Aquaveo & Water Resources Engineering News

The next version of GMS (10.1) will include the ability to export TOUGH2 MESH files. Users will be able to build complex 3D UGrids using all the tools available in GMS and use those UGrids in their TOUGH2 models. Voronoi grids, triangular grids, grids created from horizons and solids - all will be exportable. This feature is implemented in our development branch but needs more test cases. If you would like to help us test this feature, please send an email to mkennard@aquaveo.com.

The GMS team just added support for the ghost node correction (GNC) package in GMS. This will be available in GMS 10.1. A wiki page has already been created to document the interface.

The GNC package can be used to improve the accuracy of a model in which there are grid cells that do not follow the assumptions of the finite volume method - namely that the line connecting the center points of adjacent cells is also a perpendicular bisector of the face between the cells. This situation arises with quadtree, octree, and nested type grids, as well as with many other types of grids. Regular (quadrilateral) grids and voronoi grids do not have this issue. All of these types of grids can easily be created in GMS.

¹

GMS can calculate all the ghost node data automatically, or the user can enter the data manually. When done automatically, GMS uses an algorithm developed with input from Sorab Panday (author of MODFLOW-USG). The algorithm is illustrated in the following figure:

Ghost node locations are calculated using the perpendicular bisector of the face between adjacent cells and its intersection with the lines connecting cell centers. AlphaJ values are calculated based on the distances from the ghost nodes to the adjacent cells. A new tutorial was created to teach how to use the new GNC functionality in GMS. The GNC package typically increases the accuracy of a model by a small amount and is therefore often unnecessary.

---

^{1. Panday, Sorab, Langevin, C.D., Niswonger, R.G., Ibaraki, Motomu, and Hughes, J.D., (2013), MODFLOW-USG, version 1: An unstructured grid version of MODFLOW for simulating groundwater flow and tightly coupled processes using a control volume finite-difference formulation: U.S. Geological Survey Techniques and Methods, book 6, chap. A45, 66 p., Reston, Virginia↩}

Some of the more notable things completed by the GMS team during the month.

• Fixed 8 bugs
• MODFLOW-USG duplicated when duplicating UGrid
• More progress on observations with MODFLOW-USG and PEST
• Expanded/collapsed state of items in the Project Explorer now saved
• mod-PATH3DU interface mostly completed
• Progress on new datasets to be used with VTK
• User now asked about inactive cells when converting to MODFLOW-USG
• Switching to MODFLOW-NWT now prompts to switch the solver to NWT
• GUI testing for UGrid interpolation added
• MODFLOW code and solution changed to reduce all the dlls

Thanks to our friends at SRIT for organizing another outstanding training course in Australia. The four day GMS class was a great success.