Posted Mon, 26 Nov 2012 11:02:26 GMT by abaietto
Hello,

I would like to ask if someone has any clue of how to solve this problem:

I'm on a 3D model of an unconfined aquifer with an irregularly steeping base surface: upstream the the aquifer is very thin (4 m) while downstream it becomes thick (30 m).
Along the model, the groundwater level is supposed to follow the base of the acquifer standing just few metres above it (the substratum has a very low K). Hydraulic Head BC are imposed on the two sides of the model.

I'm trying to calibrate the groundwater level in several points but I can't found a way to keep it constantly above the aquifer base level (attached figure), even just few mm above. Obviously this is not physically correct....

I'm using this scheme:

top of the model                  Slice 1: FIXED
slice of the aquifer              Slice 2: UNSPECIFIED
slice of the aquifer              Slice 3: UNSPECIFIED
slice of the aquifer              Slice 4: UNSPECIFIED
slice of the aquifer              Slice 5: PHREATIC (where part of the water table is expected to be in the deeper part of the aquifer)
slice of the aquifer              Slice 6: PHREATIC (where part of the water table is expected to be in the deeper part of the aquifer)
Aquifer base                        Slice 7: FIXED
Base of the impervious layer Slice 8: FIXED

with the constraint of FALLING DRY AT BOTTOM as watertable (residual water depth: considered several values from 0.001 to 1)

Thank you for any help,

Alessandro
Posted Wed, 28 Nov 2012 17:58:21 GMT by psinton@aquageo.us
phreatic mode does not work well where the water table crosses slices...use richard's equation instead and review white papers
Posted Thu, 29 Nov 2012 06:10:15 GMT by Blair Thornburrow Groundwater Modeller
Or you could use free and movable surface on the top slice and undefined beneath down to the top of the aquitard.

However, there is a classic problem with simulating "thin" (low saturated thickness) flow in an aquifer above an aquitard.  This is very difficult to solve numerically because of the non-linear behaviour associated with alternating wetting and drying.  This requires very small time steps.

You must be signed in to post in this forum.