Fix inertial gravity wave test

[xylar]

The viz step didn't get updated when we added refinement in just space and just time separately, so that is the main fix here. The viz step is now more similar to the manufactured solution step, which was fixed in #250.

This merge also sets the namelist option config_disable_tr_all_tend = .true., which was correctly dropped from the single_layer.yaml in #262 but which now is correctly included in this test's forward.yaml.

Checklist

• Testing comment in the PR documents testing used to verify the changes

[xylar]

Testing

I ran all three inertial gravity wave tests on Chrysalis, including viz. The results look as expected. We're not seeing any convergence in time, so that analysis check fails (presumably as expected).

Both:

Space:

Time:

[xylar]

Here are the analysis plots, for completeness:

Both:

Space:

Time:

[xylar]

@cbegeman, could you give this a quick look when you have time? I don't think you have to spend too much time looking over the code (and you don't need to do any testing) as long as you're happy with the results I posted above.

[sbrus89]

It seems odd that we aren't seeing convergence in time. Is there any reason to expect that?

[xylar]

I think it takes a pretty specific setup for convergence not to be defined more by errors in space or in time. So I'm not that surprised.

[cbegeman]

@xylar Thanks for catching this! Approving based on visual inspection.

[xylar]

Thanks, @cbegeman!