search

WCRP-CMIP / CMIP6_CVs

Controlled Vocabularies (CVs) for use in CMIP6
Creative Commons Attribution 4.0 International
155 stars 78 forks source link

source_id registration for multiple CNRM configurations #115

Closed durack1 closed 7 years ago

durack1 commented 7 years ago

Thanks David, can you please complete the following template for each of the 4 new model configurations: CNRM-CM6-1, CNRM-CM6-1-HR, CNRM-ESM2-1, CNRM-ESM2-1-HR. I have provided an example of a model description below:

"ACCESS-1-0":{
"aerosol":"CLASSIC (v1.0)",
"atmosphere":"HadGAM2 (r1.1; 192 x 145 N96; 38 levels; top level 39255m)",
"atmospheric_chemistry":"None",
"cohort":[
    "CMIP5" # This field will be populated by the WCRP Infrastructure Panel
],
"institution_id":[
    "CSIRO-BOM"
],
"label":"ACCESS 1.0", # limited to 16 characters
"label_extended":"ACCESS 1.0", # This entry is free text for users to contribute verbose information
"land_ice":"None",
"land_surface":"MOSES2.2",
"ocean":"ACCESS-OM (MOM4p1; tripolar primarily 1deg latitude/longitude; 50 levels; top grid cell 0-10m)",
"ocean_biogeochemistry":"None",
"release_year":"2011",
"sea_ice":"CICE4.1",
"source_id":"ACCESS-1-0"
}
durack1 commented 7 years ago 
From: SENESI Stéphane
Date: Thursday, November 3, 2016 at 9:12 AM
To: Paul Durack
Subject: Re: CMIP6 contributing institution and model details request

Hi Paul

Thanks for the URLs. I read in https://github.com/WCRP-CMIP/CMIP6_CVs/issues/new :
    'institution_id' -- list all institutions (by institution_id) who are responsible for one or more CMIP6 simulations with this model version. 
where a list is an alternate choice w.r.t. to what you quote, namely "each of the centre acronyms hyphenated".

Is it actually a choice, or should a list of acronyms be hyphenated ?

As an aside, please note that CNRS-Cerfacs means : "Cerfacs, a lab associated with french CNRS (the Centre National de la Recherche Scientifique)" while 'CNRM-Cerfacs' means : a joint venture between CNRM (Centre National de Recherches Météorologiques) and Cerfacs. I am with CNRM.

S
durack1 commented 7 years ago

Stephane, that is up to you, so you in the institution_id file you could have "CNRM-Cerfacs", or you could list both "CNRM" and "Cerfacs" as separate institutions (if they were), and then in the source_id file (where each of the models are described), we would then list each of the contributing institutions in a list, so [“CNRM”, “Cerfacs”] and in the directory structure the first entry from this list would be used.

senesis commented 7 years ago

Any issue with using "CNRM-CERFACS" as source_id and ["CNRM", "CERFACS" ] as institution_id (provided all three are registered) ? This seems to be in line with guidelines at https://github.com/WCRP-CMIP/CMIP6_CVs/issues/new regarding institution_id (namely "_list all institutions (by institutionid) who are responsible for one or more CMIP6 simulations with this model version."). And, by the way, there is no explanation there for semantics of entry 'source_id'

durack1 commented 7 years ago

@senesis the model names (or source_id values) are already CNRM-CM6-1, CNRM-CM6-1-HR, CNRM-ESM2-1, CNRM-ESM2-1-HR, so what we're talking about here is the institution_id, the host institution that developed these models.

So if you like we can register the institution_id and follow the example of CSIR-CSIRO, so:

"CNRM-CERFACS":"CNRM (Centre National de Recherches Météorologiques, XXX, XXX, France), CERFACS (XXX, XXX, XXX, France)"

And you can provide the XXX fields that I am unsure about

taylor13 commented 7 years ago

@senesis @durack1 I think there may remain some confusion. We agree that "source_id" is supposed to record the name (or "nickname") of the model. The institution_id that appears in the source_id CV should name the institution(s) responsible for the model's simulations. If several institutions form a consortium, federation, partnership, etc., and together they are responsible for a simulation, the group of institutions should be recorded as in a single institution_id (e.g. CNRM-CERFACS). In some cases an institution will acquire a model developed elsewhere and run CMIP6 experiments completely independent of the model originator. Suppose, for example, both the UKMO and the NIMR-KMA institutions independently carry out simulations with HadCM3. Then each of these two institutions should be registered separately in the "institution_id" CV, but for source_id=HadCM3, the institution_id entry should read:
"institution_id":[ "UKMO", "NIMR-KMA" ],

because this entry must include every institution that independently performs one or more simulations with the model. (Note that NIMR-KMA represents a partnership between NIMR and KMA, and the partnership is represented by a single compound institution). Karl

senesis commented 7 years ago

Thanks Paul and Karl. Semantics of source_id was obvious for you, not for me. We will go joint (CNRM-CERFACS). I will create an issue for that. And I will complete the source_id issue next week.

