Investigate AMIP paperplots change

[juliasloan25]

After the update to ClimaAtmos v0.11.0 in PR #293, the amip_paperplots.png plots generated by AMIP and AMIP modular look slightly different (see before and after).

The ncep_paperplots.png for the same runs look identical despite the ClimaAtmos change (see before and after). All other moist_mpi_earth, seabreeze, and slabplanet tests also produce the same output as was previously seen.

It's possible that the change in the AMIP paper plots is due to incorrect radiative fluxes. We need to investigate this further and correct the physics to match the original results.

Paperplots before and after (the above links appear to be broken)

[valeriabarra]

After running some investigation, here are the shortwave and longwave (respectively) face radiation flux plots:

[valeriabarra]

We also checked that the regridding before and after the PR is unchanged:

And that the tot energy is conserved before and after the PR:

So the issue seems to be stemming not from the Coupler, but from some internal ClimaAtmos changes that made some configurations unstable. We need to find a new set of parameters in the Coupler to make them stable (probably changing dt or try clearsky radiation).

[LenkaNovak]

this was resolved here