search

NOAA-EMC / global-workflow

Global Superstructure/Workflow supporting the Global Forecast System (GFS)
https://global-workflow.readthedocs.io/en/latest
GNU Lesser General Public License v3.0
70 stars 162 forks source link

Test cases for JEDI DA configurations #1821

Open aerorahul opened 10 months ago

aerorahul commented 10 months ago

Description A lot of new and exciting JEDI DA features have been added in the past few months. JEDI-based DA for atmosphere with and without ensemble JEDI-based DA for snow updates JEDI-based DA for aerosol updates JEDI-based DA for ocean and sea-ice

We would like to ensure that these capabilities are tested regularly with every update of the system where relevant. For e.g. an update in the ufs-weather-model hash would trigger all tests, but a change in the land_analysis will only likely impact snow DA, but that will be deferred to the SME.

Requirements The developers of these systems should provide and maintain a cycled test case for each of these individual configurations. At an appropriate time, a weakly coupled test will be added combining these configurations based on the needs of the science teams.

Acceptance Criteria (Definition of Done)

(Optional): Suggest A Solution

AndrewEichmann-NOAA commented 10 months ago

@guillaumevernieres should I be in on this?

guillaumevernieres commented 10 months ago

@guillaumevernieres should I be in on this?

probably @AndrewEichmann-NOAA , we'll figure it out internally.

CoryMartin-NOAA commented 8 months ago

@aerorahul if it's okay with you, I will wait to prepare the aerosol case once I've been able to implement and test FGAT as described in https://github.com/NOAA-EMC/global-workflow/issues/1930 as that will be our target configuration (assuming the atm has IAU on). I expect to test/debug this early next week.

RussTreadon-NOAA commented 5 months ago

PR #2199 adds C96_atmsnowDA.yaml. PR #1963 adds C48mx500_3DVarAOWCDA.yaml. Might either of these be leveraged to create cases CXX_atmUFSDA or CXXCYY_hybatmUFSDA. These cases would be JEDI-based DA for atmosphere without and with ensemble, respectively.

RussTreadon-NOAA commented 5 months ago

@jiaruidong2017 , where may I find the staged ICs used to run the Hera tests reported in PR #2287?. This PR ran C96_atmsnowDA with, I presume, an ensemble. If true, it seems the same ICs could be used to create C96_atmUFSDA and C96_hybatmUFSDA CI cases.

jiaruidong2017 commented 5 months ago

@RussTreadon-NOAA you can find the ICs used in PR #2287 as below:

/scratch2/NCEPDEV/stmp1/Jiarui.Dong/gdas.init/landid9/C96C48/output/gdas.20211220/
/scratch2/NCEPDEV/stmp1/Jiarui.Dong/gdas.init/landid9/C96C48/output/enkfgdas.20211220/
RussTreadon-NOAA commented 5 months ago

Thank you @jiaruidong2017 . These ICs work with the GFS v17 UFS weather model using the fractional grid global_cycle changes in PR #2262. Is this correct?

jiaruidong2017 commented 5 months ago

@RussTreadon-NOAA Yes, it is correct.

RussTreadon-NOAA commented 5 months ago

g-w issue #2294 provides the JEDI-based DA for atmosphere with and without ensemble test requested by this issue.