mperego

@rppawlo I tried to replace C2 with Cn in [tFieldAggPattern.cpp ](https://github.com/trilinos/Trilinos/blob/master/packages/panzer/dof-mgr/test/dofmngr_test/tFieldAggPattern.cpp) but I didn't understand it well and didn't want to change the hardcoded values. Is the `GeometricAggFieldPattern` supposed to...

right!

thanks @alanw0 and @mcarlson801!

I'm fine with removing them. At some point we'll implement a low-order model, but it won't be the l1l2 and it will be likely based on StokesFO problem that now...

L1L2 is gone. Not sure about full Stokes. There is no plan at the moment to use it, but things might change. We could add an option: -D ENABLE_LANDICE_STOKES which...

Yes, that was also what I had in mind.

Assuming you are correct ;-), I agree with that.

@jewatkins I'm not convinced it has to with tolerances, but it could if for some reason this peoblem is very ill-posed. I tried running this w/ gcc and I get...

The Humboldt results look reasonable to me.

OK, so we don't get significant changes on the Humbolt mesh. @jewatkins what solver are you using on the Greenland mesh? Belos or AztecOO? Does the choice of the solver...