[Go-essp-tech] basin variable in CMIP5

[Drach, Bob]

Hi Stephen,

To fix this problem, data publishers should remove 'basin' from the configuration option 'thredds_exclude_variables' in esg.ini. Since for CMIP5 the publisher requires the variable name to match the leading portion of the file name, the dimension variables named basin will still not be indexed (as should be the case).

Thanks, Stephen, for pointing out the problem.

--Bob

From: "Taylor, Karl Taylor" <taylor13 at llnl.gov<mailto:taylor13 at llnl.gov>>
Date: Tue, 29 May 2012 11:33:07 -0700
To: "stephen.pascoe at stfc.ac.uk<mailto:stephen.pascoe at stfc.ac.uk>" <stephen.pascoe at stfc.ac.uk<mailto:stephen.pascoe at stfc.ac.uk>>
Cc: "go-essp-tech at ucar.edu<mailto:go-essp-tech at ucar.edu>" <go-essp-tech at ucar.edu<mailto:go-essp-tech at ucar.edu>>, "esg-node-dev at lists.llnl.gov<mailto:esg-node-dev at lists.llnl.gov>" <esg-node-dev at lists.llnl.gov<mailto:esg-node-dev at lists.llnl.gov>>
Subject: Re: [Go-essp-tech] basin variable in CMIP5

Hi all,

The reason "basin" is on the exclude list (I think) is that "basin" is also a "dimension" variable (e.g. for msftmyz).  We don't want a lot of "dimensions" showing in the database as "variables", so they get excluded.  Problem is that when "basin" is a legitimate variable (as in the "fixed frequency" group), it gets omitted by the publisher.

Maybe Bob can suggest a work-around that won't simply remove "basin" from the exclude list, because this would result in unwanted dimensions being published.

regards,
Karl

On 5/28/12 3:25 PM, stephen.pascoe at stfc.ac.uk<mailto:stephen.pascoe at stfc.ac.uk> wrote:

Hi,

I want to warn datanode administrators about a potential problem with datasets containing the "basin" variable.

We have recently discovered that our CMIP5 datasets with fixed frequency in the ocean realm didn't contain the "basin" variable despite the underlying data containing this variable.  It turns out that the default ESGF datanode configuration lists "basin" as one of the variables to exclude (in the default esg.ini file) therefore basin was being omitted from THREDDS catalogues even though the files were in the datanode database.

A quick check of pcmdi9 shows that there are only 9 datasets containing "basin", all at GFDL.  This looks suspicious to me and it seems likely that other nodes have omitted basin in error.  The BADC has now corrected this problem.

Regards,
Stephen.



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/go-essp-tech/attachments/20120529/020e7ebf/attachment.html