Jump to content
GMS, SMS, and WMS User Forum

Saubhagya Singh

  • Posts

  • Joined

  • Last visited

Everything posted by Saubhagya Singh

  1. Dear Forum Members, I am trying to perform Monte-Carlo simulation in a vertical slice of the aquifer (x-z plane). I want to generate around 100 realizations of the spatially-correlated hydraulic-conductivity random field in 2D (x-z plane). Then I want to solve flow and transport for each realization and analyze the ensemble of results. I explored options in GMS and this is what I found out (correct me if I am wrong): 1) Stochastic Modeling Option: It is not possible to generate 2D field (spatially varying K values) using this option. 2) Interpolation using 2D Scatter Point Tool: This only allows to generate conditional realizations (kriging). I couldn't see a way to generate unconditional realizations. Also, this tool works only in plane view, hence random field can be generated only in the x-y plane. I want to generate fields in x-z plane. Hence, it seems my only option is to generate K fields outside of GMS, importing K fields and solving cases one by one. Please, let me know if there is any other efficient way to perform this. Thank You Saubhagya
  2. Good Morning Forum Members, Is it possible to import a 2D (x-z plane) K field for a vertical slice of a heterogeneous aquifer? Or my only option is to paste layer-by-layer values in the LPF package? Thank You Saubhagya
  3. Ah! Now "Rewetted" makes sense to me. Don't know how I missed that haha! Thank you so much!
  4. The figure in my previous comment shows the result when no rewetting was required as initial heads were equal to the higher boundary-head. But in the case where I have initial heads equal to lower boundary-head, the cells near the higher-head boundary never get wetted. I will try some more combinations of wetting parameters.
  5. Hi Michael, Thank you for your answer. Yes, I also think that. I did not have rewetting on. However, I tried various combinations of rewetting parameters but nothing helped. Finally, I had initial heads set to higher boundary-head, which then did not require rewetting and it worked (Figure attached). However, do you have any suggestion how can I have rewetting enabled? I used the wetting parameter as low as 0.01 still rewetting did not happen.
  6. Here is an example of a single-density flow width domain length as 53 cm, hydraulic conductivity as 20 cm/min and boundary heads as 21 cm and 22 cm. Theoretical calculation based on Darcy's law yields 8.11 cm^2/min, however, the flow budget in the model described below shows 7.9186 cm^2/min as the flow near the inland boundary which is exactly the value we would obtain if we consider a confined aquifer of thickness 21 cm and other parameters same as above. This suggests that flow budget does not account for the flow above the elevation equal to the lower boundary head. Also, there are not velocity vector generated in that region. Any guidance on how to proceed to get the accurate flow budget for the unconfined aquifers will be helpful.
  7. Dear community members, I am trying to numerically evaluate the discharge in an unconfined coastal aquifer with a constant head as the inland boundary. After performing SEAWAT simulation, the flow rate value in the flow-budget for any cross-section close to the inland boundary reflects only the flow below the sea-level. I also confirmed this by plotting velocity-vector plot, in which there were no velocities plotted in the region above the sea-level and below the water table. I have attached the picture for clarity. Is there a way I can get the total value of the freshwater discharge in the aquifer? I believe, a simple single-density flow model would also have a similar issue with the flow-budget in the unconfined case.
  8. Thank you Bill for the reply. Is there any workaround to this for now? Like, will splitting that one row/column into two of half of the original widths work? Or, are any of the older version free from this bug? Regards Saubhagya
  9. Hello respected members, The problem mentioned by me in the question above is still unresolved. I tried running MODFLOW model on both 2000 and 2005 version. The problem can also be seen in the example files provided in the Goswami/Clement Experiment** (http://www.aquaveo.com/software/gms-learning-tutorials). I tried changing few things but nothing helped. ANy help in this regard would be really appreciated. Thanks and Regards Saubhagya
  10. Dear members, I am trying to generate velocity vectors in a density dependent model. The GMS is displaying error, pic of which is attached below. I have been generating velocity vectors and never faced any such problem. I am not able to work out solution to this problem. Any help would be appreciated. Thank You
  11. Hi, It might be very basic question but I am struggling with it. Is there a feature in GMS to display streamlines after MODFLOW simulation? Thank You Saubhagya
  12. or I think, even exporting a particular contour line should also work for me. Is it possible?
  13. Good Morning, I tried doing this but couldn't find the right file to do this? Basically the task for me would be to find the location of cell with a specific concentration in the lower most layer of the aquifer, at each timestep. So the file having data of concentration in each cells at each timestep should work for me. Thank You Saubhagya
  14. Yes, that should work. Thank you very much Alan for the suggestion. Regards Saubhagya
  15. Hi Alan, Toe position would be the x-coordinate of point where a particular concentration contour (iso-chlor) would meet the base of the aquifer. Thanks
  16. Hi, I was wondering if it is possible to export toe position for each time step, for a particular isochlor while modelling seawater intrusion using SEAWAT? Thanks Saubhagya
  17. Hi, I am trying to simulate a SWI in coastal aquifer (transient). The aquifer is 1500 m long and 40 meter deep, and is discretized as 1500 cells by 80 cells. I am running the model for 50,000 days and apparently the model would take 9 hours. I have attached my time-discretion details. Am I doing something in an inefficient way? Would putting some different values to multiplies would help? Any help to reduce run time would be appreciated. Thanks
  18. Hi, I am trying to run a simulation for a very long duration (more that 80,000 days). Last few MT3D simulation crashed probably because the folder containing MT3D files exceeded 10 GB in size. This brought few queries in my mind. 1) Are such a huge MT3D file sizes normal? 2) What settings/criterion in GMS controls the file size? Is it time step size or cell sizes? Does Max number of inner and outer iterations affect the sizes of files? 3) To start with, what is the best way to estimate the optimal grid spacing and time step size? If model is huge and takes time to set up, trying different grid spacing can be tedious. These questions are bothering me since a long time. Answers would really be appreciated. Thanks and Regards Saubhagya
  19. Also, I have attached picture showing settings of my GCG solver
  20. Thanks Alan for the help. I checked output file of MT3D as suggested by you but couldn't find any clue at the bottom of the file. I have provided link to the file here below (upload restriction due to size limit). If possible, please have a look and guide me how to debug my model. https://drive.google.com/file/d/0Bw8D5CrEtgA-WGdvNmpkcll6TVE/view?usp=sharing Thank You Sincerely Saubhagya
  21. Hi, I am trying to simulate a scenario where the boundary head is gradually varying with time. Another scenario I am trying is to simulate coastal aquifer with seaside boundary head varying as per tidal variations of sea level. As far as I know, in given stress period the specified head remains constant. Therefore continuous head variation cannot be achieved and it has to be approximated by series of step changes. This will result in huge number of stress periods. Is there any alternate more efficient way to simulate such scenario? Thanks in advance. Saubhagya
  • Create New...