Jump to content
GMS, SMS, and WMS User Forum

Hydrology

Members
  • Posts

    157
  • Joined

  • Last visited

Everything posted by Hydrology

  1. I am getting write errors whenever I try to save or run my MODFLOW model. It was working just fine, and then started giving me the errors that are shown in the attached screen shot. I have tried closing out GMS and restarting but the errors still occur. I have even rebooted my computer, but the errors still occur. The model still seems to run, but I am afraid that these errors are going to come back and bight me in the butt in the future. Any help would be appreciated. Thanks
  2. I have been trying to create a groundwater contour animation in GMS. It seems to work well for short duration timeseries of 300 stress periods, but when I try to create an animation of 3,000 stress periods, the animation seems to be created properly in GMS, but when I play the video, it is all scrambled and garbled. Is there a limit to the number of stress periods that you can use to create an animation?
  3. Anybody else have this problem? After a month or 2 the grids started displaying again, but now Feb 2021, they have stopped displaying again, even for older models that worked fine a few years ago.
  4. I tried the PCGN solver and it was awesome to see it use 90% of my 24 cores. But, strangely the PCGN took 17 hours to run using all 24 cores, where the PCG solver only took 30 minutes to run basically using one core.
  5. Hi thanks for responding. I have the 64bit option and parallel option checked. I am using the PCG package. Has the PCG package not been parallized?
  6. I don't know if this should be under the MODFLOW directory, but I have been trying to see if I can speed up my GMS runs. I have a GMS MODFLOW model that is taking about an hour to run and I would like to shorten that run time. I see that I am using about 2.5% of my CPU utilization and I am running about 300 kB/s in disk access time, well below the capacity of the CPU or disk. I have a dual processore Xeon system with 12 cores for each processor running at 3.2GHz. My disk drive is 4-2TB M.2 NVME SSD's running in a Raid 0 array in a PCIe slot. The system has 256GB of RAM. This configuration gives me about an hour run time for the model. I tried loading the whole model into a 90Gig RAM Disk on the same machine, and the run time was basically the same. As a comparison, I ran the same model on my laptop that has 16GB of RAM and a single 1TB NVME drive and a single i7 4 core processor. The model actually ran about 10% faster on my laptop than on my workstation. What is the slow link in the model execution? It doesn't seem to be CPU or disk access. Is there anything that I could do (hardware wise) to speed the model run up?
  7. This is a little late for adding to this topic, but your path name is really long. Try shortening up your directory path name and see if that helps.
  8. I am setting a drain with a specific elevation in a MODFLOW run, but the final gw elevation for the cell that the drain is located in, is not the same as the elevation that I set for the drain. It seems to vary depending on the hydraulic conductivity that I set for the layer that the drain is in. For low layer conductivities, the final drain elevation in the cell is close to the elevation that i specified for the drain, but for higher layer conductivities, the gw elevation in the cell can be several feet higher than the specified drain elevation. How can I get the drain to force the elevation in the cell to what I specify?
  9. Wrong forum. Move your question to the WMS forum.
  10. I just recently started having a problem displaying the grid. I create the grid and it is visible, but as soon as I map it to Modflow, the grid within the model boundary dissapears. You can still see the grid outside the model boundary until you activate cells in coverage, then those grid lines as they should. I have played around with the display options in 3D Grid Data, but the cell edges will not display. I have this same problem on about 7 different GMS models. It seemed to just start about a week or two ago. Does anybody else have this problem, or know of a fix. I just figured out how to click on the zoom and pan button to turn on the contour label formatting, so I was hoping that there would be something similar for this.
  11. I think that the problem ended up being a conflict between GMS and our firewall "Comodo". GMS creates 5 or 6 batch files in the Temp directory every time you zoom or pan the display. It then executes the batch files sequentially to redraw the bitmap. Our firewall would send up a popup every time GMS executed one of those batch files because batch files can be a big security risk. These popups would often occur 300 to 500 times during a single GMS session. I tried turning the firewall off to avoid the warnings, and that is what caused GMS to stop displaying bitmapped graphics completely. But, I found that by uninstalling the firewall completely from the workstation, GMS started displaying bitmapped graphics again. Turning the firewall off was not enough, you need to uninstall it completely from your computer. So if you are having trouble displaying bitmapped graphics, you need to uninstall your firewall. Not a great thing to do, but you don't have much of a choice with the weird way bitmaps are redrawn in GMS. Hopefully this is the fix and GMS will continue to work. I will repost if the problem comes back.
  12. The files are local, but are on a different drive than the model files. I will try moving them to the GMS model directory and see if that helps.
  13. I have been working with a GMS model with two TIFF files for background information. After several days of working correctly, the model stopped displaying the imagery. It always took a long time to display them, but after 30 sec to a minute they would display. I deleted the images and reopened them but GMS still does not diaplay them. I tried closing GMS and reopening it, but GMS will still not display them. Does anybody know what is causing this, or have a workaround?
  14. Here is the screen capture of the model crash.
