IMOS: PRES dimension as a Z dimension

[lbesnard]

Not having a DEPTH dimension, and using a PRES one instead triggers a series of issues with the IMOS checker, while the file is CF compliant.

globalattr                             :3:    57/59 :  
    geospatial_vertical_max            :3:     0/ 1 :  
        type                           :3:     0/ 1 : Attribute
                                                      geospatial_vertical_max
                                                      should have numeric type
    geospatial_vertical_min            :3:     0/ 1 :  
        type                           :3:     0/ 1 : Attribute
                                                      geospatial_vertical_min
                                                      should have numeric type
var                                    :3:    88/96 :  
    PRES                               :3:    13/17 :  
        positive                       :3:     0/ 1 :  
            vertical                   :3:     0/ 1 : Vertical coordinate
                                                      variable (axis='Z') should
                                                      have attribute positive =
                                                      'up' or 'down'
        reference_datum                :3:     0/ 1 :  
            type                       :3:     0/ 1 : Variable attribute
                                                      PRES:reference_datum not
                                                      present
        standard_name                  :3:     1/ 2 :  
            vertical                   :3:     0/ 1 : Vertical coordinate
                                                      variable (axis='Z') should
                                                      have attribute
                                                      standard_name = 'depth' or
                                                      'height'
        units                          :3:     1/ 2 :  
            vertical                   :3:     0/ 1 : Variable PRES appears to
                                                      be a vertical coordinate,
                                                      should have units of
                                                      distance
    coordinate_variable                :3:     1/ 2 :  
        space_time_coordinate_present  :3:     0/ 1 : No space-time coordinate
                                                      variable found
    data_variable                      :3:     3/ 5 :  
        PSAL                           :3:     1/ 2 :  
            check_dimension            :3:     0/ 1 : data variable should have
                                                      at least one of the
                                                      dimensions TIME, LATITUDE,
                                                      LONGITUDE, DEPTH
        TEMP                           :3:     1/ 2 :  
            check_dimension            :3:     0/ 1 : data variable should have
                                                      at least one of the
                                                      dimensions TIME, LATITUDE,
                                                      LONGITUDE, DEPTH
    geospatial_vertical_units          :3:     0/ 1 :  
        check_attributes               :3:     0/ 1 : units is not a valid CF
                                                      depth unit

Example file : IMOS_AATAMS-SATTAG_TSP_20130302T012000Z_Q9900562_FV00.nc.zip

[mhidas]

According to v1.3 of the IMOS conventions, pressure is supposed to be converted into depth in metres to provide a vertical coordinate, but this is not stated in the new version (1.4). However both versions do say you can't label a pressure as "depth" or "height". Either way this needs some work.

[mhidas]

@lbesnard Have you made any changes to this kind of file? There may be a few things we can improve in the checker, but I think the file is technically not IMOS-1.3 compliant. We may need to have a chat with @ggalibert and @smancini to decide what we want to insist on.

[lbesnard]

@lbesnard Have you made any changes to this kind of file?

No I haven't. happy to have a chat

[mhidas]

Just a note, the message about geospatial_vertical_min/max being numeric type is a separate issue. (The offending file has these attributes as integers, and the checker is currently expecting a float.)

[mhidas]

After further discussions we've decided to leave these checks unchanged in the checker, and instead look into the possibility of adding a DEPTH variable to these files.

[mhidas]

@lbesnard What's the status of these files? Is this still an issue or have you found a way around it? If we need to keep tracking this issue, please move it to the content repo. Otherwise please close it. Thanks.

[mhidas]

This issue has been moved to https://github.com/aodn/data-services/issues/652