Add support for CADS IR cloud detection scheme in the GSI #2478
Conversation
…nd true/false flags to us CADS for airs, cris, and iasi instruments.
WalterKolczynski-NOAA
changed the title
|
@wx20jjung Is there a GSI update that needs to be included in this PR? If so, please mark it as a draft until ready. |
|
Walter, The GSI develop branch being used by the global-workflow already has the necessary CADS code. Everything should be complete. I have tested both the default IR cloud detection scheme and the new CADS on Hera. Both work as expected. |
|
@wx20jjung Thanks for the confirmation. Please update your branch to the current develop and we will run CI tests. |
|
Walter, CADS should be up to date now with global-workflow:develop |
Did you push them? I still have a button that says "Update branch" and I don't see any new commits. |
|
Sorry. Please see if I did it properly this time. |
WalterKolczynski-NOAA
added
CI-Hera-Ready
emcbot
added
CI-Orion-Building
TerrenceMcGuinness-NOAA
added
CI-Orion-Passed
|
All scheduled CI test cases passed (updated label). |
* upstream/develop: Update fbwind for COM refactor (NOAA-EMC#2479) Update parm files for atm product (NOAA-EMC#2486) Fix paths in prepocnobs task (NOAA-EMC#2459) Add oceanalecn to workflow generator (NOAA-EMC#2409) Add support for CADS IR cloud detection scheme in the GSI (NOAA-EMC#2478) Improve error messaging to resolve bugzilla (NOAA-EMC#2468) Check the DO_JEDISNOWDA condition before adding snow DA analysis to the list (NOAA-EMC#2471)
|
@WalterKolczynski-NOAA and @wx20jjung : The changes to Where do these files reside? I do not see these fix files in GSI-fix or in EIB maintained |
|
Russ,
I have my copies on hera
/scratch1/NCEPDEV/jcsda/Jim.Jung/save/global-workflow/fix
Please put them in the proper directory within github.
Thanks,
Jim
…On Mon, Apr 22, 2024 at 12:06 PM RussTreadon-NOAA ***@***.***> wrote:
@WalterKolczynski-NOAA <https://github.com/WalterKolczynski-NOAA> and
@wx20jjung <https://github.com/wx20jjung> : The changes to
scripts/exglobal_atmos_analysis.sh in this PR include three new fix files
${NLN} ${FIXgfs}/gsi/AIRS_CLDDET.NL AIRS_CLDDET.NL
${NLN} ${FIXgfs}/gsi/CRIS_CLDDET.NL CRIS_CLDDET.NL
${NLN} ${FIXgfs}/gsi/IASI_CLDDET.NL IASI_CLDDET.NL
Where do these files reside? I do not see these fix files in GSI-fix
<https://github.com/NOAA-EMC/GSI-fix> or in EIB maintained FIX_DIR. Am I
looking in the wrong places?
—
Reply to this email directly, view it on GitHub
<#2478 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMPASA4IVYQGX2XGOJKMN3TY6UYRRAVCNFSM6AAAAABGAMEACWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANZQGA2DMMJRGQ>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
|
@wx20jjung , thank you for pointing me at the required CADS fix files. After looking at the |
|
@wx20jjung , thank you for stopping by my cubicle. Looking at A closer check of GSI-fix:develop shows that RTMA has a namelist file, That said, it's my impression that @wx20jjung , how often do you think users will need to modify the contents of the To keep things simple, I'm inclined to add the |
|
Russ,
Once set, I would expect the CLDDET.NL files to remain mostly static. The
main tests used by this software use obs-background (forecast) differences
(in observation space) so the forecast model resolution has minimal
effects. Unless there are significant changes to the forward model in the
CRTM, I would not expect the need to change these parameters. Instrument,
and/or its corresponding imager (CrIS-VIIRS, IASI-AVHRR) changes/failures
would most likely require a review of the various parameters. I would
expect these files could be edited on a frequency of years or longer.
…On Mon, Apr 22, 2024 at 1:25 PM RussTreadon-NOAA ***@***.***> wrote:
@wx20jjung <https://github.com/wx20jjung> , thank you for stopping by my
cubicle. Looking at gsi_enkf.fd/src/gsi/cads.f90 confirms that the
CLDDT.NL files are fortran namelists. Subroutine CADS_Setup_Cloud in
cads.fd90 reads the contents of these files as namelist
Cloud_Detect_Coeffs.
A closer check of GSI-fix:develop <https://github.com/NOAA-EMC/gsi-fix>
shows that RTMA has a namelist file, rtma_parmcard.input
<https://github.com/NOAA-EMC/GSI-fix/blob/develop/rtma_parmcard_input> in
GSI-fix. Given this precedent, we could add the new CLDDET.NL files to
GSI-fix.
That said, it's my impression that fix/ is for static or mostly static
files. If rtma_parmcard_input or the new CLDDET.NL files change more
frequently, they should either be part of the application script or the
parm/ directory.
@wx20jjung <https://github.com/wx20jjung> , how often do you think users
will need to modify the contents of the CLDDET.NL files. You described
them as tuning files. This implies to me that these files are static once
tuning is done. Is this true? How often is tuning needed? Do we need to
retune when (a) change model resolution or physics change, (b) CRTM,is
updated, (c) the instrument changes, or ...?
To keep things simple, I'm inclined to add the CLDDET.NL files to GSI-fix
but I'd first like to get a sense of how often users will be changing these
files.
—
Reply to this email directly, view it on GitHub
<#2478 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMPASA6YGXA4Z2CTN7OZQG3Y6VBXPAVCNFSM6AAAAABGAMEACWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANZQGMZDQMZVGY>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
A new infrared cloud detection scheme "CADS" was added to the GSI. These changes allow the use of this new scheme within the global-workflow. Setting the namelist variables to .true. (e.g. cris_cads=.true.) will envoke the the new scheme for that instrument.
Resolves #2473
Type of change
Change characteristics
How has this been tested?
Ran several cycles on Hera with the default cloud detection scheme and CADS.