  15. I have a weird problem. I have a transient model that runs fine if I set the Output Control to "Output at last time step of each stress period". But when I set Output Control to "Output At Every Time Step", the program immediately crashes when I try to run it. Has anybody run into this problem before? I have tried to attach a screen capture of the crash error, but the forum wont let me even though it is only 112KB. The MODFLOW output file just contains the first the few lines of the run header.
  16. Thanks Michal, I will experiment with that and see if that approach gives me what I need.
  17. Hi Bri, thanks for the response. I am familiar with the references that you indicated. I often use the specified head and head dependent flux boundaries. I havent used the specified flux boundary, but that is not what I am looking for. What I am trying to set up is a boundary condition that will allow free flow across the boundary without restriction, so that the flow across the boundary will vary in time, depending on the amount of water that i pump into or out of the model domain. I don't want to necessarily fix the head across the boundary either because I think that would overly constrain the model. Not applying a boundary condition to that part of the model will create a no-flow boundary condition, and that is what i am trying to avoid. Is there a way to achieve the condition that I am looking for?
  18. Is the default boundary condition in MODFLOW a no flow boundary? If so how would I create a boundary condition that allows flow, but has a fixed head at either end of the boundary?
  19. I thought that might be the case, thanks.
  20. Does a horizontal flow barrier block the entire layer based on what layers your barrier coverage is assigned to, or can you set the bottom elevation of the barrier by clicking on the attributes in map view and setting the bottom elevation in the Z column. I have not seen much on line that has given me a definitive answer. All of the descriptive drawings always show the whole layer being blocked by the barrier, but that may be just for simplicity.
  21. I will try to see if they can do this. Since this type of program behavior fits a classic malware attack profile, anyone running GMS has been banished to a trailer behind the office where there is no internet connection or connection to the server. So now we have no direct access to the project files on the server, and we are not allowed to put anything developed on this computer back on the server. We have never had this problem with previous versions of GMS, so I will see if we can go back to an older version of GMS. Below is a typical script from the batch files. REM -- Merge images including a created background color image. REM -- Created a 1 pixel image with display extents. "c:\software\hydrologic\gms 10.4 64-bit\gdalbin\gdalwarp.exe" --config GDAL_DATA "c:\software\hydrologic\gms 10.4 64-bit\gdalbin" -of GTiff -dstalpha -te 6005738.0000000000000 2110135.0000000000000 6021445.0000000000000 2120129.0000000000000 -ts 1 1 "R:\Temp\GMS_19140\im00005_action1out.tif" -overwrite "R:\Temp\GMS_19140\im00031_image2out_image1merge1onePixel.tif" REM -- Give the 1 pixel image the background color. "c:\software\hydrologic\gms 10.4 64-bit\gdalbin\gdaldem.exe" --config GDAL_DATA "c:\software\hydrologic\gms 10.4 64-bit\gdalbin" color-relief "R:\Temp\GMS_19140\im00031_image2out_image1merge1onePixel.tif" "R:\Temp\GMS_19140\im00031_image2out.tif_image1merge2map.txt" "R:\Temp\GMS_19140\im00031_image2out_image1merge3Background.tif" REM -- Combine 1 or more images using current display view. "c:\software\hydrologic\gms 10.4 64-bit\gdalbin\gdalwarp.exe" --config GDAL_DATA "c:\software\hydrologic\gms 10.4 64-bit\gdalbin" -of GTiff -te 6005738.0000000000000 2110135.0000000000000 6021445.0000000000000 2120129.0000000000000 -ts 1614 1027 "R:\Temp\GMS_19140\im00031_image2out_image1merge3Background.tif" "R:\Temp\GMS_19140\im00005_action1out.tif" -overwrite "R:\Temp\GMS_19140\im00031_image2out.tif"
  22. I have just upgraded to GMS 10.4 and I am having a compatibility issue with GMS and our Network security. Apparently with 10.4, whenever a graphics file, like an aerial photograph, is displayed, the program creates a temporary batch file that it executes and then erases. This happens for each aerial file you have in your project, and the batch files are created each time you pan or zoom your display. These batch files are considered security risks by our security software, and I get a pop-up asking me to allow/not allow the batch file to run each time I move my display. Each batch file has a different random name for each time the display is redrawn, so we cannot add an exception to our security software for the batch file. A typical name is C\Temp\GMS_21380\img00247_image2.out.bat. I have 9 aerial files in my project so I have to click OK on the security pop ups 100 or more times during a working session with GMS. Is there any way to have GMS not produce these batch files.
  23. Are the pumping rates interpolated between time periods? I have pump totalizer data on irregular time steps and I am trying to make sure that I am representing them correctly in the pump time series. In the example below I show the date and the totalizer readings in (gals). I convert the total gals to an average gpm computed over the time period since the last reading, then to a ft3/day for input into modflow. Does modflow interpolate the flow between time steps, or does it use a specified flow rate for each time step until another one is specified (stepped pump rate). For example is the pump rate on 3/13/12 in the example below 0.00 ft3/day, or is it 77.33 ft3/day? I want to make sure that I get the mass balance for pumped volume correct. Date gals) (gpm) (ft3/day) 3/12/12 0.0 0.0 0.0 3/15/12 5,212.0 1.2 -232.3 3/16/12 8,674.0 2.4 -462.8 3/19/12 9,752.0 0.2 -48.0
  24. Is there a way that I can rotate the display so that north is not straight up? I have an existing model and I would like to rotate it around 45 degrees before I make an animation of groundwater drawdown due to pumping.
×
×
  • Create New...