Wednesday, November 28, 2007

GV 5.09 Build 9 through 11

Sorry, got behind over Thanksgiving...

Here are a few minor changes for Version 5.09 Build 9 through 11:

- When using Pest with the BCF Package and estimating K in zones for confined layers, there was a problem if not all zones were being estimated. T values were written in that case for the non-estimated zones.

- The ETS Package was not being imported if it showed up in a MODFLOW2000 name file.

- When importing a MODFLOW2000 model with the Well Package that used parameters to define wells, there was a problem if the order of parameter names was not the same as the way they were defined in the start of the file.

- Polygon shapefiles were being recognized by ArcGIS as having errors. Other GIS software did not have a problem. Turns out the vertices were being written counter-clockwise instead of clockwise.

Wednesday, November 14, 2007

GV 5.09 Build 8

Several minor modifications for today...

- If you write a run log after doing a "Save As" and changing the name of the Vistas file, the previous Vistas file name was written to the log file.

- When using MODFLOW-SURFACT version 3 with transport, Vistas was not writing the *.con file to the name file.

- The DXF export of particle traces never saved the traces in different colors for different layers if that option was used to display particles within Vistas. Added that to this version.

Monday, November 12, 2007

GV 5.09 Build 7

We found a bug where the CHD Package was not written correctly in the last stress period if you did not define an extra head for the very last time step (NPER+1) and you did not check the option to make CHD heads constant within each stress period. Because CHD linearly interpolates heads from the beginning of a stress period to the end, you need one extra head value. Anyway, we changed it so that if you do not had that extra head value, then the heads in the last stress period will not be linearly interpolated.

Also, GV does error checking on pilot point location for PEST. Pilot points of the same type cannot be at exactly the same location. GV was not distinguishing between pilot point types though (Kx, Kz, Storage) so if you did have Kx and Kz at the same location then there would be an error in the PEST error/warning file. This did not effect the run.

We discovered that MODFLOW2005 does NOT support the DDREFERENCE option in the output control file (word format only) like MODFLOW2000 does. This is a very handy feature that allows drawdowns after this flag to be computed based on the results of the time step where the flag is entered. We have asked the USGS to add that flag to MODFLOW2005.

Friday, November 09, 2007

MODFLOW2005 V1.4

MODFLOW2005 Version 1.4 was released on November 2nd. Groundwater Vistas Version 5.09 Build 6 has this new version. Not much really changed but here is what changed:

- Adaptive damping added to the GMG solver (Note that this was added to MODFLOW2000 in Version 1.17.00)
- A new package called SUB-WT (aquifer subsidence in water table aquifers) was added. Groundwater Vistas supports the SUB and IBS compaction packages but not this new one yet.
- A couple of minor fixes in the LAK3 Package

There is still a bug in MODFLOW2005 in the UZF (Unsaturated Zone Flow) Package where the ET rates are not properly exported to the cell-by-cell flow file. Dave Prudic told me that this is fixed but it has not been released yet.

Thursday, November 08, 2007

Better HUF Support

GV 5.09 Build 6

HUF (Hydrogeologic Unit Flow) was a new package added to MODFLOW2000 and MODFLOW2005. It allows you to define hydrostratigraphic layers that are independent of actual model layers. MODFLOW then computes a composite transmissivity based on the HUF layers that intersect each model layer. Not a bad idea.

GV5 supports HUF but the trick is how to get the HUF data into the GUI since GV really only tracks actual model layers. Here are the basic steps you need to follow to use HUF in Groundwater Vistas:

(1) Use Model/MODFLOW2000/Packages and enter a non-zero number next to HUF and check the 2nd column to create it. Put a zero next to LPF to turn that one off and uncheck 2nd column.
(2) select Model/MODFLOW2000/Options - HUF tab. Enter the number of HUF layers you want to use.
(3) add property matrices. This is done using Props/Surface Data. Each "surface" is a 2D matrix in row-column order just like MODFLOW reads it. You will need a matrix of top elevation, thickness, Kx and Kz at a minimum for each HUF layer. You add them by first creating a new one and then using either import or one of the edit commands to fill it.
(4) after adding all of the matrices you need, go to Model/MODFLOW2000/Huf data and select the surfaces for each huf property you need for each huf layer.

So that gets a bit tedious defining those surfaces or arrays. In order to make it easier to use HUF, we added a new export/import. The idea is that you create two GV models. One has the definition of all HUF layers (you don't need boundary conditions or other stuff) and the other has the number of model layers you actually want to use when running MODFLOW. Each must have the same number of rows and columns. You define the HUF model first and then use Props/Export/HUF Data Arrays. Then go into the final model and use Model/MODFLOW2000/HUF Data Import.

New Approach to My Blog

Well, my first attempt at blogging did not go very far. I guess I am too focused on our software and my consulting and was not able to set aside time for blogging. So now I'm going to use this blog to write about Groundwater Vistas and MODFLOW in general. If you want to keep up with what changes in GV and why, here is where you will find it.