taylor13 commented 7 years ago

thanks for fighting through the weeds. Please let us know as other questions come up.

durack1 commented 7 years ago

@senesis there is no need to create another issue, are you happy with the below for institution_id:

"CNRM-CERFACS": "CNRM (Centre National de Recherches Meteorologiques, Toulouse, 31057, France), CERFACS (Centre Europeen de Recherche et de Formation Avancee en Calcul Scientifique, Toulouse 31100, France)"
durack1 commented 7 years ago

@senesis we have now finalised your institution_id: CNRM-CERFACS

It would be great to get the model configurations source_id values also filled out so we can close this open issue - for a peruse of the currently registered models, see here

senesis commented 7 years ago

While the text below shows without any "e acute" in Meteorologiques, the text I received by email did show 2 'e acute'. This is representative of the kind of issue we still may have with coding systems. We can just give up with any accent and use plain 'e's I also modified 'CERFACS' to 'and CERFACS' The source_id entries will follow today Thanks , Paul "CNRM-CERFACS": "CNRM (Centre National de Recherches Meteorologiques, Toulouse, 31057, France), and CERFACS (Centre Europeen de Recherche et de Formation Avancee en Calcul Scientifique, Toulouse 31100, France)"

senesis commented 7 years ago

Here are the entries for CNRM CMs and ESMs :

""CNRM-ESM2-1-HR":{ "aerosol":"TACTIC (v2.0)", "atmosphere":"Arpege 6.2 (Tl359; 91 levels; top level 78.4 km)", "atmospheric_chemistry":"None", "cohort":[ "CMIP5" ], "institution_id":[ "CNRM-CERFACS" ], "label":"CNRM-ESM2-1-HR", "label_extended":"CNRM-ESM2-1-HR", "land_ice":"None", "land_surface":"Surfex 8.0c", "ocean":"Nemo 3.6 (eORCA025; tripolar primarily 1/4deg latitude/longitude; 75 levels; top grid cell 0-1 m)", "ocean_biogeochemistry":"Pisces 2.s", "release_year":"2016", "sea_ice":"Gelato 6.1", "source_id":"CNRM-ESM2-1-HR" }

"CNRM-ESM2-1":{ "aerosol":"TACTIC (v2.0)", "atmosphere":"Arpege 6.2 (Tl127; 91 levels; top level 78.4 km)", "atmospheric_chemistry":"REPROBUS-C (v2.0)", "cohort":[ "CMIP5" ], "institution_id":[ "CNRM-CERFACS" ], "label":"CNRM-ESM2-1", "label_extended":"CNRM-ESM2-1", "land_ice":"None", "land_surface":"Surfex 8.0c", "ocean":"Nemo 3.6 (eORCA1; tripolar primarily 1deg latitude/longitude; 75 levels; top grid cell 0-1 m)", "ocean_biogeochemistry":"Pisces 2.s", "release_year":"2016", "sea_ice":"Gelato 6.1", "source_id":"CNRM-ESM2-1" }

"CNRM-CM6-1-HR":{ "aerosol":"climatological fields computed by TACTIC (v2.0)", "atmosphere":"Arpege 6.2 (Tl359; 91 levels; top level 78.4 km)", "atmospheric_chemistry":"None", "cohort":[ "CMIP5" ], "institution_id":[ "CNRM-CERFACS" ], "label":"CNRM-CM6-1-HR", "label_extended":"CNRM-CM6-1-HR", "land_ice":"None", "land_surface":"Surfex 8.0c", "ocean":"Nemo 3.6 (eORCA025; tripolar primarily 1/4deg latitude/longitude; 75 levels; top grid cell 0-1 m)", "ocean_biogeochemistry":"None", "release_year":"2016", "sea_ice":"Gelato 6.1", "source_id":"CNRM-CM6-1-HR" }

"CNRM-CM6-1":{ "aerosol":"climatological fields computed by TACTIC (v2.0)", "atmosphere":"Arpege 6.2 (Tl127; 91 levels; top level 78.4 km)", "atmospheric_chemistry":"None", "cohort":["CMIP5"], "institution_id":[ "CNRM-CERFACS"], "label":"CNRM-CM6-1", "label_extended":"CNRM-CM6-1", "land_ice":"None (except for some FAFMIP experiments)", "land_surface":"Surfex 8.0c", "ocean":"Nemo 3.6 (eORCA1; tripolar primarily 1deg latitude/longitude; 75 levels; top grid cell 0-1 m)", "ocean_biogeochemistry":"None", "release_year":"2016", "sea_ice":"Gelato 6.1", "source_id":"CNRM-CM6-1" }

durack1 commented 7 years ago

@senesis apologies for missing this one, it's been sitting in the queue for sometime. These changes have now been made and you should be able to view your entries at CMIP6_source_id.html - if any other tweaks are required, please reopen this issue and indicate any required changes

senesis commented 7 years ago

Could you please modifiy the atmospheric chemistry component name from None to 'OZL_v2' in the 3 entries where it applies (namely CNRM-CM6-1, CNRM-CM6-1-HR and CNRM-ESM2-1-HR) ?

durack1 commented 7 years ago

@senesis that's now done, please take a peek at CMIP6_source_id.html to double check the results

durack1 commented 7 years ago

@senesis we have recently augmented the information we are collecting for registration, so we require some additional information from you. I have made some amendments to the existing entries below (noted between the ### entry ### text), and will require you to fill out the fields with the ### informational text directing you to descriptive information about the request, or lists of potential entries (I have only included this for the first model CNRM-CM6-1 as an example)

"CNRM-CM6-1":{
   "activity_participation":[
      "" ### See activity list at https://github.com/WCRP-CMIP/CMIP6_CVs/blob/master/CMIP6_activity_id.json
   ],
"aerosol":
   "description":"climatological fields computed by TACTIC (v2.0)",
   "nominal_resolution":"" ### See Appendix 2 in https://goo.gl/v1drZl (likely same as atmos below)
"atmos":
   "description":"Arpege 6.2 (Tl127; ### XXX x XXX longitude/latitude###; 91 levels; top level 78.4 km)",
   "nominal_resolution":"" ### See Appendix 2 in https://goo.gl/v1drZl
"atmosChem":
   "description":"OZL_v2",
   "nominal_resolution":"" ### See Appendix 2 in https://goo.gl/v1drZl (likely same as atmos above)
"land":
   "description":"Surfex 8.0c",
   "nominal_resolution":""### See Appendix 2 in https://goo.gl/v1drZl (likely same as atmos above)
"landIce":
   "description":"none (except for some FAFMIP experiments)",
   "nominal_resolution":"none"
"ocean":
   "description":"Nemo 3.6 (eORCA1; tripolar primarily 1deg###; 360 x 180### longitude/latitude; 75 levels; top grid cell 0-1 m)",
   "nominal_resolution":""### See Appendix 2 in https://goo.gl/v1drZl
"seaIce":
   "description":"Gelato 6.1",
   "nominal_resolution":""### See Appendix 2 in https://goo.gl/v1drZl (likely same as ocean above)

"CNRM-CM6-1-HR":{
   "activity_participation":[
      ""
   ],
"aerosol":{
   "description":"climatological fields computed by TACTIC (v2.0)",
   "nominal_resolution":""
"atmos":{
   "description":"Arpege 6.2 (Tl359; ### XXX x XXX longitude/latitude###; 91 levels; top level 78.4 km)",
   "nominal_resolution":""
"atmosChem":{
   "description":"OZL_v2",
   "nominal_resolution":""
"land":{
   "description":"Surfex 8.0c",
   "nominal_resolution":""
"ocean":{
   "description":"Nemo 3.6 (eORCA025; tripolar primarily 1/4deg###; 1440 x 720### longitude/latitude; 75 levels; top grid cell 0-1 m)",
   "nominal_resolution":""
"seaIce":{
   "description":"Gelato 6.1",
   "nominal_resolution":""

"CNRM-ESM2-1":{
   "activity_participation":[
      ""
   ],
"aerosol":
   "description":"TACTIC (v2.0)",
   "nominal_resolution":""
"atmos":{
   "description":"Arpege 6.2 (Tl127; ### XXX x XXX longitude/latitude###; 91 levels; top level 78.4 km)",
   "nominal_resolution":""
"atmosChem":{
   "description":"REPROBUS-C (v2.0)",
   "nominal_resolution":""
"land":{
   "description":"Surfex 8.0c",
   "nominal_resolution":""
"ocean":{
   "description":"Nemo 3.6 (eORCA1; tripolar primarily 1deg###; 360 x 180### longitude/latitude; 75 levels; top grid cell 0-1 m)",
   "nominal_resolution":""
"ocnBgchem":{
   "description":"Pisces 2.s",
   "nominal_resolution":""
"seaIce":{
   "description":"Gelato 6.1",
   "nominal_resolution":""

