Implement equation of time in RRTMG shortwave radiation

[clark-evans]

This PR implements the equation of time shortwave radiation correction in the RRTMG shortwave radiation parameterization. It is based on MPAS-Dev/MPAS-Model PR MPAS-Dev#1172 (MPAS-Dev#1172). This is expected to correct for the ~15-min insolation time biases seen within our real-time MPAS forecasts over the last 15 months.

I expect this to be answer changing, as it affects the incident shortwave radiation. The regression tests indicate small differences in forecast output at all forecast times (12, 24, 36, 48, 60 min).

Priority Reviewers

• Please list the developers/collaborators you'd like to prioritize for review
  • @joeolson42
  • @AndersJensen-NOAA

[AndersJensen-NOAA]

Thanks Clark! To make this more general, I would write a function in mpas_atmphys_driver_radiation_sw.F (below the subroutine radconst), call the function after radconst is called and then add eot=eot to the sw procedure. That's my 5 cents (they don't make pennies anymore so I'm rounding).

[clark-evans]

Thanks Clark! To make this more general, I would write a function in mpas_atmphys_driver_radiation_sw.F (below the subroutine radconst), call the function after radconst is called and then add eot=eot to the sw procedure. That's my 5 cents (they don't make pennies anymore so I'm rounding).

I'm open to doing that, but I'd like to minimize changes relative to the upstream as much as possible to ease future code merges. Barring Wei re-doing her upstream PR, which I haven't gotten a sense is likely, I think WYSIWYG (or at least close to it). Happy to discuss more, though.

[dustinswales]

@clark-evans For MPAS in the UWM, we will be using RRTMG (or RRTMGP) from the CCPP, which is different than RRTMG in MPAS-Dev. So we will need to implement this change there if needed.