Document requirements and usage of ENVO for earth-specific vs interplanetary use cases

[cmungall]

ENVO was originally designed for earth-based microbiome sample metadata and is being generalized for broader use cases, including those that are not-earth bound.

Generalizing ENVO to interplanetary use cases will have impacts on its usability for earth-based use cases. There are examples in various tickets, e.g #1034. I will add more to this ticket later.

In many ways this is analogous to the species-specific vs species-neutral anatomy ontology scenario that we have >10 years experience with (and >20 for GO). We have various techniques for dealing with this that can be applied (e.g. taxon-GCIs => planetary GCIs TODO add link). But the situation still involves compromise and things are often left unsatisfactory.

Sometimes it's necessary to have specific ontologies synchronized with the general one. This should not be off the table for ENVO. Although a complicating factor here is that there is already considerable uptake of ENVO for earth use cases (e.g microbiome) and managing a transition will be hard. But we should consider all options.

Stub ticket, will fill out later

[kaiiam]

(cc @rduerr @brandonnodnarb @lewismc @dr-shorthair)

As discussed in the recent ESIP semantic harmonization sessions @rduerr's new (I believe NASA-associated) work involves the need for interplanetary semantics. She said (correct me if I'm wrong) that some NASA systems are currently using ENVO.

This thread is also relevant to the ESIP Summer 2020 meeting about an ontology federation, which floated the idea of a Planetary Data Systems ontology.

Also see this issue about linking SWEET to the PDS4 Data Dictionary

[cmungall]

Do you have any more documentation on the NASA use case (and is this really interplanetary)

Even if there is a genuine need for interplanetary terms, ENVO should be earth-centric, otherwise it will be far too confusing for users.

For example, when ENVO uses a term like "glacier" it should mean a glacier consisting of water ice, because that is what the overwhelming majority of users and annotators expect. Similarly, "ocean" should mean an ocean consisting of water, because that is what normal users will expect.

If there is a use case for annotating nitrogen ice glaciers and ice oceans on Neptune or Pluto then we can have terms "nitrogen ice glacier" and "nitrogen ocean" but these should not be subclasses of the earth-centric ENVO terms, but related via another relation. We could consider a planet-neutral grouping but these have to be balanced against complexity and difficulty of maintenance.

[kaiiam]

I'm not in any position to comment about anyone else's projects e.g., NASA but yes I agree that if ENVO were to host non earth-centric terms they shouldn't confuse or conflate the more common use cases. "le mieux est l’ennemi du bien". Shouldn't degrade the earth semantics overreaching for the stars.

[rduerr]

While I am not in a position to argue for or against non earth-centric terms in ENVO, I must note that methane clathrate ice is an important Earth material and important to any climate change discussion.  So ices other than water ice are needed in ENVO.Sent from my iPhoneOn Jul 28, 2023, at 2:15 PM, Kai Blumberg @.***> wrote: I'm not in any position to comment about anyone else's projects e.g., NASA but yes I agree that if ENVO were to host non earth-centric terms they shouldn't confuse or conflate the more common use cases. "le mieux est l’ennemi du bien". Shouldn't degrade the earth semantics overreaching for the stars.

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: @.***>