Can an unconfined aquifer in a complicated topography of high and low altitude be modelled as a confined aquifer due to convergence issues?

209 views
Skip to first unread message

Syed Faisal

unread,
Oct 10, 2023, 9:42:54 AM10/10/23
to MODFLOW Users Group
Hi,
Can an unconfined aquifer in a complicated topography of high and low altitude be modelled as a confined aquifer due to convergence issues?
I tried my best to model it as an unconfined aquifer but always getting errors while running the model. Please help for a justification.

Regards
Faisal

Robert Lanning

unread,
Oct 10, 2023, 11:34:46 PM10/10/23
to MODFLOW Users Group
Hello Faisal

Without knowing more details about your model, we can't answer your question.  If the aquifer you mention is the top layer in your model, the answer is probably no - it should be unconfined, unless you have some impermeable layer near the top of that layer.  If the model won't converge, it is more likely due to your conceptual model inputs (recharge amount, layer setup, hydraulic conductivity, boundary heads for example).  Please tell us more about your model and folks may have some suggestions:
  • What MODFLOW program and what User Interface are you using?
  • What is the topography (range in altiude)? Maybe include a topographic map of your model area
  • How many layers in your model?
  • Are layers "cut-off" by the topography?
  • Is there both unconsolidated and bedrock?
  • What are the lithologies?
  • What are you using for recharge?
  • What boundary conditions are you using?
  • Hydraulic conductivities of layers?
  • Any surface water features (lakes, rivers)?
  • Any karst features, springs, or "drain" features?
  • Perhaps include a sketch cross-section of your model with some of the model attributes
Regards,
GeoBob

Jakab Andras - Gmail

unread,
Oct 10, 2023, 11:34:57 PM10/10/23
to mod...@googlegroups.com
It certainly can, but indeed it might be quite difficult. As a first approach, depending on the version of MODFLOW you are using, try using either MODFLOW-NWT as the solution engine provided that your model is a MODFLOW-2005 model, or if you have a MODFLOW 6 or MODFLOW-USG model, use the Newton option (with MODFLOW-USG it is default). These MODFLOW versions have quite similar solver settings but there are differences between them. It is worth, however, reading at least the MODFLOW-NWT manual prior to start manipulating the solver settings as they are more in number and are more complex than the standard MODFLOW-2000/2005 solver settings.

--
This group was created in 2004 by Mr. C. P. Kumar, Former Scientist 'G', National Institute of Hydrology, Roorkee. Please visit his webpage at https://www.angelfire.com/nh/cpkumar/
---
You received this message because you are subscribed to the Google Groups "MODFLOW Users Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to modflow+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/modflow/dded57b5-8cd2-4752-8a5e-a7882d809f5dn%40googlegroups.com.

An Ho Antonio Taylor

unread,
Oct 10, 2023, 11:35:05 PM10/10/23
to mod...@googlegroups.com
Have you tried to use another solver like Newton NWT package? I imagine that the confined aquifer assumption keeps wetted the cell even though it is supposed that the head has the possibility to surpass below that threshold, giving you a wrong Head value, have you tried changing the Solver criteria vs volumetric Balance?

BR
An Ho

--

ashutosh singh

unread,
Oct 10, 2023, 11:35:16 PM10/10/23
to mod...@googlegroups.com
Hi,

My experience is that complex topography requires a correct layerwise geology, which means a good three dimensional model with good knowledge of hydraulic conductivities.
Another thing is to use Modflow-NWT for these simulations where one can use convertible layers.

//Ashutosh Singh

--

Syed

unread,
Oct 13, 2023, 10:20:39 AM10/13/23
to mod...@googlegroups.com
Dear Jakab,

Thanks for your response. Actually, I have tried all of the above options and none of them is working for convergence of the unconfined model. The only way the model is converging with good calibration is through the confined aquifer with MODFLOW-2005. 
Best regards
Syed

Syed

unread,
Oct 13, 2023, 10:20:51 AM10/13/23
to mod...@googlegroups.com
Hi Ashutosh,

I think you got it right. This area with complex topography is having less knowledge about geology, lithology and conductivity.

Regards
Syed

Jakab Andras - Gmail

unread,
Oct 13, 2023, 11:54:01 PM10/13/23
to mod...@googlegroups.com
As others pointed out, without knowing more about your model it is hard to provide more help. Solvers of MODFLOW-USG/-NWT/-6 should normally have a setting that leads to convergence, but as I said, there are lots of possibilities, and one should be familiar with their meanings. Sorab Panday the author of MODFLOW-USG and also co-author of MODFLOW 6 has a couple of webinars in which he discusses these settings. You may try to contact GSI to get a copy for a fee, but I'm not sure if it's available.

If you are absolutely sure you tried all possibilities, then my recommendation is to simplify the conceptual model. Often, the model includes much more complexity than it is required, so you might be able to redefine your goals or simplify your conceptualization.

Francisca Clemo Garcia

unread,
Nov 29, 2023, 11:22:22 PM11/29/23
to MODFLOW Users Group
Hi everyone, I'm joining this conversation because I'm having the same issue. Have an area with 2 narrow gorges and another gorge a little bit less narrow. These are all separated by hills, so topography and geometry are complex. I have 3 layers: 2 layers represent units of unconsolidated materials (the upper one is only saturated in places near an infiltration zone, included as recharge), and the 3rd layer is fractured rock. I exported the grid from Leapfrog with a minimum thickness of 10 m per layer. Right now, my boundary conditions are 2 GHB at the entrance of 2 gorges (level taken from less than more stable boreholes), 1 with an injection well (I don't have any wells there, so I don't know the water level) and as an exit, I have a drain with a high conductance (just to try to let the model take out all the water that is needed). The model converges only if every layer is confined (0), very rapidly, but when I change the conditions to unconfined to layer 1, and unconfined with T varies to layers 2 and 3, the model doesn´t converge and has high errors.
I'm using GWV 8, and solver XMD, and also tried with 2 PCGU.
I would really appreciate some of your ideas.

Raghwendra pratap

unread,
Nov 30, 2023, 12:24:35 AM11/30/23
to mod...@googlegroups.com
Hi Francisca,
can you give a little detail about how you calculated the conductance for GHB and Drain package and is it in direct form or calculated?

--
This group was created in 2004 by Mr. C. P. Kumar, Former Scientist 'G', National Institute of Hydrology, Roorkee. Please visit his webpage at https://www.angelfire.com/nh/cpkumar/
---
You received this message because you are subscribed to the Google Groups "MODFLOW Users Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to modflow+u...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages