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.

Sign up for GitHub

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

Jump to bottom

Corrects icepack initialization for bgc. #175

Closed
wants to merge 1 commit into from
Closed

Corrects icepack initialization for bgc. #175

wants to merge 1 commit into from

Conversation

njeffery
Copy link
Contributor

Adds missing namelist fields.

Also added gx1 tests for zbgc (bgc,bgcgx1) and skl (bgcskl,sklgx1)
Required nutrient forcing filenames are hard coded (nitrate_climatologyWOA_gx1v6f.nc and silicate_climatologyWOA_gx1v6f.nc)
Namelist field bgc_data_dir needs to be properly specified.
Note gx1 zbgc test is really slow and fails on wolf
Tested with icepack@1e5f2fa

@eclare108213
Copy link
Contributor

@njeffery Thank you for adding tests! Can these be run with gx3, even if the results aren't realistic (nothing about gx3 is realistic)? Maybe set nitrate and silicate forcing as constants for short-time runs?

Your Icepack conflicts with the one on master, and I think we can't merge this PR like this. Could you update to the master version, please?

We'll need to treat the forcing files and directories consistently with other forcing in the code. This is a little different from how we've done things in the past.

And finally, we will need to update the documentation for all of this.

@eclare108213 eclare108213 mentioned this pull request Aug 14, 2018
Closed
@njeffery
Copy link
Contributor Author

njeffery commented Aug 14, 2018 via email

@eclare108213
Copy link
Contributor

Sounds like a memory access issue. If we had a gx3 test, it would be easier to debug... or maybe gx3 wouldn't fail, and that would be a useful data point too. How far into the gx1 run does it fail? Do bgc restarts work?

@njeffery
Corrects icepack initialization for bgc.
b681ff9 
Adds missing namelist fields.
Also added gx1 tests for zbgc (bgc,bgcgx1) and skl (bgcskl,sklgx1)
Required nutrient forcing filenames are hard coded (nitrate_climatologyWOA_gx1v6f.nc and silicate_climatologyWOA_gx1v6f.nc)
Namelist field bgc_data_dir needs to be properly specified.
Note  *gx1 zbgc test is really slow and fails on wolf*
@njeffery njeffery force-pushed the njeffery_zbgc3DForcing branch from bd82b41 to b681ff9 Compare August 15, 2018 19:39
@eclare108213
Copy link
Contributor

@njeffery Let me know when you might have a chance to sit down together and talk through this. Maybe I can help?

@duvivier
Copy link
Contributor

duvivier commented Sep 6, 2018

@njeffery I have the files downloaded and will add them to the tarballs shortly. Is there anything else you need me to comment on at this time or is the review on hold till the memory issues are sorted?

@njeffery
Copy link
Contributor Author

njeffery commented Sep 7, 2018

Thanks @duvivier. The review is on hold due to the memory issues, though I suspect they are already present in master.

@duvivier
Copy link
Contributor

@njeffery Just a quick question here since I'm making sure these files (and all the others) are updated on the ftp but in the right directories.

In /CICE/cicecore/cicedynB/general/ice_forcing_bgc.F90 it appears that For both the nitrate_climatologyWOA_gx1v6f.nc and silicate_climatologyWOA_gx1v6f.nc files it should look for them in bgc_data_dir, which the user would need to specify in the namelist.

Can you specify the directory structure we should use? The current COREII forcing we provide uses this structure:
CICE_data/forcing/gx1/COREII/{4xDAILY,MONTHLY}/

@njeffery
Copy link
Contributor Author

njeffery commented Sep 17, 2018 via email

@eclare108213
Copy link
Contributor

My intent was to get rid of bgc_data_dir, and just use ocn_data_dir for the bgc data. Is there any chance we'd have atm bgc data? Given that change, I'd prefer this directory structure:
CICE_data/forcing/gx1/WOA/MONTHLY

@njeffery
Copy link
Contributor Author

njeffery commented Sep 18, 2018 via email

@eclare108213
Copy link
Contributor

eclare108213 commented Sep 18, 2018 via email

@njeffery
Copy link
Contributor Author

njeffery commented Sep 18, 2018 via email

@duvivier
Copy link
Contributor

duvivier commented Sep 18, 2018 •
edited
Loading

I'm also inclined to the CICE_data/forcing/gx1/WOA/MONTHLY format instead of adding bgc in there just for the flexibility. I'll go with that and hopefully we can eliminate bgc_data_dir that way.

For future forcing from the atmosphere with BGC I'd think we could just do: CICE_data/forcing/gx1/CAM/MONTHLY
Since we could potentially have CAM forcing files for the physical state as well.

@eclare108213
Copy link
Contributor

sounds good to me

@duvivier
Copy link
Contributor

Nicole, we want to have all new files added use the date created in case there are future updates. Is there any problem with me appending yyyyddmm to these (20180814 is when you emailed them to me)? You could also re-make them and send them to me with the date? Then a softlink could just be made in the data directory for the dated file name to the general file name.

@njeffery
Copy link
Contributor Author

njeffery commented Sep 20, 2018 via email

@duvivier
Copy link
Contributor

These files have been added to the CICE forcing and are now available on the ftp (CICE_data_new.tar.gz, CICE_data_forcing_gx1_MONTHLY.tar.gz, and CICE_data_all.tar.gz all have the new files, but people with the data already can just download the first file).

I've added documentation to the wiki about new files as well for these and setting the stage for future additions. https://github.com/CICE-Consortium/CICE/wiki/CICE-Input-Data

Let me know if there's anything else you need from me for this PR.

@eclare108213
Copy link
Contributor

E's summary of where this PR is:
Includes new tests for zbgc and also skeletal layer bgc on both gx3 and gx1.

(Some of?) these tests utilize new forcing files which are hard-coded. The files need to be handled the same way as other forcing files. @duvivier updated the file names on the ftp site, adding dates.

Documentation still needs to be added for the new tests. It's not clear from this github conversation, which tests have forcing data in files and which use internally generated data. It looks like gx1 has files and gx3 doesn't?

The gx1 configuration crashes on wolf, which @njeffery says might be due to a problem already in master. We need to run several sets of tests, and post the results to the wiki:

  1. a complete set of these new tests using the most recent file updates. It would be nice to run the tests on the various machines used for our regular testing as well as on wolf, to see if the problem is isolated to wolf or is more general.
  2. run the test suite on wolf without these changes, to see if the issue is already on master.

@eclare108213
Copy link
Contributor

@njeffery is running the various tests with debug flags on as a first step to tracking down the issue with this part of the code.

@apcraig
Copy link
Contributor

apcraig commented Oct 13, 2018

Is this redundant with #207 or are those two separate things?

This branch will need to be updated to master to address changes in how tracers are handled (as namelist, not cpps). This will require changes to set_env and set_nml files. Sorry about that. Let me know if you need some help.

@eclare108213
Copy link
Contributor

@njeffery Is this PR is superseded by #207, so I can close it?

@njeffery
Copy link
Contributor Author

njeffery commented Oct 17, 2018 via email

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@eclare108213 eclare108213 Awaiting requested review from eclare108213

@duvivier duvivier Awaiting requested review from duvivier

Assignees

@eclare108213 eclare108213

@apcraig apcraig

Labels
Testing Type: Bug
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
4 participants
@njeffery @eclare108213 @duvivier @apcraig