skybristol
commented
6 years ago Cool! I forgot that I had Philip Goldstein (CU-Boulder and formerly with the OBIS activity) post the last version of what was called the "Marine Biogeography" specification up to the ESIP Community Ontology Repository. I think you will find references and definitions for many of these terms here:
http://cor.esipfed.org/ont?iri=http://mmisw.org/ont/ioos/marine_biogeography
I assume that your set of variables in the file of 'mystery' variable names come from a compilation of what you discovered across all of the OBIS-USA source files. Many of the older files were processed to the "MBG" specification and should line up with what's in the COR.
Per our discussion a few minutes ago, I'd be interested to know if you can write code to work directly against that iteration of the MBG as "registered" in the ESIP COR. This would be a good test to see if we can leverage that system directly. We can somewhat count on it persisting as long as we stay active in ESIP, talk about how we are using it, and take ownership of that and any other "ontologies" that we register.
If you do hit that with your code (either directly or by pulling the information and getting it into your expected JSON format) and still come up with some terms from the OBIS-USA sources that you don't find, please post that back here to the issue, and we'll get Abby to take a look and fill in the blanks.
In an effort to fully describe data variables and unit specifications, NetCDF files have been infused where applicable with Darwin Core terminology and url tracebacks to the source material (http://www.tdwg.org/activities/darwincore/) along with explicitly prescribed units both for key spatiotemporal features (lat,long,depth) and in cases where variable names include unit indicators directly (e.g. "AttributeName_in_meters"). The preliminary *.json structure is shown here. Currently, the file is only a running list of terms, not parsed out by item number.
Other variables shown here are not Darwin Core terms. A similarly structured *.json vocabulary file of missing terms has been created to populate pertinent information patterned after the Darwin Core descriptions, for consistency. The missing definitions and information should probably be populated at least in part from SB records or inferred in some way.
Also, neither *.json vocabulary files contain a "unit" keyword at present, but it may be advantageous to include one, where applicable for dimensional variables, for the sake of clarity. Thus far, we're adhering to UDINT data standards.
@skybristol