Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update global_enkf ctest for tapering of perturbations (#806) #807

Merged

Conversation

CatherineThomas-NOAA
Copy link
Collaborator

Description

The global_enkf ctest needs to be updated to include a change to the configuration of GFSv17. The upper layer ensemble spread is too large, so we are enabling a new feature in the EnKF to reduce the spread as part of the ensemble analysis. This feature was previously included in the GSI EnKF source code (Issue #728) and being turned on in the global-workflow through NOAA-EMC/global-workflow#3097. The g-w PR is not required to exercise this PR.

Fixes #806

Type of change

  • New feature (non-breaking change which adds functionality)

How Has This Been Tested?

The taperanalperts feature was tested in an ATM-only C384/C192 cycling parallel on Hera. The GSI regression tests were run on Hera (results in following comment).

Checklist

  • My code follows the style guidelines of this project
  • I have performed a self-review of my own code
  • I have commented my code, particularly in hard-to-understand areas
  • New and existing tests pass with my changes
  • Any dependent changes have been merged and published

@CatherineThomas-NOAA
Copy link
Collaborator Author

CatherineThomas-NOAA commented Nov 14, 2024

Hera ctest results

Test project /scratch1/NCEPDEV/da/Catherine.Thomas/gsi/feature_taper/build
    Start 1: global_4denvar
    Start 2: rtma
    Start 5: hafs_3denvar_hybens
    Start 4: hafs_4denvar_glbens
    Start 6: global_enkf
    Start 3: rrfs_3denvar_rdasens
1/6 Test #3: rrfs_3denvar_rdasens .............   Passed  1216.42 sec
2/6 Test #4: hafs_4denvar_glbens ..............***Failed  1880.73 sec
3/6 Test #6: global_enkf ......................   Passed  1950.77 sec
4/6 Test #1: global_4denvar ...................   Passed  2568.77 sec
5/6 Test #5: hafs_3denvar_hybens ..............   Passed  6740.22 sec
6/6 Test #2: rtma .............................   Passed  13935.89 sec

83% tests passed, 1 tests failed out of 6

Total Test time (real) = 13935.93 sec

The following tests FAILED:
	  4 - hafs_4denvar_glbens (Failed)

The hafs_4denvar_glbens test failed due to runtime:

The runtime for hafs_4denvar_glbens_loproc_updat is 297.647849 seconds.  This has exceeded maximum allowable threshold time of 294.302768 seconds,
resulting in Failure time-thresh of the regression test.

A rerun of the hafs test succeeded:

Test project /scratch1/NCEPDEV/da/Catherine.Thomas/gsi/feature_taper/build
    Start 4: hafs_4denvar_glbens
1/1 Test #4: hafs_4denvar_glbens ..............   Passed  1408.14 sec

100% tests passed, 0 tests failed out of 1

Total Test time (real) = 1408.18 sec

I will run the tests on additional machines tomorrow.

@RussTreadon-NOAA
Copy link
Contributor

@CatherineThomas-NOAA , who would you like to request two peer reviews from?

@CatherineThomas-NOAA
Copy link
Collaborator Author

@TravisElless-NOAA and @AndrewEichmann-NOAA: Would you be available to review this PR?

Copy link
Contributor

@RussTreadon-NOAA RussTreadon-NOAA left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good to me. Hera global_enkf passed. Did we confirm

  1. enkf.x ran with taperanalperts=.true.
  2. perturbations are indeed tapered

@CatherineThomas-NOAA
Copy link
Collaborator Author

From the Hera ctest global_enkf_hiproc_updat stdout:

 TAPERANALPERTS  = T,
 TAPERANALPERTS_AKBOT    =   500.000000000000     ,
 TAPERANALPERTS_AKTOP    =  -1.00000000000000

For the effect of tapering itself, I usually use the spread netcdf files produced from the epos step. Using the data from the ctest, I tried to look at the differences between the increments of a couple members, but it's hard to tell if there's an effect there. The tapering does not set spread to zero at the model top, just reduces it.

For previous comparisons when taperanalperts is turned on, see these slides

@CatherineThomas-NOAA
Copy link
Collaborator Author

Dogwood ctests

Test project /lfs/h2/emc/da/noscrub/catherine.thomas/gsi/feature_taper/build
    Start 1: global_4denvar
    Start 2: rtma
    Start 3: rrfs_3denvar_rdasens
    Start 4: hafs_4denvar_glbens
    Start 5: hafs_3denvar_hybens
    Start 6: global_enkf
1/6 Test #3: rrfs_3denvar_rdasens .............   Passed  728.50 sec
2/6 Test #6: global_enkf ......................   Passed  852.35 sec
3/6 Test #2: rtma .............................   Passed  969.81 sec
4/6 Test #5: hafs_3denvar_hybens ..............   Passed  1153.78 sec
5/6 Test #4: hafs_4denvar_glbens ..............   Passed  1214.00 sec
6/6 Test #1: global_4denvar ...................   Passed  1683.33 sec

100% tests passed, 0 tests failed out of 6

Copy link
Contributor

@RussTreadon-NOAA RussTreadon-NOAA left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Approve

Copy link

@TravisElless-NOAA TravisElless-NOAA left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Changes here correctly reflect changes made in the associated GW PR

@RussTreadon-NOAA
Copy link
Contributor

This PR can be scheduled for merger once we get a second approval from a peer reviewer.

@AndrewEichmann-NOAA , can you review this PR?

Copy link
Contributor

@AndrewEichmann-NOAA AndrewEichmann-NOAA left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

looks reasonable

@RussTreadon-NOAA RussTreadon-NOAA merged commit 997f526 into NOAA-EMC:develop Nov 19, 2024
4 checks passed
@CatherineThomas-NOAA CatherineThomas-NOAA deleted the feature/taper branch November 25, 2024 23:33
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Update global_enkf ctest for tapering of perturbations
4 participants