WCRP-CMIP / CMIP6_CVs

Controlled Vocabularies (CVs) for use in CMIP6
Creative Commons Attribution 4.0 International
159 stars 80 forks source link

Update GFDL-AM4 registration to include ALL activity_id entries #904

Closed durack1 closed 4 years ago

durack1 commented 4 years ago

@jgjgfdl it appears that GFDL-AM4 has published RFMIP data on ESGF, but this is not listed in the activity participation field of the model registration.

Are there any other missing *MIPs/activity_id entries that we should also add? Or any other details that require amendment whilst this correction is implemented?

ping @aradhakrishnanGFDL @ceblanton @MartinaSt

Ref https://github.com/WCRP-CMIP/CMIP6_CVs/issues/900#issuecomment-617085459

durack1 commented 4 years ago
From: "Durack, Paul J."
Date: Wednesday, April 22, 2020 at 5:09 PM
To: Jasmin John - NOAA Federal, David Paynter - NOAA Federal
Cc: Kristopher Rand - NOAA Affiliate, Chris Blanton - NOAA Affiliate, Aparna Radhakrishnan - NOAA Affiliate
Subject: Re: RFMIP with AM4?

Thanks folks, that seems like we now have a path forward.

GFDL-CM4 has RFMIP as an activity_id (listed in the activity participation field in https://wcrp-cmip.github.io/CMIP6_CVs/docs/CMIP6_source_id.html, with a total of 8 MIPs/activity_id entries), whereas GFDL-AM4 only lists CMIP.

Jasmin, I think your path forward is exactly the right approach. I will copy this thread back into github and close the issue as this resolution doesn’t involve a change to activity_id entries for GFDL-AM4.

Thanks for the prompt resolution.

P

From: Jasmin John - NOAA Federal
Date: Wednesday, April 22, 2020 at 4:58 PM
To: David Paynter - NOAA Federal
Cc: "Durack, Paul J.", Kristopher Rand - NOAA Affiliate, Chris Blanton - NOAA Affiliate, Aparna Radhakrishnan - NOAA Affiliate, Jasmin John - NOAA Federal
Subject: Re: RFMIP with AM4?

Hi, 

Thanks David.

So we (GFDL) will have to retract the 4 AM4/RFMIP variables, issue an errata, and republish under CM4.
The other AM4 AMIP variables are for the AM4.0 documentation paper and there should be no issue with those (that I am aware of).

Thanks for the quick resolution,
Jasmin

On Wed, Apr 22, 2020 at 7:34 PM David Paynter - NOAA Federal wrote:
Dear Paul, 
Thanks for bringing this to my attention. So it appears that there are 4 RFMIP variables set to AM4 rlu,rld,rsd, rsu. These are runs that were performed with the offline radiation code of CM4 for the RFMIP instantaneous time-slice experiments (rad-irf ). As this is the CM4 radiation code I should have changed the model from AM4 to CM4 to be consistent with how we have named the other experiments. Hopefully, this is an easy fix to make. I am very sorry for any issues this may have /will cause. I notice there are another 69 variables Amip  that have AM4 that are not associated with RFMIP, i am not sure if these need fixing too.  
Best Wishes, 
David 

On Wed, Apr 22, 2020 at 7:06 PM Jasmin John - NOAA Federal wrote:
Hi, 

Thanks Paul.

I'm aware of the data citation issue and we have tried very hard to make sure no data gets served without first having all the necessary bits and pieces (including the CMIP6 CV registration info with the appropriate Model/MIP Activities), in place internally and externally.
However a lot of this is manual, so perhaps a step was missed.

Thanks for offering to paste the thread back into gitHub once we resolve the issue - much appreciated!

Hope you and your family are staying well.

Best,
Jasmin

On Wed, Apr 22, 2020 at 6:45 PM Durack, Paul J. <durack1@llnl.gov> wrote:
Thanks Jasmin, to explain why the update is necessary. Many of the infrastructure pieces (like the data citation service provided by DKRZ) is depending on the MIP/activity_id registration alongside the model to configure the system behind the scenes. As data is published on the ESGF, but there is no registration for a MIP, this link isn’t made, and consequently the service breaks.

There is no problem with revising the CMIP6_CVs registration info, but best to do this BEFORE any data is published.

I’ll copy this thread back in as a comment on the issue (https://github.com/WCRP-CMIP/CMIP6_CVs/issues/904) once we have a resolution, and so we have a trace of this offline/email discussion that led to a change.

Cheers,

P

From: Jasmin John - NOAA Federal
Date: Wednesday, April 22, 2020 at 3:21 PM
To: "Durack, Paul J."
Cc: David Paynter - NOAA Federal, Kristopher Rand - NOAA Affiliate, Chris Blanton - NOAA Affiliate Aparna Radhakrishnan - NOAA Affiliate, Jasmin John - NOAA Federal
Subject: RFMIP with AM4?

Hi Paul, 

Thanks for the note.
I'm only aware of RFMIP associated with CM4, ESM4, and offline models (GRTCODE).

cc'ing David Paynter, Kris, Chris, and Aparna for clarification.

Thanks,
Jasmin