iobis / env-data

ENV-DATA related issues and documentation
2 stars 0 forks source link

What BODC collections to use? #23

Open rubenpp7 opened 4 years ago

rubenpp7 commented 4 years ago

Hi,

Sometimes I have received synonym (or easily mistakable) BODC vocabs from our adventurous data providers such as:

Sampling net mesh size

Chlorophyll (not specified) concentration in the water:

I have always tried to stick to a small number of collections such as P01, P06, L05, L22, S10, S11, Q01, etc. However, when people dive into many BODC collections with "uncommon terms" to search, they get to situations like the shown above, where several terms are valid for one measurementType.

In these cases, do any of the collections (like the ones previously mentioned) have priority over the others in case we find synonym terms? what should we base our selection on? added date of the term? are some collections restricted or not recommended to use? should we restrict our search to the 7 or 8 common collections? is it OK to explore all the collections to search for our term if it is not found in the previously mentioned collections? are there any best practices developed?

I believe that BODC should be informed if situations like this arise, so they can take measures, has this been done?

For the vocabs of the example, I would suggest to use the Q01 term for the mesh size and the P02 for the chlorophyll, because I think that those ones probably fit better. What is the recommended best practice?

meliezer commented 4 years ago

Hi, These are good questions. IMHO it's very important being interoperabile with SeaDataNet, so whenever possible P01 is better than Q1. P02 is not a data variables vocabulary, nor C47. I believe that extra vocabularies could stay without causing problems,

P.S. BODC has https. Do I create problems using https for the vocabularies URL?

meliezer commented 4 years ago

Hi, I see in 4.8.3. of https://iode.org/index.php?option=com_oe&task=viewDocumentRecord&docID=26385 : The project team is directed to review and evaluate the existing vocabulary identifiers in the MoF data to determine what vocabularies are in use for the measurementType and 39 measurementTypeUnit but also for other terms. The team should produce a registry of the vocabularies recommended for use in OBIS data with regard to measurementType and measurementUnit.