Feature/yury/s2s3 merge #293
Conversation
Sync develop into master
…s of salt flux due to SSS relaxation to SSS_MIN, which prevents model from crashing due to low salinity.
According to Atanas, this update fixes a bug to set the proper mode to write to RAM in dual ocean mode for OGCM and children.
yvikhlya
added
0 diff
yvikhlya
added
the
communication/coordination
|
@mathomp4 By the way, I just noticed that Externals.cfg has geos/v1.0.1 tag for MOM5, which does not have the required export, and this will not compile. Should I change it to geos5 branch? |
No. The correct thing would be to make a new release of MOM5 (geos/v1.0.2) and then @sdrabenh can update the Externals/components.yaml to that. There are some changes in the geos5 branch that haven't been released yet: |
|
I could make a new release (or you could) and then you could edit the Externals.cfg at the same time. (Note, the |
|
Yury,
Matt (and Ben and I) are in a meeting at the moment. We are trying to move to use mepo (and the related config file “components.yaml”. The syntax there is self explanatory)
Just run at the top level (i.e. GEOSgcm)
mepo init
mepo clone (of course after you make a change to components.yaml)
Atanas
From: Yury Vikhliaev <notifications@github.com>
Reply-To: GEOS-ESM/GEOSgcm_GridComp <reply@reply.github.com>
Date: Friday, June 5, 2020 at 1:50 PM
To: GEOS-ESM/GEOSgcm_GridComp <GEOSgcm_GridComp@noreply.github.com>
Cc: "Trayanov, Atanas L. (GSFC-610.1)[SCIENCE SYSTEMS AND APPLICATIONS INC]" <atanas.l.trayanov@nasa.gov>, Review requested <review_requested@noreply.github.com>
Subject: [EXTERNAL] Re: [GEOS-ESM/GEOSgcm_GridComp] Feature/yury/s2s3 merge (#293)
@mathomp4<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_mathomp4&d=DwMCaQ&c=ApwzowJNAKKw3xye91w7BE1XMRKi2LN9kiMk5Csz9Zk&r=_r6-SBvtAqSGH0t7NYy2fPpq1wyPKQV-8ZiMEmovLb4&m=IlgB4S7FyuAFjEEq0Ha4YH5tl-m-MC8uFXN0Bu6LHfc&s=g4E8e_qxvfwOkV_0REmDHYaC2Xh-ISDk8v0WBMPcN3U&e=> By the way, I just noticed that Externals.cfg has geos/v1.0.1 tag for MOM5, which does not have the required export, and this will not compile. Should I change it to geos5 branch?
—
You are receiving this because your review was requested.
Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_GEOS-2DESM_GEOSgcm-5FGridComp_pull_293-23issuecomment-2D639664481&d=DwMCaQ&c=ApwzowJNAKKw3xye91w7BE1XMRKi2LN9kiMk5Csz9Zk&r=_r6-SBvtAqSGH0t7NYy2fPpq1wyPKQV-8ZiMEmovLb4&m=IlgB4S7FyuAFjEEq0Ha4YH5tl-m-MC8uFXN0Bu6LHfc&s=m6kUl5mgnGqPmsUYjNZgDkWt8k0caJFodS_qxRPK17c&e=>, or unsubscribe<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AL6ZCVPZ3E3SYFHGGPARQR3RVEV3HANCNFSM4NUIUUFA&d=DwMCaQ&c=ApwzowJNAKKw3xye91w7BE1XMRKi2LN9kiMk5Csz9Zk&r=_r6-SBvtAqSGH0t7NYy2fPpq1wyPKQV-8ZiMEmovLb4&m=IlgB4S7FyuAFjEEq0Ha4YH5tl-m-MC8uFXN0Bu6LHfc&s=Dn9kKurEChuIGuLTjKF4E--VmNMbxnhhHE9c56qzUw4&e=>.
|
Can I ask you to handle the release and dependency fix? You know the rules better than me. |
|
Yury,
I’ve never done a release. We can ask Matt, could you cc him?
From: Yury Vikhliaev <notifications@github.com>
Reply-To: GEOS-ESM/GEOSgcm_GridComp <reply@reply.github.com>
Date: Friday, June 5, 2020 at 2:05 PM
To: GEOS-ESM/GEOSgcm_GridComp <GEOSgcm_GridComp@noreply.github.com>
Cc: "Trayanov, Atanas L. (GSFC-610.1)[SCIENCE SYSTEMS AND APPLICATIONS INC]" <atanas.l.trayanov@nasa.gov>, Review requested <review_requested@noreply.github.com>
Subject: [EXTERNAL] Re: [GEOS-ESM/GEOSgcm_GridComp] Feature/yury/s2s3 merge (#293)
I could make a new release (or you could) and then you could edit the Externals.cfg at the same time. (Note, the components.yaml in GEOSgcm would need to change at the same time)
Can I ask you to handle the release and dependency fix? You know the rules better than me.
—
You are receiving this because your review was requested.
Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_GEOS-2DESM_GEOSgcm-5FGridComp_pull_293-23issuecomment-2D639672282&d=DwMCaQ&c=ApwzowJNAKKw3xye91w7BE1XMRKi2LN9kiMk5Csz9Zk&r=_r6-SBvtAqSGH0t7NYy2fPpq1wyPKQV-8ZiMEmovLb4&m=vI8Y5RIlRI_CS91aqdQ1atWgAzF_n3YtJiWy24TuDcs&s=mUxX4816K0UDzL-p0xEflBC3K2i4qEInGPUM5CSqREc&e=>, or unsubscribe<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AL6ZCVN6TPL63TLAMRQ3LWDRVEXOXANCNFSM4NUIUUFA&d=DwMCaQ&c=ApwzowJNAKKw3xye91w7BE1XMRKi2LN9kiMk5Csz9Zk&r=_r6-SBvtAqSGH0t7NYy2fPpq1wyPKQV-8ZiMEmovLb4&m=vI8Y5RIlRI_CS91aqdQ1atWgAzF_n3YtJiWy24TuDcs&s=eRh0NXZEqMZXqxE-4jyiDQY0PD5ZudNd8U8RqDg6zqA&e=>.
|
|
I actually meant @mathomp4, could you handle the release and dependency updates in config files? |
|
I have released a new version of MOM5 with all the bug fixes since the last. I've updated the I also made GEOS-ESM/GEOSgcm#153 which updates |
|
@mathomp4 Thanks Matt. It still needs the dependency on GEOS-ESM/MAPL#405 to be satisfied before it can build. |
There was a problem hiding this comment.
@yvikhlya
Just 3 comments/requests:
1. Are you all using a particular notation for dual ocean variables,
e.g., in GEOSogcm_GridComp/GEOS_OgcmGridComp.F90
Things like:
- DST_NAME = (/'FRACICEd'/), &
- SRC_ID = SEAICEd,
with adin the end?
If so, please state that explicitly in the preamble. That will help a long ways!
2.1 In GEOSogcm_GridComp/GEOSocean_GridComp/GuestOcean_GridComp/GuestOcean_GridComp.F90
The saltrestore, will not work with MOM6 since we are not exporting DH, hence DH(:,:,1) isn't available. But then I see no reason, why this block cannot be moved into MOM5_plug.
FYI: MOM6 salt flux restore https://github.com/sanAkel/MOM6/blob/17339577cec87bace340f46515155a80eaaf20f1/config_src/coupled_driver/MOM_surface_forcing.F90#L346
For now, would you please move this block from Guest ocean to MOM5 plug. Thanks!
2.2 In calculation of DEL_TMP, restoring temperature difference, please put in a comment with what the value of -0.054 is with its units. @zhaobin74
Question
What the addition in GEOS_GcmGridComp.F90 do??
Was it needed by MAPL2? Or something else?
@yvikhlya It's possible you might have to push one more change. One of the MAPL calls that @atrayano changed for you might need you to use keyword arguments so one of them can become optional. |
|
|
@yvikhlya I already had sent you a Team's chat message with instructions what you need to change. Here it is again: doRecord = MAPL_RecordAlarmIsRinging(MAPL, MAPL_Write2Disk, RC=status) to doRecord = MAPL_RecordAlarmIsRinging(MAPL, MODE=MAPL_Write2Disk, RC=status) |
requested by Atanas.
|
MAPL 2.1.5 has been released. |
Will you update components.yaml? |
Done! I've updated the branch on GEOSgcm to have MAPL 2.1.5 in all the files. |
|
Note: CircleCI will fail because the CI isn't as smart as it should be. The reason is that it clones GEOSgcm first. But it doesn't know it needs a different branch of the fixture to get the CI to pass. If the GEOSgcm was updated to have MAPL 2.1.5 and mom 1.0.2, it would build and pass. |
|
@mathomp4 correct, but for testing, the branch should be to be up-to-date with develop |
Yes. For testing, you'll want to first do: after cloning the fixture. Then you can do a |
|
@yvikhlya my tests do not show this PR is 0-diff. In both amip and replay mode, openwater_internal_checkpoint has different values for the SSKINW field ...
|
…diff with use_skin_layer=1 and affected other computations in coupled mode.
…_GridComp into feature/yury/s2s3-merge
|
@sdrabenh I amended changes to GEOS_OpenWaterGridComp.F90 which caused non-zero diff under condition when use_skin_layer=1. We will deal with SSKINW computation later in separate pull request. |
yvikhlya
removed
the
Contingent - DNA
This PR has a number of zero diff updates from S2S3 tag in CVS. It depends on GEOS-ESM/MAPL#405 and GEOS-ESM/MOM5#14. @sanAkel please review GEOS_SurfGridComp.F90 and GEOS_OpenWaterGridComp.F90. Updates in GEOS_GcmGridComp.F90, GEOS_OgcmGridComp.F90 and GuestOcean_GridComp.F90 affect dual ocean mode. Updates to MOM_GEOS5PlugMod.F90 fix export of mixed layer depth from MOM5 and record restart in dual ocean mode.