Jump to content
GMS, SMS, and WMS User Forum
piccolr

Transient TINs

Recommended Posts

Dear all,

I'm currently dealing with a MODFLOW simulation where the head of an unconfined aquifer changes in time. The assumption is that for a given time and in correspondence of the channel network, the head of the aquifer stays dz meters below the surface. In order to impose suitable boundary conditions, I created a TIN based on the GMS grid (i.e. the two coincide) and I let the TIN have different time step (in order to do this I first exported the TIN, modified it adding nt columns that account for the distance dz of the aquifer from the surface, where nt is the number of time step considered, and successively imported it in GMS).

The information contained in the time-varying TIN are imported in the conceptual model: 1) I created a coverage for the channel network and I assigned to channels' arcs the CHD property, 2) I selected all the nodes of the channel network and I associated them the TIN.

Finally I exported the conceptual model to my MODFLOW project using Map-->MODFLOW. In doing this, I kept "Constant value throughout data set time step (step function)" (I had a look at this page http://xmswiki.com/xms/GMS:Map_to_MODFLOW_with_TINs)

Now, the problem is that the boundary condition in correspondence to the channel network is not time-varying, but rather the initial value is maintained throughout the whole simulation. Can you help me in finding out the problem?

Thanks

Share this post


Link to post
Share on other sites

What are the time values for the different data set time steps on your TIN? How do those time values compare to the time values assigned to the stress periods?

Share this post


Link to post
Share on other sites

Alan, thanks for the reply.

I exported the TIN as a txt file and modified it off-line adding one column for each time step. The time step correspond to the stress periods (i.e. both stress periods and time steps of the TIN data set have a 1 day frequency). TheTIN data set has this structure:

X Y head|1 head|2 head|3 .... head|nt

.. .. .. .. .. .... ..

where head|i indicates the aquifer head at the i-th time step.

This file is imported as 2D scattered data and interpolated to the existing TIN. Finally, the TIN is used for imposing the BC along the arc representing the channel network in the conceptual model.

Actually, I'm now trying a different solution, which allows me to avoid a couple of steps, that is to directly interpolate the 2D scattered time dependent dataset to MODFLOW, enabling the option "CHD BCs" and "Use all time steps" in the interpolation menu. What do you think about this option?

However, still I'm curious about using transient TINs.

Thanks

Edited by piccolr

Share this post


Link to post
Share on other sites

Either option will work. With both options you will need to assign the actual time to the data set time steps. So if your MODFLOW stress periods begin at 0.0, 1.0, 2.0 then what you have described should work. However, if your MODFLOW stress periods begin at 0.0, 30.0, 45.0, 60.0 (or something like that) then the time assigned to your data set time steps should overlap with the MODFLOW stress periods.

Let me know if this is not clear.

Share this post


Link to post
Share on other sites

Yes, the MODFLOW stress periods begin at 0.0, 1.0 etc. By interpolating the 2D scattered data, things seem to work correctly!

Concerning TINs, the problem is that when I associate the current TIN to the nodes of the arcs in the conceptual model, GMS associates only the TIN corresponding to the last time step selected in the "time step window" (we could say: the current time step of the current TIN), whereas I would like all time steps were associated.

Edited by piccolr

Share this post


Link to post
Share on other sites

Glad to hear that using the interpolation worked for you.

Regarding the TIN with Map->MODFLOW...

If the active data set on the TIN is transient then it should use all of the time steps when doing Map-> MODFLOW. We have a few simulations that we use to test this feature to make sure it is working but maybe we have missed something. Could we look at your simulation and see if there is a bug?

Share this post


Link to post
Share on other sites

If you will email support@aquaveo.com and tell them that you have a model to upload for the GMS developers to look at then they will give you the instructions. Thanks for your help.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...