Dear Randy, all,
I don't know if there's a "CF-compliant" answer to this, but will note that this is something that is anticipated by UncertML. A means for encoding UncertML concepts within NetCDF is not yet agreed upon, but the NetCDF-U convention proposal is one method. Personally I would like to see a CF-compliant mechanism for this, and I think that the "umbrella variables" concept that is being discussed now on Trac is a good candidate .
The idea behind this would be to have two "child" variables, one for the lower end of the range and one for the upper, which are linked by an "umbrella" variable that groups them. The "umbrella" variable would be somehow tagged with metadata to indicate that it represents a "credible range" or whichever UncertML concept is most accurate.
I stress that this is still under discussion but you might like to peruse the Trac ticket (
https://cf-pcmdi.llnl.gov/trac/ticket/79).
Hope this helps,
Jon
-----Original Message-----
From: CF-metadata [mailto:cf-metadata-bounces at cgd.ucar.edu] On Behalf Of Randy Horne
Sent: 15 September 2012 12:40
To: cf-metadata at cgd.ucar.edu
Cc: Randy Horne
Subject: [CF-metadata] how to express data variable element values that need to cover a range
Folks:
We have a type of product we are generating where the data value is not precisely known...only a range between two values.
Cells / boundary variables would work, but according to appendix A, they are to be used for coordinate variables only.
Is there a CF compliant approach to handling this ?
very respectfully,
randy
_______________________________________________
CF-metadata mailing list
CF-metadata at cgd.ucar.edu
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
Received on Tue Sep 18 2012 - 12:54:17 BST
