Validate GOES ABI radiance between FV3_GSI and MPAS_JEDI

[xyzemc]

NOTE: To use this template, remove any sections or lines that are not needed, and delete this line before submitting your PR.

Description

Validating the GOES ABI clear sky radiance assimilated in FV3 GSI and MPAS_JEDI with the case 2024052700
To make sure the difference of data count after QC between GSI and MPAS_JEDI are acceptable

Acceptance Criteria (Definition of Done)

What does it mean for this to be finished? Please add links to relevant PRs. Links may be added later if a PR has not yet been created.

• Link any relevant pull requests here:
  • PR #
  • PR #

Dependencies

What must be done before this can be done? Add issue or PR dependencies as appropriate
Does this block progress on other issues? Add this issue as a dependency to issues as appropriate

Additional information (optional)

Add any other context about the problem here.

[xyzemc]

Channel 8
GSI simulated Brightness Temperature after QC

MPAS_JEDI simulated Brightness Temperature after QC

[xyzemc]

Channel 9
GSI simulated Brightness Temperature after QC

MPAS_JEDI simulated Brightness Temperature after QC

[xyzemc]

Channel 10
GSI simulated Brightness Temperature after QC

MPAS_JEDI simulated Brightness Temperature after QC

[xyzemc]

To investigate why ABI channel 10 has fewer observations passing quality control (QC) in JEDI compared to GSI, I hypothesized that the thinning method in JEDI might be influencing other QC filters for channel 10, despite the total number of rejected observations being consistent across all channels.
To test this hypothesis, I utilized the unthinned ABI data written out from GSI and conducted a run using FV3-JEDI. The results showed that the number of observations passing QC for channel 10 in JEDI remained lower than in GSI, while channels 8 and 9 saw an increase in the number of observations passing QC. These findings suggest that factors beyond the thinning process may be contributing to the observed discrepancies, warranting further investigation.

         Unthinned data passed QC in GSI  Ch.8

Unthinned data passed QC in JEDI Ch.8

          Unthinned data passed QC in GSI  Ch.9

Unthinned data passed QC in JEDI Ch.9

          Unthinned data passed QC in GSI  Ch.10

Unthinned data passed QC in JEDI Ch.10

[xyzemc]

Phase 1 Validation for FV3-JEDI
To run the Phase 1 validation using the gsidiag-converted IODA observation and geoval files, I implemented the following fixes:
Variable Name Updates: Modified six variable names in the geoval file to align with the latest UFO changes.

#./change_variable.sh surface_temperature_where_sea  skin_temperature_at_surface_where_sea
#./change_variable.sh surface_temperature_where_land  skin_temperature_at_surface_where_land
#./change_variable.sh surface_temperature_where_ice skin_temperature_at_surface_where_ice
#./change_variable.sh surface_temperature_where_snow  skin_temperature_at_surface_where_snow
#./change_variable.sh humidity_mixing_ratio water_vapor_mixing_ratio_wrt_dry_air
#./change_variable.sh surface_wind_speed  wind_speed_at_surface
#./change_variable.sh surface_wind_from_direction wind_from_direction_at_surface

CRTM Input Adjustment: Removed CO₂ as an absorber in yaml file, ensuring compliance with UFO-required CRTM inputs.
The GSI code used for write out the geoval file is: https://github.com/NOAA-EMC/GSI/tree/feature/ufo_geovals
I did a minor modification for ABI in this GSI.

Following how global did the phase1 validation for satellite radiance data:
srun -l -n 160 ${RDASApp}/build/bin/test_ObsFilters.x

[xyzemc]

The the unit of mixing ratio from gsi geovals is g/kg, but JEDI changed the kg/kg as stated in https://github.com/orgs/JCSDA-internal/discussions/157#discussioncomment-11387594
So I did the unit conversion in gsi setuprad code:
if (iabsorb == 1) then !! Changed the unit of humidity mixing ratio from g/kg to kg/kg in UFO in Oct 2024. call nc_diag_data2d(trim(fieldname), atmosphere(1)%absorber(:,iabsorb)*1e-3) ! check %absorber_units else call nc_diag_data2d(trim(fieldname), atmosphere(1)%absorber(:,iabsorb)) ! check %absorber_units end if
Before the mixing ratio unit converted (g/kg)

After the mixing ratio unit converted to kg/kg