"CNRM-ESM2-1-HR":{
   "activity_participation":[
      ""
   ],
"aerosol":{
   "description":"TACTIC (v2.0)",
   "nominal_resolution":""
"atmos":{
   "description":"Arpege 6.2 (Tl359; ### XXX x XXX longitude/latitude###; 91 levels; top level 78.4 km)",
   "nominal_resolution":""
"atmosChem":{
   "description":"OZL_v2",
   "nominal_resolution":""
"land":
   "description":"Surfex 8.0c",
   "nominal_resolution":""
"ocean":
   "description":"Nemo 3.6 (eORCA025; tripolar primarily 1/4deg###; 1440 x 720### longitude/latitude; 75 levels; top grid cell 0-1 m)",
   "nominal_resolution":""
"ocnBgchem":{
   "description":"Pisces 2.s",
   "nominal_resolution":""
"seaIce":
   "description":"Gelato 6.1",
   "nominal_resolution":""
durack1 commented 7 years ago

@senesis have you had any luck collecting the information requested above?

svalcke commented 7 years ago

Hi, Stéphane being on holidays, he ask me to take over this ticket. Here below is the information for the different CNRM-CM6 configurations. Also we have the following additional requests: • In CMIP6_institution_id.json, in « institution_id » for « CNRM-CERFACS », could you please change « Toulouse 31100 » for « Toulouse 31057 » • In CMIP6_source_id.json, for «CNRM-CM6-1 », «CNRM-CM6-1-HR », « CNRM-ESM2-1 », and « CNRM-ESM2-1-HR », could you please change « release_year » from « 2016 » to « 2017

Finally for the source_id information. Please note that we changed more details than asked for, see the attached Word file in track change mode for details. CNRM_CMIP6_CV.docx :

"CNRM-CM6-1":{
   "activity_participation":["ScenarioMIP",  "DCPP", "LS3MIP",  "GMMIP",  "CFMIP",  "DAMIP",  "FAFMIP", "ISMIP6", "PMIP", "RFMIP", ],
"aerosol":
   "description":"prescribed monthly fields computed by TACTIC_v2 scheme",
   "nominal_resolution":"250 km" 
"atmos":
   "description":"Arpege 6.3 (Tl127; 24572 x 1; 91 levels; top level 78.4 km)",
   "nominal_resolution":"250 km"
"atmosChem":
   "description":"OZL_v2",
   "nominal_resolution":"250 km" 
"land":
   "description":"Surfex 8.0c",
   "nominal_resolution":"250 km"
"landIce":
   "description":"none",
   "nominal_resolution":"none"
"ocean":
   "description":"Nemo 3.6 (eORCA1; tripolar primarily 1deg; 362 x 294 ; 75 levels; top grid cell 0-1 m)",
   "nominal_resolution":"100 km
"seaIce":
   "description":"Gelato 6.1",
   "nominal_resolution":"100 km"

"CNRM-CM6-1-HR":{
   "activity_participation":["ScenarioMIP",  "DCPP", "HighResMIP"
   ],
"aerosol":{
   "description": "prescribed monthly fields computed by TACTIC_v2 scheme",
   "nominal_resolution":"100 km"
"atmos":{
   "description":"Arpege 6.3 (Tl359; 181724 x 1; 91 levels; top level 78.4 km)",
   "nominal_resolution":"100 km"
"atmosChem":{
   "description":"OZL_v2",
   "nominal_resolution":"100 km"
"land":{
   "description":"Surfex 8.0c",
   "nominal_resolution":"100 km"
"ocean":{
   "description":"Nemo 3.6 (eORCA025; tripolar primarily 1/4deg ; 1442 x 1050; 75 levels; top grid cell 0-1 m)",
   "nominal_resolution":"25 km"
"seaIce":{
   "description":"Gelato 6.1",
   "nominal_resolution":"25 km"

"CNRM-ESM2-1":{
   "activity_participation":["ScenarioMIP", "GeoMIP" ,"OMIP" ,"AerChemMIP","C4MIP", "LUMIP"],
"aerosol":
   "description":"TACTIC_v2",
   "nominal_resolution":"250 km"
"atmos":{
   "description":"Arpege 6.3 (Tl127; 24572 x 1; 91 levels; top level 78.4 km)",
   "nominal_resolution":"250 km"
"atmosChem":{
   "description":"REPROBUS-C_v2",
   "nominal_resolution":"250 km"
"land":{
   "description":"Surfex 8.0c",
   "nominal_resolution":"250 km"
"ocean":{
   "description":"Nemo 3.6 (eORCA1; tripolar primarily 1deg###; 362 x 294; 75 levels; top grid cell 0-1 m)",
   "nominal_resolution":"100 km"
"ocnBgchem":{
   "description":"Pisces 2.s",
   "nominal_resolution":"100 km"
"seaIce":{
   "description":"Gelato 6.1",
   "nominal_resolution":"100 km"

