NTR: depth in soil

[cmungall]

Working with @kaiiam and NMDC workshop ontology team, need key mapping for depth in soil package

[kaiiam]

Thanks @cmungall, there are many of these MIXS terms across all the packages which can be make use of existing templates such as the chemical_concentration robot templates. As I will likely be contributing to the NMDC effort to help cover MIXS, I can take the lead on the creation of the terms using the templates.

[pbuttigieg]

Xref #897 @kaiiam note the issue with "depth of what?" which brings BCO into the mix (no pun intended)...

[dr-shorthair]

What is the BCO connection here? (I don't see it.)

[kaiiam]

At the moment I don't think there is a BCO connection for this type of term dept of X classes these would be in ENVO or in the spinoff quality ontology if we decide to do that. Perhaps BCO could be used if the MIxS Structured comment fields need to be put into an ontology maybe but that's not the priority for now. I presume that in an RDF implementation not every comment in the MIxS spreadseet would a full class associated with it anyway.

If there is funding to work on an RDF implementation of MIxS we can pick the BCO discussion up again. For now though I think the priority should be the creation of the MIxS quality terms such as dept of X classes in ENVO or the spinoff ontology.

[ramonawalls]

Depth is soil is really about location, not a property of soil. Of course the location is important, because it tells you something about the qualities of the soil where the sample was collected, but I fear the discussion of this term has morphed it into a soil quality. E.g., if depth of X is depth of soil, then that is a quality of the soil, not metadata about the sampling process, which is the key information needed for MIxS.

[pbuttigieg]

Refreshing this - use the patterns approach to generate depth of soil similar to http://purl.obolibrary.org/obo/ENVO_09200023

[dr-shorthair]

Term not found ...

[kaiiam]

I'd be similar to http://purl.obolibrary.org/obo/ENVO_3100031 I think ENVO_09200023 might either be a term in progress or a typo.

[smrgeoinfo]

Depth of some thing requires 2 things: -- depth of what. this must be some entity that has a discrete location, e.g. the top of a soil horizon, the top (bottom) of a stratigraphic unit), the upper boundary of some object (e.g. a reef)... -- a spatial reference system. What is the 'depth' relative too-- e.g. mean sea level, local surface elevation (useful for depth in a borehole)...

Depth of 'soil' can't be quantified-- what are we talking about-- depth of top of soil horizon, depth of bottom of soil horizon, depth of mid-point in soil horizon, depth of a sample location...

Depth 'in' soil sound more like the depth of a sampling site (w.r.t. some spatial reference system).

[kaiiam]

Depth 'in' soil sound more like the depth of a sampling site (w.r.t. some spatial reference system).

That would be the most common annotation use case yes. This follows a classic OBO entity attribute pattern, where a quality/characteristic/attribute depth inheres in some entity e.g., soil or water. Lets not over think this.

[smrgeoinfo]

Hope I'm not overthinking... but 'soil' and 'water' are mass nouns, and thus I would think they don't have location properties. The boundary of a 'portion' of soil is an object (countable), and can have a location (e.g. depth...)

[kaiiam]

We aren't saying location per say, we are modeling these as qualities (specifically dependent continuants). So the depth of the soil or water is dependent on that soil or water being sampled. OBO does this similarly with concentrations we also have them as qualities that inhere in some chemical entity that is part of some material. Is the the perfect model who knows but its what we've settled on for OBO projects.

[smrgeoinfo]

Thanks for the explanation