[Go-essp-tech] Fwd: scientist reviews of trackback

[Cecelia DeLuca]

  Had sent this to the wrong address - committee instead of tech, so 
forwarding.
Some follow-up is forthcoming on Gavin (Schmidt)'s requests regarding the
search.

Best,
- Cecelia

-------- Original Message --------
Subject: 	[GO-ESSP-COMMITTEE] scientist reviews of trackback
Date: 	Wed, 29 Jun 2011 07:50:26 -0600
From: 	Cecelia DeLuca <Cecelia.DeLuca at noaa.gov>
To: 	'Metafor List' <metafor at lists.enes.org>, go-essp-committee at ucar.edu
CC: 	Sylvia Murphy <Sylvia.Murphy at noaa.gov>



   Hi all,

When Sylvia and I were at the CESM meeting last week, we pulled a couple
of scientists aside to demo the trackback interface to them.  Below
are notes from Gavin Schmidt and Dave Bader.

Gavin:
- Was confused that the experiment description may not reflect what
a group was actually able to do, which is under conformances.  A note
at the top of the experiments page would help to clarify this.
- Very much wants to search on forcing.  The dynamic comparison table
would also be useful here, so you could see a property across a range
of components.
- Diffs against a standard run would make the display more useful for
day-to-day work.  (Here we were talking about being able to generate
and use this kind of metadata in a more routine way.)  Of course the
diffs are useful in general as well.
- Under conformances, allow (but more like encourage, really) users
to describe what else was done to their model besides what was asked
for in order to comply, and other configuration details of note.
- Have a way for the originator to clarify and update metadata and
ensure that the versions and updates are clearly noted.
- *Have a way for people to comment on metadata.* This was a big one,
since he immediately had questions about the way the HadGEM runs were
performed.  We talked about aliasing or something
like Craigslist where there might be a screen between the contacter
and the contactee.
- What he really wants is a space-time filter for multiple variables that
he can apply across a range of simulations so that he can look at features
(but that's someone else's job ...)

Dave:
- Wants a tag in the netCDF file that connects to the metadata.  He
would like the DOI to be connected to the full provenance of the simulation,
including the post-processing.
- Wants to see more information that gets us closer to reproducibility
in the technical section (OS version, compiler version, etc.).
- Wants to diff or look at properties of members of the same ensemble
run.
- Define a retention period for simulations - I don't remember this
thread, but I  think we were talking about "long term storage" of climate
model runs, what was meant by that, and the role of metadata.

-- Cecelia




-- 
===================================================================
Cecelia DeLuca
NESII/CIRES/NOAA Earth System Research Laboratory
325 Broadway, Boulder 80305-337
Email: cecelia.deluca at noaa.gov
Phone: 303-497-3604

_______________________________________________
GO-ESSP-COMMITTEE mailing list
GO-ESSP-COMMITTEE at ucar.edu
http://mailman.ucar.edu/mailman/listinfo/go-essp-committee

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/go-essp-tech/attachments/20110707/5c2c8ca4/attachment.html