Dear Daniel,
I agree with Jonathan's interpretation: "as self-describing as possible" can only go so far. I don't see any reason in principle not to include pointers to external resources. However, the problem is always in the detail.
The problem appears to be that you consider the amount of information which can be put in a valid flag_meanings word is too limited in some cases, which I find plausible.
You can place extra information in the "comment" attribute, but this is not semantically tied to the flags, so there is a plausible case for having a specific attribute to provide additional information about the flags.
There is a precedent for referring to external definitions of terms in the "land_cover_lccs" standard name, which refers to the UN Land Cover Classification System. Not a great precedent, as the reference given is a web page from 2000 explaining the system and motivation behind the terms, but getting at the terms requires running the database on a computer with windows 95/98 or NT, apparently. It appears that LCCS has evolved to become a software tool and the terms the community is now using are in the Land Cover Meta Language (LCML), which is an ISO standard. This suggests, I think, we need to be cautious about how external references are introduced.
regards,
Martin
________________________________
From: CF-metadata <cf-metadata-bounces at cgd.ucar.edu> on behalf of Jonathan Gregory <j.m.gregory at reading.ac.uk>
Sent: 24 July 2019 16:58
To: cf-metadata at cgd.ucar.edu
Subject: [CF-metadata] PID to external description of Quality and Status states
Dear Daniel
I assume you mean a data variable which has strings describing quality or
status states, probably encoded with flag_values and flag_meanings. Would
the flag_meanings be meaningful to a human reading them? If so, you could
regard the file as self-describing. Self-describing doesn't have to mean
self-documenting, I would say. Standard names are self-explanatory to some
extent, and make the file self-describing, but they also have definitions
which aren't in the file, and papers in the peer-reviewed literature about
the quantities. If the flag_meanings are self-explanatory, I think that's OK,
but I'm not in favour of putting codes in the CF-netCDF file.
Best wishes
Jonathan
----- Forwarded message from Daniel Neumann <daniel.neumann at dkrz.de> -----
> Date: Wed, 24 Jul 2019 14:45:33 +0200
> From: Daniel Neumann <daniel.neumann at dkrz.de>
> To: cf-metadata at cgd.ucar.edu
> Subject: [CF-metadata] PID to external description of Quality and Status
> states
> User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:60.0) Gecko/20100101
> Thunderbird/60.8.0
>
> Dear list,
>
> A CF-conformal NetCDF file should be as self-describing as possible.
> URIs/PIDs poitining to external data sources are problematic in this
> context. However, IDs pointing to external data sources were already
> discussed for taxa in the CF Trac Ticket #99 "Taxon Names and
> Identifiers" (https://cf-trac.llnl.gov/trac/ticket/99).
>
> Let's assume we have a long description of quality control measures
> performed resulting in N possible quality states for a NetCDF
> variable. The full description could be part of a peer-reviewed
> paper or grey literature (with assigned PID). However, it is too
> long to be included in the attribute "description" of the
> quality/status flag variable. One solution would be to add a brief
> description of the quality states to the quality_flag/status_flag
> variable and point to the external full description via it's PID.
>
> What is your opinion on this?
>
> This is not a request for a feature but rather meant to collect some
> opinions on this topic. It is somehow related to the request of a
> global PID attribute in CF GitHub Issue 160 "Add attribute
> citation_id"
> (https://github.com/cf-convention/cf-conventions/issues/160).
>
> Regards,
> Daniel
>
>
> --
> Dr. Daniel Neumann
> Abteilung Datenmanagement
>
> Deutsches Klimarechenzentrum GmbH (DKRZ)
> Bundesstra?e 45 a ? D-20146 Hamburg ? Germany
>
> Phone: +49 40 460094-120
> Email: daniel.neumann at dkrz.de
> URL: www.dkrz.de<http://www.dkrz.de>
>
> Gesch?ftsf?hrer: Prof. Dr. Thomas Ludwig
> Sitz der Gesellschaft: Hamburg
> Amtsgericht Hamburg HRB 39784
>
>
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
----- End forwarded message -----
_______________________________________________
CF-metadata mailing list
CF-metadata at cgd.ucar.edu
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
Received on Thu Jul 25 2019 - 12:36:43 BST
