New ScenarioMIP experiments using CMIP5 forcing

[martinjuckes]

Experiments discussed in #805.

[martinjuckes]

@taylor13 , @durack1, @claudiatebaldi : we need to define the data request for these experiments. Paul has provided a partial answer. This raises some questions:

(1) the historical experiment has a very large data request because of its central role as a baseline for many studies. The historical-cmip5 experiment does not have this role, so it does not make sense to request the same volume of data. Groups of variables which we might want to omit are: 3hr, 6hrPlev, 6hrLev, Omon-3d, esmval-3d?

(2) rcp26-cmip5 etc: these are tier 2 scenarioMIP experiments. The Paul's suggestion is to treat them the same as existing ScenarioMIP experiments -- it might me more appropriate to treat them as Tier 2 (ssp119 etc -- );

(3) piControl-cmip5 : as for historical, I don't see the need to request the full volume of data requested from piControl. Can we omit sub-daily and Omon-3d data?

[claudiatebaldi]

Dear Martin, all that you propose sounds good to me. I do not see the reason to save subdaily under 1 or 3 and I agree that the cmip5 future experiments should be treated as Tier 2. Claudia From: Martin <notifications@github.com> Reply-To: cmip6dr/Request <reply@reply.github.com> Date: Friday, March 27, 2020 at 8:08 AM To: cmip6dr/Request <Request@noreply.github.com> Cc: "Tebaldi, Claudia" <claudia.tebaldi@pnnl.gov>, Mention <mention@noreply.github.com> Subject: Re: [cmip6dr/Request] New ScenarioMIP experiments using CMIP5 forcing (#118) @taylor13<https://protect2.fireeye.com/v1/url?k=6cdb4543-306e7b8c-6cdb6f56-0cc47adc5e60-471747b9c26b7fbe&q=1&e=0213d629-e0af-4c98-aa03-b8881e2c4295&u=https%3A%2F%2Fgithub.com%2Ftaylor13> , @durack1<https://protect2.fireeye.com/v1/url?k=c4cd2aab-98781464-c4cd00be-0cc47adc5e60-2d875dda9b4e1630&q=1&e=0213d629-e0af-4c98-aa03-b8881e2c4295&u=https%3A%2F%2Fgithub.com%2Fdurack1>, @claudiatebaldi<https://protect2.fireeye.com/v1/url?k=810102fd-ddb43c32-810128e8-0cc47adc5e60-11d8571990214a13&q=1&e=0213d629-e0af-4c98-aa03-b8881e2c4295&u=https%3A%2F%2Fgithub.com%2Fclaudiatebaldi> : we need to define the data request for these experiments. Paul has provided a partial answer<https://protect2.fireeye.com/v1/url?k=dc04acdc-80b19213-dc0486c9-0cc47adc5e60-afeab395ac67e631&q=1&e=0213d629-e0af-4c98-aa03-b8881e2c4295&u=https%3A%2F%2Fgithub.com%2FWCRP-CMIP%2FCMIP6_CVs%2Fissues%2F805%23issuecomment-558282479>. This raises some questions: (1) the historical experiment has a very large data request because of its central role as a baseline for many studies. The historical-cmip5 experiment does not have this role, so it does not make sense to request the same volume of data. Groups of variables which we might want to omit are: 3hr, 6hrPlev, 6hrLev, Omon-3d, esmval-3d? (2) rcp26-cmip5 etc: these are tier 2 scenarioMIP experiments. The Paul's suggestion is to treat them the same as existing ScenarioMIP experiments -- it might me more appropriate to treat them as Tier 2 (ssp119 etc -- ); (3) piControl-cmip5 : as for historical, I don't see the need to request the full volume of data requested from piControl. Can we omit sub-daily and Omon-3d data? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<https://protect2.fireeye.com/v1/url?k=8e2970cf-d29c4e00-8e295ada-0cc47adc5e60-d60d3c37918548a9&q=1&e=0213d629-e0af-4c98-aa03-b8881e2c4295&u=https%3A%2F%2Fgithub.com%2Fcmip6dr%2FRequest%2Fissues%2F118%23issuecomment-604964766>, or unsubscribe<https://protect2.fireeye.com/v1/url?k=a0a9afd2-fc1c911d-a0a985c7-0cc47adc5e60-9d645cc8b6c72969&q=1&e=0213d629-e0af-4c98-aa03-b8881e2c4295&u=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAFZSZZWBUGZUTMJQJFQRSFTRJSJKTANCNFSM4LU6EIYQ>.

[durack1]

@martinjuckes the omission of the Omon-3d will limit any meaningful ocean assessment, or energy budget assessment contrasting the forcing responses. In my head this was the main reason for requesting these cmip5-comparison simulations.

I was trying to ascertain what is included in the Omon-3d "table" or request, can you point me to the page on http://clipc-services.ceda.ac.uk/dreq/index.html where these variables are listed? Do we have tiers in this request?

I would recommend that thetao and so are be primary variables we'd want to include in a request, with this true for all the experiments

[durack1]

@martinjuckes @taylor13 the relevant linked issue is
WCRP-CMIP/CMIP6_CVs#805

[martinjuckes]

@durack1 thanks -- the link is also given at the top of this discussion.

Yes, omitting variables limits options; including too many is also a problem, so we have to strike a balance.

If you don't agree with Claudia, please respond to her.

[durack1]

@martinjuckes got it. There is no disagreement with Claudia, she made a statement about "I do not see the reason to save subdaily under 1 [historical-cmip5] or 3 [piControl-cmip5] and I agree that the cmip5 future experiments should be treated as Tier 2" I agree, the high (temporal) frequency data is heavy in storage volume, in addition to being heavy on computational demand, so I am not advocating sub-monthly data is saved.

My comment was that the Omon-3D (can you point to the variable list that is requested for this data request group?) should be included

[martinjuckes]

Omon-3d refers to the 3d variables in the Omon table (http://clipc-services.ceda.ac.uk/dreq/u/MIPtable::Omon.html). These variables are not requested for ScenarioMIP tier 2 experiments.

[durack1]

@martinjuckes the huge number of variables on that table would be overkill for these [historical-cmip5, piControl-cmip5, rcp*-cmip5] experiments. Is there a template for a subset of the full 3D variable list that we can use, or start with?