Reviewing and editing geospatial metadata

  A presentation to the Association of Earth Science Editors
  by Peter N. Schweitzer (U.S. Geological Survey, Reston VA)
  in Shepherdstown, WV on 7 September 2005

Overview of the concepts of metadata: what's it for

  Overall motivation
    Why publish anything?
      So other people can use the information
        to see what we have done
          so they can judge our interpretations and conclusions
          so they will value us and our organizations
        to get their work done

    The more detailed the information we make available,
    the more confidence people have in their use of it, and
    consequently in us.

Relationship of metadata to data: semantics and structure of files

  Metadata is structured documentation.  FGDC metadata is
  specifically intended for documenting geospatial data.
  You can use it for documenting almost anything, if you
  want. But to help people understand the data, show them
  maps and provide usable legends as data.

  Resource:  Metadata in Plain Language
    http://geology.usgs.gov/tools/metadata/tools/doc/ctc/

  FGDC metadata is not a complete explanation of the
  knowledge represented in the data; it is a summary whose
  degree of detail is a compromise between what we think we
  know and what we think the user knows already.  There is
  no way to prescribe the proper level of detail that
  should be given in a metadata record.

  FGDC metadata does not normally include a table of contents
  or list of files that come with the publication.  You can
  include it in the metadata as an extension but it's also
  important to provide a text file containing this information.

Overview of software tools that may be helpful for review

  Reviewing data and metadata
    Structure and format of information
    Content (readability, consistency, accuracy, precision)

  Tools
    GIS and other scientific software
    mp, tkme, dbfmeta
    specialized programs and scripts (mq)

  Resource: Formal metadata: information and software
    http://geology.usgs.gov/tools/metadata/

Consistency and common problems, conceptual and detailed

  Too little information
    Missing pieces
    Perfunctory answers
  Too much information
  Putting things in the wrong places
    Difficult elements
      Attribute_Accuracy_Report
      Completeness_Report
      Logical_Consistency_Report
      Supplemental_Information
    See-sickness

  Resource: Top ten metadata mistakes
    http://geology.usgs.gov/tools/metadata/tools/doc/opinion/mistakes.html

What's worth arguing about and what isn't

  Do insist on
    Readability
    Answering the key questions
    Understanding what you want to enable people to do
    Understanding what you expect people to know already
      not because you want to state prerequisite knowledge,
      but because your expectations will color your
      opinions during review.
    Reasonable packaging
    Economy of scale
      If there's a simple change that will help lots of
      users, don't make everybody do it.

  Don't fret about metadata elements that
    should have reasonable defaults
      Attribute_Definition_Source
      Enumerated_Domain_Value_Definition_Source
    aren't done by the community with any consistency
      Latitude_Resolution
      Longitude_Resolution
      Abscissa_Resolution
      Ordinate_Resolution
    aren't the responsibility of the authors
      URL for the data and web site: web people
      disclaimers, liability for use: policy people
      formal keywords: librarians or collection managers

  Resource: How to review metadata
    http://geology.usgs.gov/tools/metadata/tools/doc/opinion/review.html

A word about technical support

  No amount of review will catch all errors.  People will
  do things with the data that nobody can anticipate. When
  they do, they will find problems, some of which can or
  should be fixed.  Do it.  Don't argue for a hopeless
  principle (for example, the sanctity of the approval of
  the publication) over usability of the data.  Yes, your
  organization will act as though technical support is not
  its job.  Argue against them, and then do it anyway.
  The result is that you will become known as someone who
  is useful, and this will reflect well on your role within
  the organization.

(end)