Setup the snow DA analysis to update to the enkf ensemble members #2033
Conversation
jobs/JGLOBAL_LANDENS_ANALYSIS
Outdated
| export WIPE_DATA="NO" | ||
| export DATA=${DATA:-${DATAROOT}/${RUN}landensanl_${cyc}} |
There was a problem hiding this comment.
These are only needed if the job is using the working directory of a previous job. It appears this job is doing initialize/run/finalize all in one job.
| export WIPE_DATA="NO" | |
| export DATA=${DATA:-${DATAROOT}/${RUN}landensanl_${cyc}} |
There was a problem hiding this comment.
Removed as you suggested. Thanks.
jobs/JGLOBAL_LANDENS_ANALYSIS
Outdated
| # Begin JOB SPECIFIC work | ||
| ############################################## | ||
| # Generate COM variables from templates | ||
| YMD=${PDY} HH=${cyc} generate_com -rx COM_LANDENS_ANALYSIS |
There was a problem hiding this comment.
COM_LANDENS_ANALYSIS_TMPL is not defined anywhere in order to build this variable. Should this be
| YMD=${PDY} HH=${cyc} generate_com -rx COM_LANDENS_ANALYSIS | |
| YMD=${PDY} HH=${cyc} generate_com -rx COM_LAND_ANALYSIS |
(COM_LAND_ANALYSIS is already defined in config.com)
I also don't see a MEMDIR definition, which this will need if it is part of RUN=enkf. What RUN(s) is this part of, and if it is part of enkf, is it trying to write stuff for a bunch of different members (either as a metatask or looping over members within a single job)?
There was a problem hiding this comment.
Thanks @WalterKolczynski-NOAA I removed the whole sections because the COM directories for the ensemble members will be defined at the runtime and the MEMDIR is part of RUN=enkf.
|
@jiaruidong2017 marked this as draft because we plan to make this a metatask within the workflow in order to parallel process the 80 ensemble members |
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
|
@jiaruidong2017 |
@aerorahul Both the deterministic test and the EnKF test on C96C48 resolution succeeded as below:
What and where do I need to provide to confirm the above tests? Thanks. |
|
@jiaruidong2017 like we chatted about the other day, @aerorahul would like a test case for the workflow CI. Does his date of Dec 21, 2021 18z work for your purpose? (it looks like it does based on above). If so, please provide him with the input obs, and all the instructions to run the test (including setup_expt.py options and any changes needed to the config.* files) |
|
@CoryMartin-NOAA Yes, I used the initial conditions (/scratch1/NCEPDEV/global/glopara/data/ICSDIR/C96C48/) on Dec 21, 2021 at 18z, and conducted the test run for the four cycles to Dec 22, 2021 at 18z because the IMS assimilation is only at the 18z cycle. I made some changes to the initial conditions because of the changes of COM directory. @aerorahul How do I provide the input obs to you? |
scripts/exglobal_archive.sh
Outdated
| @@ -242,6 +243,11 @@ if [[ ${HPSSARCH} = "YES" || ${LOCALARCH} = "YES" ]]; then | |||
| targrp_list="${targrp_list} gdasice" | |||
| fi | |||
|
|
|||
| #gdasland | |||
| if [ "${DO_JEDILANDDA}" = "YES" ]; then | |||
| targrp_list="${targrp_list} gdasland" | |||
There was a problem hiding this comment.
I don't see an additional file list being created in hpssarch_gen.sh that matches this name.
There was a problem hiding this comment.
Removed the gdasland list, and we added it to gdas.txt. Thanks for your review.
parm/config/gfs/config.preplandobs
Outdated
| ##export FIMS_NML_TMPL="${HOMEgfs}/sorc/gdas.cd/parm/land/prep/fims.nml.j2" | ||
| ##export IMS_OBS_LIST="${HOMEgfs}/sorc/gdas.cd/parm/land/prep/prep_ims.yaml" | ||
| ##export CALCFIMSEXE="${HOMEgfs}/exec/calcfIMS.exe" | ||
| ##export IMS2IODACONV="${HOMEgfs}/ush/imsfv3_scf2ioda.py" |
There was a problem hiding this comment.
If these aren't needed anymore, they should just be removed.
env/HERA.env
Outdated
| echo "atmanlrun atmensanlrun aeroanlrun landanl" | ||
| echo "anal sfcanl fcst post metp" | ||
| echo "atmanlrun atmensanlrun aeroanlrun landanl landensanl" | ||
| echo "anal sfcanl fcst post vrfy metp" |
There was a problem hiding this comment.
Why is vrfy being added back in?
There was a problem hiding this comment.
Removed vrfy.
|
@jiaruidong2017 is it easy enough to provide the ICs for 12z, so that the first DA cycle is at 18z, then we only need to run 2.5 cycles instead of 4.5 |
|
@CoryMartin-NOAA Good suggestion. The ICs I used are provided by you. Do you mean I can generate the ICs at 12z by myself? Thanks. |
|
@jiaruidong2017 if you could, that would be great. @aerorahul would prefer to run:
|
|
@CoryMartin-NOAA Okay, I will generate an ICs at 12z, and conduct a 2-cycle run. |
|
The ICs and OBS are updated below: The ICs on Dec 20, 2021 at 12z are saved at: The observation files with 2 cycles are provided below: |
|
Thanks @jiaruidong2017 ! @aerorahul is this all you need now? |
|
I have copied the new files into the GDA on Hera for testing: The |
|
@aerorahul Please use branch |
|
@jiaruidong2017 @CoryMartin-NOAA We have copied the observations and the ICs to the GDA and the I have exercised the setup of the experiment as a CI test would do as follows: Please verify the experiment configuration in: |
- The UFO filters are rearranged by using the explicit filter ordering with `obs pre filters`, `obs prior filters` and `obs post filters` options. - Add `Temporal Thinning filter` into the `obs post filters` section and and arrange it after the `Background Check filter` and before the `Buddy check filter`. - Rename the IMS IODA file name and copy it into the `rundir/obs` directory for the late use in the snow analysis in the global workflow (NOAA-EMC/global-workflow#2033). Co-authored-by: Cory Martin <cory.r.martin@noaa.gov>
|
@KateFriedman-NOAA Would you please copy the new observation files below into the GDA on Hera for the |
|
@jiaruidong2017 Done: The |
|
This PR is on hold until a series of subsequent PRs are created for the deterministic case. At that point, we will either supersede this PR with a new one, or modify this PR to handle the ensemble case. |
Description
Add/Modify config.* and job files for JEDI-based Land DA analysis to update the EnKF ensemble members.
config.baseto turn on/off land DA (addDO_JEDILANDENS).config.resourcesfor the new task (addlandensanl)This PR adds in initial configuration files for support of future global land analysis capabilities for the EnKF ensemble members.
New files:
parm/config/gfs/config.landensanl
jobs/JGLOBAL_LANDENS_ANALYSIS
jobs/rocoto/landensanl.sh
scripts/exglobal_landens_analysis.py
Modified files:
config.resources
env/HERA.env
workflow/rocoto/gfs_tasks.py
workflow/rocoto/tasks.py
workflow/applications/gfs_cycled.py
How Has This Been Tested?
These are place holders for future experiment setup and XML generation.
Checklist