AVHRR Obs Validation

[azadeh-gh]

Need to validate observations from AVHRR and evaluate QC procedures in the AVHRR YAML file, as a AVHRR YAML file already exists in the GDASApp.

This issue will be used to track progress on validation of radiances in UFO for GDAS in four stages:

• HofX computation using GSI produced GeoVaLs
• QC count checks (and YAML changes) using GSI produced GeoVaLs (see comment)
• Obs errors using GSI produced GeoVaLs (see comment)
• HofX computation in FV3-JEDI

[azadeh-gh]

Naming issue: geoval and obs files name is avhrr but the name in bc is avhrr3, so cannot run the run_ufo_hofx script by using one name for obstype, I used avhrr for obstype name and inserted avhrr3_n18 in copy BC files part of the code.
QC filters: Used Andrew's updated geoval to have :

1. "VarMetaData" in naming
2. Correct troposphere pressure unit.
   GeoDir=/work2/noaa/da/acollard/UFO_eval/data/gsi_geovals_l127/nofgat_aug2021/20221013/geovals/

Used Andrew's updated obs file for corrected obs errors:
ObsDir=/work2/noaa/da/acollard/UFO_eval/data/gsi_geovals_l127/nofgat_aug2021/20221013/obs/

[emilyhcliu]

Move the contents of issue #56 to this issue.

[emilyhcliu]

Revisit AVHRR QC

• Prepare updated test input dataset (obs and geoval files)
• Test with the new input dataset
• Update YAMLs with the latest UFO features

QC flowchart for AVHRR - similar to IASI/CrIS

[emilyhcliu]

AVHRR data available for the QC and Obs Error validation period (2021080100)

• NOAA-18 --- 100% replication
• NOAA-19 --- ~100% one observation difference, likely due to numerical round-off error
• MetOp-A --- 100% replication
• MetOp-B --- 100% replication

Final QC and Observation Error Comparison between UFO and GSI (using AVHRR N19 and MetOp-A as examples)

AVHRR NOAA-19

UFO Channel 3

GSI Channel 3

UFO Channel 4

GSI Channel 4

UFO Channel 5

GSI Channel 5

AVHRR MetOp-A

UFO Channel 3

GSI Channel 3

UFO Channel 4

GSI Channel 4

UFO Channel 5

GSI Channel 5