"CNRM-ESM2-1-HR":{
   "activity_participation":["ScenarioMIP","OMIP"
],
"aerosol":{
   "description":"TACTIC_v2",
   "nominal_resolution":"100 km"
"atmos":{
   "description":"Arpege 6.3 (Tl359; 181724 x 1; 91 levels; top level 78.4 km)",
   "nominal_resolution":"100 km"
"atmosChem":{
   "description":"REPROBUS-C_v2",
   "nominal_resolution":"100 km"
"land":
   "description":"Surfex 8.0c",
   "nominal_resolution":"100 km"
"ocean":
   "description":"Nemo 3.6 (eORCA025; tripolar primarily 1/4deg ; 1442 x 1050; 75 levels; top grid cell 0-1 m)",
   "nominal_resolution":"25 km"
"ocnBgchem":{
   "description":"Pisces 2.s",
   "nominal_resolution":"25 km"
"seaIce":
   "description":"Gelato 6.1",
   "nominal_resolution":"25 km"
durack1 commented 7 years ago

@svalcke @senesis can you please take a look at the CMIP6_source_id.json entry in the pull request and markup any additional tweaks required. To conform to the format I have made some amendments to the submitted info, in particular the XXX x XXX longitude/latitude information needs to be worked out. I have also included the CMIP activity for each registered model, as I believe you will be contributing a piControl and other experiments for each model as well?

svalcke commented 7 years ago

Hi Paul,

I check your pull request and everything seems fine to me besides the following: -For +key = 'CNRM-CM6-1’, why do you have "# Ice sheet MIP?" besides + ‘ISMIP6’? -For “ocnBgchem” for "description":"Pisces 2.s”, the "nominal_resolution":”” should be the same as the “ocean” “nominal_resolution”, i.e 100 km for CNRM-ESM2-1 and 25 km for CNRM-ESM2-1-HR. -In different places, I see a "XXX x XXX longitude/latitude” as you also mention. I would be happy to provide this single correct value if I understood what is expected here. Can you explain? If it is the "number of grid points in longitude"x ""number of grid points in latitude” it simply cannot be specified for our atmosphere Gaussien reduced grid which is unstructured. And even if you kept something for the ocean, e.g. "362 x 294 longitude/latitude”, it does not strictly apply either as the grid is logically rectangular but stretched and distorted.

With best regards, Sophie

On 06 Jun 2017, at 21:22, Paul J. Durack notifications@github.com wrote:

@svalcke @senesis can you please take a look at the source_id.json entry in the pull request and markup any additional tweaks required. To conform to the format I have made some amendments to the submitted info, in particular the XXX x XXX longitude/latitude information needs to be worked out

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

Sophie Valcke

Research Engineer Climate Modelling and Global Change team CECI UMR 5318 CNRS/CERFACS 42 Av. G. Coriolis 31057 Toulouse Cedex 01, FRANCE Tel: 05.61.19.30.76 Fax: 05.61.19.30.00

durack1 commented 7 years ago

@svalcke thanks for responding, you had three queries I'll address each in turn.

  1. For ISMIP6 my understanding is that this would require a valid landIce model, and for CNRM-CM6-1 there is no landIce component specified. So am I correct in querying your contribution to ISMIP6?

  2. An ocnBgchem component was erroneously excluded for CNRM-ESM2-1, this entry has been correctly added

  3. We are attempting to standardize the format of the model component descriptions, and the total cell count in longitude/latitude was one way to do this. Exactly as you assumed, the total number of grid points in longitude and total number of grid points in latitude. If you have a stretched grid then I would suggest including some more descriptive text such as primarily 1deg latitude/longitude, down to 1/3deg within 30deg of the equatorial tropics; 362 x 294 longitude/latitude; like in the BNU-ESM-1-1 description (you can see examples of these in the CMIP6_source_id.html). In the case of an unstructured or alternative grid, VRESM-1-0 is using a cubic-conformal model atmospheric described as VCAM-1.0 (192 x 192 x 6 C192; 35 levels; top level 35km) and with some amendments we're working on I hope to expand this with a more verbose descriptor including the words "cubic-conformal". For an unstructured grid, AWI-CM-1-0-HR includes the following for the model ocean FESOM 1.4 (unstructured grid in the horizontal with 1306775 wet nodes; 46 levels; top grid cell 0-5 m). I hope these existing examples provide you enough information to tweak the entries above

davidhassell commented 7 years ago

Hello,

My name is David Hassell and I am part of the ES-DOC team that is working on CMIP6 documentation. As a part of this process, we need a liaison at CNRM-CERFACS who can act as the interface between the scientists who actually provide the documentation and the ES-DOC team. I was wondering if you are able to help with identifying such a person?

The documentation process has changed considerably since CMIP5 (in response to the feedback that we received), and you can now choose from a variety of documentation methods. Our aim is to train the ES-DOC liaison who can then work with people at CNRM-CERFACS. No specialist knowledge about ES-DOC nor climate modelling is required, but neither would it be a problem, of course. Some institutes have appointed a member of the modelling team, others not.

I would be very grateful f you could let me know by e-mail (david.hassell at ncas.ac.uk) if you have any suggestions.

