WOD (World Ocean Database) connection

[jmckenna]

Endpoints

• S3 buckets: https://noaa-wod-pds.s3.amazonaws.com/index.html
• WOD Select tool: http://wod.iode.org/SELECT/dbsearch/dbsearch.html
• ODISCat entry: https://catalogue.odis.org/view/182
• ~18 million records

Existing ISO 19115 metadata process

• a WOD process exists to generate ISO metadata files (although project was suspended in 2018)
• sample metadata file: https://gatewaygeomatics.com/dl/wod/wod_018917457O.xml
• The file name for the ISO file can be interpreted as
  • wod - World Ocean Database
  • 018917457 - WOD unique cast number
  • O - observed level (S=standard level)

Challenges of database size

• how to aggregate?
• WOD can be grouped by instrument & year
  • 10 instruments

Possible Next Steps

• examine existing metadata record
• examine how WOD data fits existing ODIS patterns (possibly create a new pattern if needed)
• adapt existing ISO process for JSON-LD support

cc @BoyerWOD

[pbuttigieg]

Starting development of a cast-by-cast dataset description with @BoyerWOD, @jmckenna, and @fils

[jmckenna]

@BoyerWOD we (@fils and myself) will provide you with a Python script to generate the necessary JSON-LD for each cast.

[jmckenna]

@BoyerWOD the Python script to generate all of the necessary JSON-LD files (direct from your S3 bucket) is ready for your review/run. See wod-s3-ncparser.py and sample output