Observer End-to-End Test for ASCAT

[emilyhcliu]

End-to-End Tests for Scatterometer Wind (ASCAT) processing

• [x] Create BUFR converter application for scatwind- python script (Note: wind direction and speed are converted to u- and v-wind components)
• [x] Test BUFR converted IODA obs input to H(x) ---- issue found and fixes
• [x] Test BUFR converted IODA obs input to H(x) + initial error assignment + read routine QC
• [x] Test BUFR converted IODA obs input to H(x) + initial error assignment + read routine QC + setup routine QC --- work in progress ...

This is a plumbing test between IODA and UFO. Scientific assessment will be performed later by satellite wind experts.

Check results for ASCAT MetOp-A Check results for ASCAT MetOp-B

[emilyhcliu]

Test ASCAT winds from MetOp-A from BUFR converted IODA file to HofX + QC (GSI read routine QC)

UFO O-F values do not match closely with GSI O-F values UFO simulated ascat winds fit worse to observations

U-Component Wind O-F : UFO vs GSI :

[emilyhcliu]

V-Component Wind: UFO vs. GSI

[emilyhcliu]

Re-check HofX (no QC) between GSI and UFO (using geoval from GSI) --- should be 100% matched.

GSI vs. JEDI U-wind	GSI vs. JEDI V-wind

[emilyhcliu]

Why does UFO result show larger biases? Make the long story into a few bullets below:

• In UFO evaluation, we use GSI nc-diagnostic file
• GSI calculated pressures at observation locations are provided in the obs file from GSI.
• However, pressures at observation locations do not come with BUFR file.
• So, the observation pressure is unknown, but we do know the observation height is 10 meters.
• Therefore, we need to use alternative options for the observation operator in UFO (VertInterp)
• This also needs a code change in VertInterp (one-line fix) to make it right.

Original Configuration

    obs operator:
      name: VertInterp
      apply near surface wind scaling: true

Modified Configuration:

    obs operator:
      name: VertInterp
      vertical coordinate: geopotential_height
      observation vertical coordinate: heightOfObservation
      apply near surface wind scaling: true

Here is the result from the modified configuration: U-Wind (UFO vs. GSI)

U-Wind UFO vs. Obs	U-Wind GSI vs. Obs

V-Wind (UFO vs. GSI)

V-Wind UFO vs. GSI	V-Wind UFO vs. GSI

[emilyhcliu]

Run the standard CTEST for SCATWIND HofX to verify the YAML (options for obs operator changed) and code change are good. (This is using geoval and obs files from GSI )

U-Wind GSI vs. UFO	V-Wind GSI vs. UFO

[emilyhcliu]

More Investigation: Scatwind Fix....

[emilyhcliu]

GMAO made some changes in the UFO repository related to surface winds: There are several UFO updates from GMAO regarding winds.
This PR is to test the updates from GMAO for Scatterometer Winds

Here is the collection of related UFO PRs PR #3061 PR #3050 PR #3026

So, Let's evaluate the changes ....

[emilyhcliu]

MetOp-A (UFO vs. GSI)

Note:

• For UFO: No data thinning
• For GSI: Input data already thinned, data over non-water surfaces and surface temperature less than 273 are removed

O-F without QC

O-F with QC

Final Obs Error

[emilyhcliu]

MetOp-B (UFO vs. GSI)

Note:

• For UFO: No data thinning
• For GSI: Input data already thinned, data over non-water surfaces and surface temperature less than 273 are removed

O-F without QC

O-F with QC

Final Obs Error