source: CMIP6dreqbuild/trunk/inputs/steve_em1.txt @ 322

Subversion URL: http://proj.badc.rl.ac.uk/svn/exarch/CMIP6dreqbuild/trunk/inputs/steve_em1.txt@947
Revision 322, 4.0 KB checked in by mjuckes, 5 years ago (diff)

added mip inputs

Line 
1################
2Note Jonathan's input will need modifying -- have a selection from Tables, not whole tables.
3################
4Hello Steve,
5
6thanks, that makes things a lot clearer. My suggestion of a diagnostic MIP is clearly not appropriate, I hadn't understood all that was behind Jonathon's original question.
7
8The paper you have provided, and the level of consensus it represents, is clearly very valuable. Nevertheless, WGCM has made it clear that they want the proposers to commit to analysing the data they ask for. Given the concerns that modelling centres have expressed about the volume of data requested I would expect some objections to simply specifying "all" experiments, even though you have obviously taken pains elsewhere to moderate the volume of the data request. e.g. do you expect LUMIP 3d ocean fields to be analyzed?
9
10Personally I can see a good case for having a core set of variables across all experiments, but that is not the way WGCM is playing this. If that approach had been followed I'm sure there would have been a lenghty debate about how much 3d ocean data was needed in such a core, which may be why WGCM wanted an individual to take responsibility for each request.
11
12I realise it is a lot of work to povide explicit justification for each variable/experiment combination, but there has been strong feedback on the need for a significant improvement in this aspect of the request. The danger is that modelling centres start making their own decisions about what to include, and this resulted in a patchwork of gaps in CMIP5. In the formulation I've tried to simplify this slightly by asking for experiment groups to be specified.
13
14E.g. what is the objective of cfc11 for the historical expt? I would guess it is something to do with diagnosing and validating circulation? The idea behind formulating the template as it has been done was to try to capture this sort of information. And also about who is leading it.  ## this should come in the OCMIP submission.
15
16Your document does not mention priorities of variables. As for CMIP5, each variable should be assigned a priority.
17
18e.g. reduced resolution lat-lon for generic diagnostic analysis across whole range of experiments.
19
20Have you been in touch with the GeoMIP team, led by Ben Kravitz? They have asked for a slightly different set of ocean variables. I hope they can be persuaded to converge onto your selection. Can I pass the CMIP6_ocean.pdf document ont Ben?
21
22
23------------
24
25Thanks for the feedback.  Here are some follow up queries.
26--The OMDP document provides guidance for all ocean model efforts associated with CMIP projects.  We provided the same document for CMIP5.  Both documents were requested by WGCM and CLIVAR SSG.  These documents provide a scientific and engineering rationale for ocean diagnostics, some of which have origins in CMIP1. 
27-Karl Taylor used the CMIP5 version of this document as the basis for the CMIP5 ocean model spreadsheet.  We wrote the CMIP6 edition under the presumption that the same transfer of information would happen for CMIP6. 
28--As written in the CMIP6/OMDP edition, each variable has an associated "owner" associated with it. Some "owners" are CMIP6, including such generic variables as SST, sea level part of this group.  Other variables point to specific MIPs such as FAFMIP and OCMIP.   I thus see no reason to design a "diagnostic MIP", since the other MIPs have already been put forward and they will presumably take charge of the variables.
29--Based on the feedback I have received from the scientific community, including WGCM, this document is a benchmark to which other groups should aspire, such as atmospheric, sea ice, land, biogeochemistry, etc.  It is an umbrella document that aims to bring together many ocean related threads spread throughout various MIPs.  Hence, it is not a single MIP itself, but it is a rationale for the many diagnostics requested for the various MIPs.  Without scientific rationale, along with specific guidelines for engineering specifications, the data accumulated in CMIP6 will be diffuse and less useful to the scientific community. 
Note: See TracBrowser for help on using the repository browser.