Revise geologic semantics for improved alignment

[pbuttigieg]

Based on the below, determine how to create improved (but less detailed) semantic structures in ENVO to link off to specialist geological / minerology ontologies.

At high levels of consolidation - do the grains exist?

In general that is how rocks are described, e.g. a metamorphic rock, strongly consolidated (indurated), consists of 30% muscovite in 2-3 mm diameter flakes aligned in a schistose fabric; 15% equant almandine garnet 1-2 mm; 40% anhedral quartz grains 1-4 mm diameter, some with oblate shapes aligned in the schistosity; 15 % plagioclase (An 20-30) in subhedral 1-2 mm grains. The fundamental nature of a rock or sediment is its particulate constitution. Of course, being the Earth, there are edge cases that are interesting -- volcanic glass, aphanitic materials, monomineralic rocks.

I've looked at ENVO several times thinking about how it might be better harmonized with GSO/GeoSciML, but always come away concluding that its a BIG job that would be really interesting, but needs some resourcing!

Originally posted by @smrgeoinfo in https://github.com/EnvironmentOntology/envo/issues/1042#issuecomment-1215191815

[pbuttigieg]

related and based on an unsatisfiability error on "limestone" in #1345

2022-08-15 17:58:49,206 ERROR org.obolibrary.robot.ReasonerHelper - There are 6 unsatisfiable classes in the ontology.
2022-08-15 17:58:49,207 ERROR org.obolibrary.robot.ReasonerHelper -     unsatisfiable: http://purl.obolibrary.org/obo/ENVO_00003914
2022-08-15 17:58:49,207 ERROR org.obolibrary.robot.ReasonerHelper -     unsatisfiable: http://purl.obolibrary.org/obo/ENVO_00000567
2022-08-15 17:58:49,207 ERROR org.obolibrary.robot.ReasonerHelper -     unsatisfiable: http://purl.obolibrary.org/obo/ENVO_01000[27](https://github.com/EnvironmentOntology/envo/runs/7843247136?check_suite_focus=true#step:4:28)3
2022-08-15 17:58:49,207 ERROR org.obolibrary.robot.ReasonerHelper -     unsatisfiable: http://purl.obolibrary.org/obo/ENVO_00002053
2022-08-15 17:58:49,207 ERROR org.obolibrary.robot.ReasonerHelper -     unsatisfiable: http://purl.obolibrary.org/obo/ENVO_00002054
2022-08-15 17:58:49,207 ERROR org.obolibrary.robot.ReasonerHelper -     unsatisfiable: http://purl.obolibrary.org/obo/ENVO_00003982
make: *** [Makefile:170: envo-full.owl] Error 1

An interesting reasoning error was generated when this class had "composed primariy of some CHEBI:calcium carbonate". As that was a CHEBI:mineral, and minerals are disjoint from ENVO:rock, it led to an unsatisfiable class. The current axiom works around this, but this is requires some sort of geological convention to diffferentiate between masses of minerals and rocks that are composed primarily of minerals.

Right now, we can't say that limestone is composed primarily of calcium carbonate without getting an unsatisfiability error (rock and mineral are disjoint, and if you're mostly made up of a mineral, aren't you a mineral mass?)

[smrgeoinfo]

Yes, the solution is that a rock is composed of particles, and the particles can be composed of rock (e.g. a conglomerate) or minerals (dodging questions about porosity and what's in the pores...). So a limestone is a rock composed of particles that are mostly composed of calcite, which is composed of mostly calcium carbonate. The nature of those particles is how geologists differentiate kinds of limestone. The real stuff almost always has some silica particles, clay particles, and the calcite almost always has some (or maybe alot of) Mg.