All the best,

David Hassell

National Centre for Atmospheric Science University of Reading UK

svalcke commented 7 years ago

Dear David,

Our group CNRM-CERFACS already identified such a person as Laurent Franchisteguy. Laurent sent a mail confirming this to Eric Guilyardi and yourself on May 5th. Here it is again in case you did not receive it at the time …

With best regards, Sophie

From: FRANCHISTEGUY Laurent laurent.franchisteguy@meteo.fr Subject: Re: CMIP6 documentation review phase 1 until May 5th Date: 5 May 2017 17:12:45 GMT+2 To: eric.guilyardi@locean-ipsl.upmc.fr, david.hassel@ncas.ac.uk Cc: SALAS Y MELIA David david.salas@meteo.fr, SENESI Stéphane stephane.senesi@meteo.fr, FRANCHISTEGUY Laurent laurent.franchisteguy@meteo.fr

Dear Eric and David,

Please find below the answer of CNRM-CERFACS group for your request :

1) Review of white paper 'Formal CMIP6 documentation'

  • the group really appreciates the quality of the document and your aim to simplify and streamline the process in order to minimise the contribution from the modelling groups
  • just one minor remark concerning possible redundancy and consistency issue between the realms 'Aerosols', 'Atmospheric chemistry' and 'Atmosphere'.

2) Science review of model realms documentation

  • members of our group (contributors : M. Chevallier, C. Delire, R. Roehrig, R. Séférian and D. Salas) have already filled feedback forms for the realms 'land surface', 'ocean-biogeochemistry', 'ocean' and 'sea-ice'
  • concerning the description of models, the lack of information on coupling aspects (eg ocean/atmosphere, ocean/seaice, seaice/atmosphere) is pinpointed
  • we've a specific remark on the 'specialisation' atmosphere and the associated mindmap. This mindmap is already very complete but we're not sure that it will cover our model description. If you are willing to add new items to cover all model descriptions, we would need more time to describe our model (~6 months). If not, then we already now that our model description won't generally match the last level and that we won't probably provide this level of information.

3) ES-DOC officer

ES-DOC aspects will be covered in our group by S. Sénési and myself. Thus, the contact point will be Laurent Franchistéguy (laurent.franchisteguy@meteo.fr).

Best regards, Laurent

FRANCHISTEGUY Laurent CNRM, Météo France/CNRS, UMR 3589 CNRM/GMGEC/EST Tél. +33 5 61 07 97 47 , http://www.umr-cnrm.fr/

On 08 Jun 2017, at 09:54, David Hassell notifications@github.com wrote:

Hello,

My name is David Hassell and I am part of the ES-DOC team that is working on CMIP6 documentation. As a part of this process, we need a liaison at CNRM-CERFACS who can act as the interface between the scientists who actually provide the documentation and the ES-DOC team. I was wondering if you are able to help with identifying such a person?

The documentation process has changed considerably since CMIP5 (in response to the feedback that we received), and you can now choose from a variety of documentation methods. Our aim is to train the ES-DOC liaison who can then work with people at CNRM-CERFACS. No specialist knowledge about ES-DOC nor climate modelling is required, but neither would it be a problem, of course. Some institutes have appointed a member of the modelling team, others not.

I would be very grateful f you could let me know by e-mail (david.hassell at ncas.ac.uk) if you have any suggestions.

All the best,

David Hassell

National Centre for Atmospheric Science University of Reading UK

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

Sophie Valcke

Research Engineer Climate Modelling and Global Change team CECI UMR 5318 CNRS/CERFACS 42 Av. G. Coriolis 31057 Toulouse Cedex 01, FRANCE Tel: 05.61.19.30.76 Fax: 05.61.19.30.00

svalcke commented 7 years ago

@durack1 For 1. we will indeed participate to ISMIP6 with CNRM-CM6 but collaborating with l’Institut des Géosciences de l’Environnement in Grenoble that use the Grisli (http://www.cmcc.it/wp-content/uploads/2015/02/rp0249-oda-01-2015.pdf) land ice model: 1 - we will realise the 1% CO2 & SSP85 simulations with CNRM-CM6 2 - we will provide the resulting forcings to IGE 3 - IGE will realise forced experiments taking into account these forcings and will diagnose the water flux linked to the land ice melting 4- we will use these water flux with CNRM-CM6 and will redo 1% CO2 & SSP85 simulations I am not sure how (and if) we should describe all that, any advice is welcome.

For 2. The ocnBgchem component was not erroneously excluded for CNRM-ESM2-1, it was only the “nominal_resolution” 100 km for CNRM-ESM2-1 and 25 km for CNRM-ESM2-1-HR that was missing.

