modradrrtmg: merge duplicate definitions of orbital parameters

[fjansson]

Some orbital parameters were defined both in modradrrtmg and modraddata. Moved all orbital parameters to modraddata, which is where shr_orb_decl() looks for them. This is to resolve use of undefined variables found on Fugaku (debug build). Note the orbit calculations aren't actually used anyway.

[CasparJungbacker]

Another reason to explicitly pass variables as arguments to subroutines :^).

Looks good, but when you say "the orbit calculations aren't actually used", do you mean that they're just not used by RRTMG, or at all?

[fjansson]

I mean at all, at least that's what I believe but a double check would be good. There's somewhere a much simpler orbit calculation which is used instead, with the assumption that the orbit is a perfect circle. So the times of sunrise, sunset are off by up to ~30 min.
And yes, this is a case where function arguments would be great.

[CasparJungbacker]

Hmmm, we have setupSW() in both RRTMG and RRTMGP, which both call shr_orb_decl() from shr_orb_mod, so at least it's called. Whether the call actually does something is a different question, I think the values it touches in modraddata are not used at all.

[fjansson]

the setup is called, but then to actually calculate the zenith angles, the function zenith in modraddata is used, and that one seems to not care about the fancy orbit parameters.

We should fix that too, but let's do another PR for it, and check the zenith angles with some online astronomic calendar.