[Go-essp-tech] Meeting minutes from DRS discussion on 03/20/2013

[Galia Guentchev]

Dear all,

The meeting minutes from our DRS discussion last Wednesday 03/20/2013 
have been archived here: 
http://earthsystemcog.org/projects/downscalingmetadata/meetings

Please find below summary and TODO items.


_Summary:_

1. We are discussing the "DRS" structure (which does not necessarily 
explicitly map to a directory structure)
Focusing the discussion on: facets / sequence of facets; what is and 
isn't required;

  * Proposed facets -

  * /projectID/sub-project/product/institution/predictorModel/experimentID/frequency/realm/MIPtable/Predictor_experiment_rip/predictorversion/downscalingMethod/variableName/region/DownscaledDataversion/file_name.nc

  * Example:

  * /ncpp2013/perfectModel/downscaled/NOAA-GFDL/GFDL-HIRAM-C360-coarsened/amip/day/atmos/day/r1i1p1/v20121024/GFDL-ARRMv1/tasmax/US48/v20120227/tasmax_day_amip_r1i1p1_downscaled_US48_GFDLARRMv1_19790101-19831231.nc


  * All agreed on following facets and values:

      * *project*

      * *sub-project*- not sure about the value of this facet for NASA
        data; for GFDL - "perfectModel"; NASA would like to use this for
        quantile or quartile data as a sub-group - recommended to use
        VariableID for that;

      * *product*= downscaled;

      * *institution *- (important as a facet, NO need for search on it)

      * *predictor model*= model that is downscaled or coarse version of
        high-res model for perfect model setting; (this facet is used
        instead of ModelID);

      * *experimentID*= amip, or other; value = 'historical' is
        acceptable as long as the data are not CMIP5 data;

      * *frequency, realm, MIP table*- all agreed on these facets and
        their values to follow CMIP5 requirements - for ex. mon or day,
        atmos, amon, respectively;

      * *Predictor_experiment_rip*- all agreed on these facets; value
        comes from the original model; TODO: KT will compare this with
        CORDEX document on archive definition, and with CMIP5 requirements;

          * Seems that there is no rip for ensembles; Needs to be added
            - TODO: KT would look into this.

      * *predictorversion*= the version of the predictor dataset;

      * *downscalingMethod = *specific statistical method or a specific
        dynamical model;

      * *variableName *-  accepted as a facet by all;

      * *region *- CONUS or US48 - TODO: KT - will send around
        information about how to indicate region in CMIP5;

      * *Downscaleddataversion*= date the downscaled data were produced;

      * *Filenames*for NASA have been following CMIP5 convention; same
        for GFDL;

          * filenames need to be unique;

          * Important to have "predictor model" also in addition to
            "downscaling method" within the filename;

          * need to have "region";

          * could skip product="*downscaled*" from the filename;


  * Other issues:

      * "tracking ID" can be in netCDF headers - should have that
        information; not in the DRS;


2. Questions:

  * from NASA - DOI for that data - would like to have this in the
    attributes? How others are handling that? it will not be searchable;
    TODO: could ask go-essp list about that;

  * KT - may be there could be a version number place in CMIP5
    sequencing - will check for consistency;


2. *Summary*: We agree on facets, values of facets and sequencing of 
facets - except for what Dr. Karl Taylor will check on; To communicate 
via email to obtain final decisions;


Please let me know if you find any discrepancies in the archived minutes 
or in the summary notes;


Best regards,
Galia


-- 
Galia Guentchev, PhD
Project Scientist
National CLimate
Predictions and
Projections
Platform (NCPP)
NCAR RAL CSAP
FL2 3103
3450 Mitchell Lane
Boulder, CO, 80301
phone: 303 497 2743

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/go-essp-tech/attachments/20130326/25e19e46/attachment.html