For 3., for "atmos" "description" I answered directly in your pull request ( https://github.com/WCRP-CMIP/CMIP6_CVs/pull/340/files_) i.e. they should be for CNRM-CM6-1 and CNRM-ESM2-1: "description":"Arpege 6.3, T127, Gaussian Reduced with 24572 grid points in total distributed over 128 latitude circles (with 256 grid points per latitude circle between 30degN and 30degS reducing to 20 grid points per latitude circle at 88.9degN and 88.9degS)" or for for CNRM-CM6-1-HR and CNRM-ESM2-1-HR: "description":"Arpege 6.3, T359, Gaussian Reduced with 181724 grid points in total distributed over 360 latitude circles (with 720 grid points per latitude circle between 33.2degN and 32.2degS reducing to 18 grid points per latitude circle at 89.6degN and 89.6degS)"

For 3. for "ocean" "description" it should be for CNRM-CM6-1 and CNRM-ESM2-1: "description":"Nemo 3.6 (eORCA1, tripolar 362 x 294; primarily 1deg longitude/latitude; 75 levels; top grid cell 0-1 m)" or for CNRM-CM6-1-HR and CNRM-ESM2-1-HR: "description":"Nemo 3.6 (eORCA025, tripolar 1442 x 1050; primarily 1/4deg longitude/latitude; 75 levels; top grid cell 0-1 m)"

durack1 commented 7 years ago

@svalcke @senesis thanks for persisting with this, peek at CMIP6_source_id.html and let me know if any further tweaks are required. I amended some of your suggestions above to maintain the same format in the description fields, we can also augment the landIce description for CNRM-CM6-1 if you think I have not summarized the relevant info appropriately

svalcke commented 7 years ago

Everything is OK but for:

1) For CNRM-CM6-1 and CNRM-ESM2-1, for the ocean, you put Nemo 3.6 (eORCA1, tripolar primarily 1deg; 362 x 294 longitude/latitude; 75 levels; top grid cell 0-1 m) where as I proposed : "Nemo 3.6 (eORCA1, tripolar 362 x 294; primarily 1deg longitude/latitude; 75 levels; top grid cell 0-1 m)" Again I insist that describing the grid as 362x294 longitude/latitude is wrong as the grid is stretched.

2) For CNRM-CM6-1-HR and CNRM-ESM2-1-HR, for the ocean, you put: Nemo 3.6 (eORCA025, tripolar primarily 1/4deg; 1442 x 1050 longitude/latitude; 75 levels; top grid cell 0-1 m) where as I proposed : "Nemo 3.6 (eORCA025, tripolar 1442 x 1050; primarily 1/4deg longitude/latitude; 75 levels; top grid cell 0-1 m)" Again I insist that describing the grid as 1442 x 1050 longitude/latitude is wrong as the grid is stretched.

3) For CNRM-CM6-1, the land ice model is GRISLI, not GRISILI

4) For CNRM-CM6-1-HR and CNRM-ESM2-1-HR, for the atmos, please change “...between 33.2degN and 32.2degS …” for “… between 32.2degN and 32.2degS …”

On 14 Jun 2017, at 20:12, Paul J. Durack notifications@github.com wrote:

@svalcke @senesis thanks for persisting with this, peek at CMIP6_source_id.html and let me know if any further tweaks are required. I amended some of your suggestions above to maintain the same format in the description fields, we can also augment the landIce description for CNRM-CM6-1 if you think I have not summarized the relevant info appropriately

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

Sophie Valcke

Research Engineer Climate Modelling and Global Change team CECI UMR 5318 CNRS/CERFACS 42 Av. G. Coriolis 31057 Toulouse Cedex 01, FRANCE Tel: 05.61.19.30.76 Fax: 05.61.19.30.00

durack1 commented 7 years ago

@svalcke thanks for checking.

We are attempting to standardize the entries so that the grid size 362 x 294 longitude/latitude is a text block that can be easily parsed. As an alternative entry, and allowing you to keep the standard entry format we could follow the BESM-2-7 exampleMOM-5 (MOM5, tripolar primarily 1 deg, 1/4 deg between 10S-10N; 360 x 300 longitude/latitude; 50 levels; top grid cell 0-10 m). For other examples see CMIP6_source_id.html

Apologies for the GRISLI typo, will fix

I will also change the 33.2 to 32.2degN

I will await your new ocean descriptor text before I finalize all these changes.

senesis commented 7 years ago

We prefer to have a realistic description, rather than an easily parsable one.

durack1 commented 7 years ago

@senesis I don't believe the goals are mutually exclusive. The BESM-2-7 example above provides comprehensive description of a stretched grid which scales from 1deg to 1/4deg as @svalcke describes above while maintaining the parseable ; 360 x 300 longitude/latitude; string

durack1 commented 7 years ago

@senesis @svalcke how about this (or a corrected version of this) for CNRM-CM6-1 and CNRM-ESM2-1:

Nemo 3.6 (eORCA1; tripolar primarily 1 degree lon/lat reducing to 1/3 degree between 20degN and 20degS; 362 x 294 longitude/latitude; 75 vertical levels; top grid cell 0-1 m)

Would that format suit?

