2021 Apr 8.
Groundwater models have evolved to encompass more aspects of the water cycle, but the incorporation of realistic boundary conditions representing surface water remains time-consuming and error-prone. We present two Python packages that robustly automate this process using readily available hydrography data as the primary input. SFRmaker creates input for the MODFLOW SFR package, while Linesink-maker creates linesink string input for the GFLOW analytic element program. These programs can reduce weeks or even months of manual effort to a few minutes of execution time, and carry the added advantages of reduced potential for error, improved reproducibility and facilitation of step-wise modeling through reduced dependency on a particular conceptual model or discretization. Two real-world examples at the county to multi-state scales are presented.
地下水模型已经发展到包含更多水循环方面的内容,但将代表地表水的实际边界条件纳入模型仍然很耗时且容易出错。我们提供了两个 Python 包,它们使用现成的水文数据作为主要输入,可靠地实现了这一过程的自动化。SFRmaker 为 MODFLOW SFR 包创建输入,而 Linesink-maker 为 GFLOW 分析元素程序创建线汇字符串输入。这些程序可以将数周甚至数月的人工工作减少到几分钟的执行时间,并且具有减少错误的潜力、提高可重复性以及通过减少对特定概念模型或离散化的依赖来促进逐步建模等优点。本文提供了两个县到多州规模的实际案例。