Skip to content

Conversation

xylar
Copy link
Collaborator

@xylar xylar commented Apr 6, 2018

...everywhere except where used by ncclimo and in namelist and streams files in config files for existing E3SM runs.

Following @akturner's work to rename MPAS-Seaice (from MPAS-CICE), we need to make the same update in MPAS-Analysis. In practice this affects relatively few parts of the code because output file names from E3SM runs are read in from the streams file and we don't happen to use any namelist options that included cice in the name.

@xylar
Copy link
Collaborator Author

xylar commented Apr 6, 2018

@akturner and @milenaveneziani, it would be good to test on a new E3SM run that includes the name change. I would be happy to run a QU240 G case and test with the results, assuming the current master branch of e3sm would produce the desired result. @akturner, is that the case?

One other thing that needs to be tested is if existing runs (before this change) still work as long as we explicitly give the old names for the namelist and streams files (as I have done in this PR). I will do some testing but would appreciate some from you, @milenaveneziani, if you're able.

One other bit of cleanup that we could include in this PR is that some of the runs we refer to in the example config files are no longer available (archive or deleted by @golaz). We will need to choose some newer runs as a substitute, probably from the v1 DECK runs currently being performed. I'll work on this when I can.

@milenaveneziani
Copy link
Collaborator

@akturner: will the change to MPAS-Seaice go in before the release?
I am not worried about MPAS-Analysis, but rather about a-prime, which is a bit more involving to change (and is also a software widely used by the coupled team, so we don't want to break it).

@akturner
Copy link
Contributor

akturner commented Apr 9, 2018

@milenaveneziani: I believe so, Jon can confirm @jonbob: Jon has confirmed, this is going into the release.

@akturner
Copy link
Contributor

akturner commented Apr 9, 2018

@xylar: Not in master yet, this is the PR to do so: E3SM-Project/E3SM#2229

@xylar
Copy link
Collaborator Author

xylar commented Apr 9, 2018

Thanks for the clarification. @milenaveneziani, I think an a-prime PR should be prepared to handle this ASAP. I believe the current MPAS-Analysis would work unchanged with this change so A-Prime should be able to be updated to support the change without this PR.

I also think it would be easy to create a version of A-Prime that is backwards compatible so I don't see any downside to addressing this before E3SM-Project/E3SM#2229 gets merged.

@milenaveneziani
Copy link
Collaborator

well, but we need at least a E3SM test run for a-prime. So things have to move in parallel with E3SM-Project/E3SM#2229, I think.

@xylar
Copy link
Collaborator Author

xylar commented Apr 9, 2018

This really isn't the right place for this discussion. Let's move to the issue I created in A-Prime.

@xylar
Copy link
Collaborator Author

xylar commented Apr 16, 2018

@akturner and @jonbob, what is the status of the MPAS-Seaice rename? Can this be merged now? It should work with both the old and the new naming conventions so I don't see any downside.

@xylar xylar removed the in progress label Apr 16, 2018
@xylar
Copy link
Collaborator Author

xylar commented Apr 16, 2018

@milenaveneziani, same question to you I guess.

@akturner
Copy link
Contributor

Hi, So I'm a little confused what Rob Jacob has actually done. I think he's allowed the changes to the namelist and streams files but not other ones. If this supports the old and new I think we should proceed with this, because we plan to change the output with v2.1. Is that ok?

Copy link
Contributor

@akturner akturner left a comment

Choose a reason for hiding this comment

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

I don't have a good enough understanding of the code to approve the actual code changes. If this supports a mixture of the old and new names (Rob only allowed some of the changes) then I approve.

@milenaveneziani
Copy link
Collaborator

streams and namelist filenames have changed for sure.
I am not sure about filenames.

What I know hasn't changed are the compset names, which don't matter for this PR, of course.

* `mpascice.rst.0002-01-01_00000.nc` (or any other mpas-cice restart
* mpas-seaice files:
* `mpasseaice.hist.am.timeSeriesStatsMonthly.*.nc`
* `mpasseaice.rst.0002-01-01_00000.nc` (or any other mpas-seaice restart
Copy link
Collaborator

Choose a reason for hiding this comment

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

For the time being, I would keep the old filenames in there as well. We can remove them all together when the change to MPAS-seaice is definite.