If so, we can also provide a similar description for CNRM-CM6-1-HR and CNRM-ESM2-1-HR

svalcke commented 7 years ago

We think the description "primarily 1 degree lon/lat reducing to 1/3 degree between 20degN and 20degS" is not right, at least not for our grid (there are other latitudes where the grid is refined and the refinement at the equator is effective between 3degS et 3degN). If we cannot avoid the (misleading) "362 x 294 longitude/latitude" formulation, than we prefer your original proposition, i.e:

durack1 commented 7 years ago

@svalcke @senesis your model registrations are now revised, take a look at CMIP6_source_id.html and let me know if any further changes are required

I guessed at the GRISLI nominal_resolution = 250 km let me know if we need to further update this entry

senesis commented 7 years ago

Sorry for a further change, Paul : Grisli's nominal resolution is 15km

svalcke commented 7 years ago

I confirm that the rest, i.e. besides Grisli nominal resolution that should be 15 km, is fine. Thanks!

durack1 commented 7 years ago

@svalcke @senesis thanks for this, I've amended the GRISLI nominal_resolution to 10 km which is the closest registered value in the nominal_resolution

Happy to further tweak any of these entries if they need tweaking in the coming months..

senesis commented 7 years ago

@durack1 , @taylor13 We discussed at length the component grid description parts for CNRM-CM6. Will any publication stage quality check enforce consistency between this description and data files metadata labeled 'grid_description' ? If yes, what is the exact algortihm for extracting the grid description part from the component description part ?

durack1 commented 7 years ago

@senesis thanks for the query. The component descriptions stored in CMIP6_source_id.json#L751-L784 are converted across to the source identifier in CMIP6_CV.json#L341. If a modeling group is using CMOR along with the cmip6-cmor-table inputs, then this information will be written in the global attributes for each CMOR-written file.

The only validation will be for the first entry in the source attribute, so in this case CNRM-CM6-1 (2017) as you have a registered entry, this will work as expected.

@taylor13 @dnadeau4 please add any additional information if required

senesis commented 7 years ago

@durack1 Paul, in that case, a misunderstanding may lead to bad results (but not in our case, as we are not using CMOR) : We thought you wanted to get the description of the model compute grid, while CMOR will use it as a description of the model output publication grid. In some cases (e.g. unstructured compute grid), data producers publish only on a grid which is more amenable to the users (as a regular grid with similar space resolution). This is our case (but, again, we do not use CMOR)

durack1 commented 7 years ago

@senesis the registration information is being used by a number of different components (ES-DOCs for e.g.), with the contributed information provide the "first pass" entries, that can be updated/corrected or augmented within the ES-DOC system. The intention of this information was to document the running model configuration, not the specification of the provided data. For GFDL models, ocean components are running at a nominal_resolution of 0.25 or 0.5 degrees, and 75 vertical levels, however, these will be contributed on a spherical 1 degree grid with 33 vertical levels. PrePARE will be used within the ESGF publisher to validate contributed data, and will use a subset of the registered source_id information to "validate" CMIP6 data contributions.

For netcdf global attributes, using CMOR means that much of the global attribute information is curated for a user, with the source information in addition to other required global attributes curated in the CMIP6_CV.json and written to files generated

svalcke commented 7 years ago

@durack1

Just to check: in our files, we now have as global attributes for example:

:grid = "data regridded to a T127 gaussian grid (128x256 latlon) from a native atmosphere T127l reduced gaussian grid" ; and :source = "CNRM-CM6-1 (2017): \naerosol: prescribed monthly fields computed by TACTIC_v2 scheme\natmos: Arpege 6.3 (T127; Gaussian Reduced with 24572 grid points in total distributed over 128 latitude circles (with 256 grid points per latitude circle between 30degN and 30degS reducing to 20 grid points per latitude circle at 88.9degN and 88.9degS); 91 levels; top level 78.4 km)\natmosChem: OZL_v2\nland: Surfex 8.0c\nocean: Nemo 3.6 (eORCA1, tripolar primarily 1deg; 362 x 294 longitude/latitude; 75 levels; top grid cell 0-1 m)\nseaIce: Gelato 6.1" ;

So "grid" describes the grid onto which the data is provided in the file, while the grid description in "source" describes the grid onto which the data is originally produced by the model.

We suppose that this is correct as PrepARE 3.2.5 is happy with our files. Can you confirm?

taylor13 commented 7 years ago

Yes, the source should describe the native grid; the "grid" attribute should describe the grid used in the file. Karl

svalcke commented 7 years ago

Thank you. You can close this issue!