Fix up ADAS CI

[mathomp4]

This PR attempts to fix the ADAS CI. The changes are:

1. Update to GMAOpyobs v1.0.6.1 which allows for not building f2py code in CI (see GEOS-ESM/GMAOpyobs@v1.0.6...v1.0.6.1)
2. Backs off the circleci-tools orb to v2.10 which is before Intel 2021.13 came in

[mathomp4]

It worked! Yay!

[rtodling]

@mathomp4 Why does this take so long to run the CI tests?

[mathomp4]

@mathomp4 Why does this take so long to run the CI tests?

Huh. I did not notice that. Something must be wonky. Let me take a look...

[mathomp4]

I just tried a push of a quick fix. I don't think it's perfect, but should make the build faster I hope.

[rtodling]

I just tried a push of a quick fix. I don't think it's perfect, but should make the build faster I hope.

I just tried a push of a quick fix. I don't think it's perfect, but should make the build faster I hope.

ok - thank you.

[mathomp4]

I just tried a push of a quick fix. I don't think it's perfect, but should make the build faster I hope.

ok - thank you.

The "real" issue seems to be the compiler in CI land is dying trying to compile a file...but then it works the next time around. Obviously the file compiles on discover, etc., so my only thought is maybe I just need less make parallelism for that one file (i.e., compiler running out of memory?). Fingers crossed!

[rtodling]

hum - that's odd - that's no happened to me once ... and I have compiled numerous times. perhaps the compilation in the CI is not quite one-to-one w/ how things get compiled in actuality, could that be?

[rtodling]

This seems good by me. Thank you.