Ticket #940 (assigned Task)
New simulation relationship required
| Reported by: | charlotte | Owned by: | gerry |
|---|---|---|---|
| Priority: | blocker | Milestone: | V1.2 Questionnaire release |
| Component: | WP6 - CMIP5 Questionnaire | Version: | 1.1 |
| Keywords: | simulation relationship | Cc: | allyn mark sylvia |
| Requirement: | http://metaforclimate.eu/Work-Package-2/Developing-the-CIM/Project-Requirements-summary.htm | ||
Description
I would like to add a new simulation relationship: isEnsembleAbout
CMIP5 specifies a set of core experiments that modelling must do their best to complete and then tiers of experiments for centres to perform if computational time allows.
Often the tiers include extra ensemble members for simulations that have already been performed (and the data already delivered). The extra ensemble runs are noted with -E extensions in the CMIP5 expeiment design but are given the same experiment name in the DRS world view (and the same name in the questionnaire output).
I would like a user returning to the questionnaire to be able to say that the simulation (ensemble) they are describing is an ensemble about a simulation they have already described.
Change History
comment:1 Changed 2 years ago by charlotte
- Owner changed from gerry to charlotte
- Status changed from new to assigned
comment:2 Changed 2 years ago by charlotte
- Owner changed from charlotte to gerry
- Status changed from assigned to new
how about "expandsEnsembleOf"
comment:3 Changed 2 years ago by gerry
- Status changed from new to assigned
Hi, did we get any resolution of this? I think Mark's "isEnsembleExtensionOf" is pretty close? I can make this change easily in the qn upon agreement of a suitable term. I'll keep the ticket open however until we have a good message to the user in place asking them to give us good information on what is different between original simulation run and this extended exnsemble run.

Charlotte
Yes you need this extension as we are no longer following the experiment plan in terms of referencing.
I don't think "isEnsembleAbout" is the right word - "isEnsembleExtensionOf" is closer to what is actually being done (but I'm still not entirely happy with it).
One of the things that often changes for the extending ensemble members is that we lose the high frequency diagnostics and even the number of diagnostics delivered. I don't think we collect this information directly, but I think we need to encourage people to spell out in the text descriptions how they differ from the 'main' run.
Regards
Mark