Sea Ice simulations discussion using CRTM (v2.4.x and v3.x)

[BenjaminTJohnson]

@KatherineLukens-NOAA @rgrumbine

https://journals.ametsoc.org/view/journals/atot/39/10/JTECH-D-21-0187.1.pdf

I'm aware that NCO/EMC still uses CRTM v2.4.1, but I typically handle all of my ongoing work in the CRTMv3 repository, that's why this issue is here.
I can apply any code modifications to both, so it's not an issue.

This issue can be used for discussion, questions, and updates. I will first work on a variational simulation of CRTM brightness temperatures as a function of all sea-ice specific variables. There's one really flubby thing about CRTM (prior to my joining) was that NESDIS wrote instrument-specific sea-ice models (and other emissivity models), so you have stuff like this:

NESDIS_AMSRE_SICEEM_Module.f90
NESDIS_AMSU_SICEEM_Module.f90
NESDIS_ATMS_SeaICE_LIB.f90
NESDIS_ATMS_SeaICE_Module.f90
NESDIS_MHS_SICEEM_Module.f90
NESDIS_SEAICE_PHYEM_MODULE.f90
NESDIS_SSMIS_SeaIceEM_Module.f90
NESDIS_SSMI_SIceEM_Module.f90

So there's these instrument specific sea-ice modules, which take observed TBs as an input and produce a retrieved emissivity, which is then used to simulate the TB... so in a sense, it's already a retrieval algorithm, but probably not a very good one. The input TBs are not required, however if they're not provided, it falls back to an emissivity estimate (NESDIS_SEAICE_PHYEM_MODULE.f90).

I'll keep digging and updating here.