Update to jcb-gdas marine obs yamls

[JohnSteffen-NOAA]

This PR updates the marine observations yamls introduced in https://github.com/NOAA-EMC/jcb-gdas/commit/41ee2a276e2688d19eef7d3dcb9567e3042f6843 by modifying the observation errors and additional filters to be consistent with the marine candidate experiment.

It addresses this GDASApp issue https://github.com/NOAA-EMC/GDASApp/issues/1316.

Additionally, yamls for new sea ice retrievals and SMAP/SMOS retrievals are included.

Close NOAA-EMC/GDASApp#1316

[guillaumevernieres]

Thanks @JohnSteffen-NOAA , can you resolve the conflict (argo yaml)? Just ignore what ever I did to the argo yaml and use your filters. I was just making stuff work!

[JohnSteffen-NOAA]

Conflict resolved

[guillaumevernieres]

Looks fine to me, did you test some of these yamls @JohnSteffen-NOAA ?

[JohnSteffen-NOAA]

No, they have not been tested from the jcb-gdas side, only from the parm/soca/obs/config @guillaumevernieres

[apchoiCMD]

Look good-

[guillaumevernieres]

Let's not merge this just yet, I don't think these yamls will work with the prep obs task. Were you able to test @JohnSteffen-NOAA ?

[JohnSteffen-NOAA]

I'm building the ctests now @guillaumevernieres . I'll update when I get to the prep obs task.

[AndrewEichmann-NOAA]

@JohnSteffen-NOAA Do you know where these are to be read in? Like from obs_list.yaml? In part I'm curious about the role of the lines

- obs space:
    name: {{observation_from_jcb}}

since the obs space name is how the file itself, at least the ones under parm/soca/obs/config, are found.

[guillaumevernieres]

@JohnSteffen-NOAA , I tested some of the obs yamls, no complaints so I'm going to merge this. However, there are a few places in our scripts that still rely in the old obs yamls ... We'll need to fix this at some point.

[guillaumevernieres]

@JohnSteffen-NOAA Do you know where these are to be read in? Like from obs_list.yaml? In part I'm curious about the role of the lines

- obs space:
    name: {{observation_from_jcb}}

since the obs space name is how the file itself, at least the ones under parm/soca/obs/config, are found.

it's coming from a processed version of obs_list @AndrewEichmann-NOAA . We have to revisit that part of the code ...

[guillaumevernieres]

The ci error seems unrelated to your work @JohnSteffen-NOAA .
@danholdaway , I'm going to ignore the failure, unless you think the issue is in this pr.

[danholdaway]

@guillaumevernieres the testing issue is fixed now.