Update for WW3 initialization #2585
Conversation
|
@JessicaMeixner-NOAA Since I don't think any RT will cover this case, have you tested this feature to ensure possible failures have been captured? For example, in an RT sandbox by requesting a binary restart for an initial run but not having one present, or having one with the wrong name, etc. |
|
@DeniseWorthen I have tested this within the global-workflow low resolution test ci test case and ensured that the significant wave height was non-zero at forecast hour 0. The global-workflow clone is here (note I only changed my branch of WW3, I didn't have this ufs-waether-model PR yet+ it's a branch of g-w): /scratch1/NCEPDEV/climate/Jessica.Meixner/UpdateFixandHiresIC/global-workflow and my output of the test is here: /scratch1/NCEPDEV/climate/Jessica.Meixner/UpdateFixandHiresIC/c48t05 I'm going to submit a higer-res case from the global-workflow that doesn't have just an initial IC to triple check things now. |
|
@DeniseWorthen - Additional sanity check from g-w with a C768 HR like set-up can be found here: /scratch1/NCEPDEV/climate/Jessica.Meixner/UpdateFixandHiresIC/c768t02/COMROOT/c768t02/gfs.20191203/00/products/wave/gridded/global.0p25 and the f000:
|
|
@JessicaMeixner-NOAA Thanks. I wasn't really concerned about the validity of starting from binary, rather that the option worked correctly w/ your code changes---that if run = initial and the binary file was not present, it would abort for example. This isn't something we would test for in the RTs, so I just wanted to be sure you had tested out various scenarios for how this could not do what you expected it to do. |
|
I have not tested that @DeniseWorthen - I can't test that from the workflow, because I have it designed to not get to that scenario. I'd have to either change the workflow to force this or find an rt.sh test that has these cases. I was more concerned that it actually read the IC. |
|
Have confirmed we get the correct error if not binary WW3 restart exists: Test directory: Error message: |
|
Great. Thanks for verifying that it fails as expected. |
|
@JessicaMeixner-NOAA Jong wanted to combine PRs so I brought in these changes into PR #2587 . Looked good from my end but feel free to look it over when you have a chance to double-check that the correct changes were brought in. Thanks |
|
Changes look correctly added to me @dpsarmie |
…ations // #2585 Update for WW3 initialization (#2587) * UFSWM - Add Land IAU variables to default_vars and allow control_c48_lnd_iau to run using the global_control.nml.IN template. * WW3 - Updated hash for WW3 initialization (#2585) --------- Co-authored-by: Jessica Meixner <jessica.meixner@noaa.gov>
3 participants
Commit Queue Requirements:
Description:
When using PIO restarts, we needed to add a capability to read binary restarts when using runtype=intial. For example, this is used in the high resolution (HR) experiments.
Commit Message:
Update for WW3 initialization
Priority:
Git Tracking
UFSWM:
Sub component Pull Requests:
UFSWM Blocking Dependencies:
Changes
Regression Test Changes (Please commit test_changes.list):
Input data Changes:
Library Changes/Upgrades:
Testing Log: