all-sky ATMS radiance assimilation and using surface-sensitive ATMS radiances

[gmao-yzhu]

This PR is to achieve the following purposes:

1. Use a common beamwidth for all channels,
2. Change ATMS radiance usage in GEOS from clear-sky approach to all-sky approach.
3. Add the assimilation of surface-sensitive channels 1-4 and 16 in GEOS.
4. Add cloud seed for rain and snow.
5. Code changes on DA side for cloud fraction and overlap are also included (Wei, Jianjun, me have tested the impact on OmFs of CrIS, GMI, ATMS but not turned on yet).

[github-actions]

Label error. Requires at least 1 of: 0 diff, 0 diff trivial, Non 0-diff, 0 diff structural, 0-diff trivial, Not 0-diff, 0-diff, automatic, 0-diff uncoupled. Found:

[gmao-wgu]

The code changes look good to me.

[gmao-wgu]

These changes look good to me!

[mhan-mars]

I was wondering why channel 15 is still set to "-1" in gmao_global_satinfo.txt?

[gmao-wgu]

I was wondering why channel 15 is still set to "-1" in gmao_global_satinfo.txt?
It does not matter. Whether it is active or passive is determined by the resource file active_channels.tbl.

[mhan-mars]

The changes look good to me.

[ahrem1234]

I had a conversation with Yanqiu on this PR, and we found out that there are several bug fix, too.

1. changing water vapor unit for JEDI-based GEOS near future
2. using only ocean data for VarBC coefficient initialization
   with this in mind, the changes looks great!

[gmao-yzhu]

Ahreum,

Here are answers to your questions:

1. It seems like you change the code to calculate BC coefficients over the ocean only. Is it applied to other observations? Yes, the initialization of BC coefficients is changed to only use data over ocean now, and this is applied to other observations.
2. Do you limit the calculation of ework value only to ATMS? Since AMSU-A is still clear-sky approach, so this ework is only applied to ATMS.
3. Do you also add the use of ATMS on NOAA21 for all-sky DA? Yes.
4. Do we need cloudcover_maximum_overlap, cloudcover_random_overlap, cloudcover_maxran_overlap, cloudcover_overcast_overlap variables for the future all-sky DA? Yes, the cloud overlap schemes are needed along with the use of cloud fraction for better CRTM brightness temperature simulation in the future.

Thank you very much for reviewing and approving this PR.

[mhan-mars]

I had a conversation with Yanqiu on this PR, and we found out that there are several bug fix, too.

1. changing water vapor unit for JEDI-based GEOS near future
2. using only ocean data for VarBC coefficient initialization
   with this in mind, the changes looks great!

The water vapor unit change, with a refinement by Ricardo, has been added to setuprad in the latest version.