* `streams.cice`
* `mpas-cice_in`
* `streams.seaice`
* `mpas-seaice_in`
Copy link
Collaborator

Choose a reason for hiding this comment

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

same goes with these.

@xylar
Copy link
Collaborator Author

xylar commented Apr 16, 2018

Okay, sounds good. I can make clear that some runs will have the old (cice) name while we expect newer runs to have only seaice.

@xylar xylar force-pushed the cice_seaice_rename branch from 652b16f to c5abde7 Compare April 16, 2018 20:49
@xylar
Copy link
Collaborator Author

xylar commented Apr 17, 2018

@milenaveneziani, could you make sure the changes I made address your concerns?

@milenaveneziani
Copy link
Collaborator

@xylar, for some reason, I don't see the new changes.

@xylar
Copy link
Collaborator Author

xylar commented Apr 17, 2018

I just amended the original commit.

* `mpascice.rst.0002-01-01_00000.nc`
* `streams.cice`
* `mpas-cice_in`

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

@milenaveneziani, here are the additions that you requested.

Copy link
Collaborator

Choose a reason for hiding this comment

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

great thanks.

``streams.cice``, (even if they will not be written out --
``output_interval=="none"``)::
``streams.seaice`` or ``streams.cice`` in older E3SM runs, (even if they will
not be written out -- ``output_interval=="none"``)::
Copy link
Collaborator Author

Choose a reason for hiding this comment

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

@milenaveneziani, you didn't explicitly request this but I also made changes here and on line 17 above to reference older E3SM runs.

@milenaveneziani
Copy link
Collaborator

I will test this on an existing E3SM run today.
We could also test on a newer run for which we have write access in the run/ directory to create some links: maybe @vanroekel's high-res G-case?

@xylar
Copy link
Collaborator Author

xylar commented Apr 17, 2018

The high-res G case seems like a heavy test. Maybe @darincomeau could help us create a mock-up of the name change on his low-res G case on Theta:
/projects/OceanClimate/dcomeau/20180328.GMPAS-IAF.T62_oEC60to30v3wLI.theta.GM600/run

@darincomeau
Copy link
Contributor

@xylar sure - what did you have in mind? Did you want me to run this on the theta low-res G case as is, to test that it works with cice names? Or rename cice files to seaice, and test that? Or both?

@milenaveneziani
Copy link
Collaborator

@darincomeau: I think you can just create links for streams.seaice and mpassi_in to streams.cice and mpas-cice_in, respectively, and you should be done.

@xylar
Copy link
Collaborator Author

xylar commented Apr 17, 2018

@darincomeau, yes, running as is on your run would not hurt. You will probably need to add the old names of the MPAS-Seaice namelist and streams files to your config file because the defaults have changed.

[input]
...
seaIceNamelistFileName = mpas-cice_in
seaIceStreamsFileName = streams.cice

Then, you could create the links above as @milenaveneziani says, though the correct names are:

mpas-seaice_in
streams.seaice

and you would remove (or edit) the two config entries above so it uses the new names.

(This is kind of a silly test because I'm sure it will work...)

@darincomeau
Copy link
Contributor

@xylar , @milenaveneziani , I've just finished running this on my theta run with both mpas-cice_in, streams.cice and mpas-seaice_in, streams.seaice listed in the config file, and both ran without problem. Let me know if there's anything else you'd like me to test.

@xylar
Copy link
Collaborator Author

xylar commented Apr 17, 2018

Glad to hear it. I think that should be it.

@milenaveneziani, can I merge or did you also want to test?

Everywhere except where used by ncclimo.

Use old input and streams file names in config files for existing E3SM.
@xylar xylar force-pushed the cice_seaice_rename branch from c5abde7 to c4f1d21 Compare April 17, 2018 19:40
@xylar xylar added the priority label Apr 17, 2018
Copy link
Collaborator

@milenaveneziani milenaveneziani left a comment

Choose a reason for hiding this comment

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

no, that's fine. go ahead.
Thanks @darincomeau.

@xylar
Copy link
Collaborator Author

xylar commented Apr 17, 2018

Thanks!

@xylar xylar merged commit 1be67f1 into MPAS-Dev:develop Apr 17, 2018
@xylar xylar deleted the cice_seaice_rename branch April 17, 2018 20:52
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants