soreag:GeologicFeature

[nicholascar]

Preferred term label

geologic feature - current Geologic Feature - proposed. Capitals for a class name

Synonyms

Geological Feature - skos:altLabel. "Geological" is used in place of Geologic in some places such as Australia

Textual definition

A Geologic Feature is a geospatial feature on the Earth that results from geological processes.

Usage Note

skos:scopeNote: Geologic Features include Stratigraphic Units, Non-stratigraphic (Lithodemic) units, Stratigraphic Event Features, Tectonic and Structural Features, GeoResource Accumulations, and Geologically Significant Sites.

Suggested parent term

The simple test for class subsumption X subset of Y, is every Object that is a member of Class X also a member of Class Y is not true, viz not every Geologic Feature X is also a Planetary Realm Y.

We think that the Class GeologicFeature isn't a subClassOf PlanetaryRealm; it is the set of things within the planetary realm so there is some other relationship that is needed, not subClassOf. The other existing restriction on the class is correct: sorel:hasRealm only sorea:Geosphere.

In summary:

• sorea:PlanetaryRealm - current, remove
• restriction: sorel:hasRealm only sorea:Geosphere - existing, retain
• sorepsg:GeometricalObject - add

Attribution

Using schema.org as well as FOAF for search engine recognition.

dcterms:contributor [
    a foaf:Organization , <http://schema.org/Organization> ;
    foaf:name "Geological Survey of Queensland" ;
    <http://schema.org/name> "Geological Survey of Queensland" ;
    <http://schema.org/identifier> <http://linked.data.gov.au/org/gsq> .
] ;

Example RDF (turtle)

@prefix sdo: <https://schema.org/> .

soreag:GeologicFeature 
    a owl:Class ;
    rdfs:subClassOf sorepsg:GeometricalObject ,
    [ 
        a owl:Restriction ;
        owl:onProperty sorel:hasRealm ;
        owl:allValuesFrom sorea:Geosphere
    ] ;
    rdfs:label "Geologic Feature"@en ;
    skos:definition "A Geologic Feature is a geospatial feature on the Earth that results from geological processes."@en ;
    skos:scopeNote "Geologic Features include Stratigraphic Units, Non-stratigraphic (Lithodemic) units, Stratigraphic Event Features, Tectonic and Structural Features, GeoResource Accumulations, and Geologically Significant Sites"@en ;
    dcterms:contributor [
        a foaf:Organization , sdo:Organization ;
        foaf:name "Geological Survey of Queensland" ;
        sdo:name "Geological Survey of Queensland" ;
        sdo:identifier <http://linked.data.gov.au/org/gsq> .
    ] ;

[nicholascar]

We - Geological Survey of Queensland - are planning to define a series of subclasses for this Class if this definition is accepted. Some of the subclasses are mentioned in the scopeNote already: Craton, Orogen, Shield etc.

[dr-shorthair]

@smrgeoinfo you may like to check the defs here

[lewismc]

A pull request would be really helpful @nicholascar :) If not, let me know and I can whip one up.

[smrgeoinfo]

your definition is broadly consistent with GeoSciML v4.0.1, so OK there, especially given the hodge-podge of things that are sub classes of 'geologic feature' in soreag :

The abstract GeologicFeature class represents a conceptual feature that is hypothesized to exist coherently in the world.

• this corresponds with a "legend item" from a traditional geologic map
• while the bounding coordinates of a Geologic Feature may be described, its shape is not.

[smrgeoinfo]

some other observations: all of the other classes have labels in lower case, so it would be odd if the label were 'Geologic Feature'. Alt label suggestion is fine.

I note that 'geologic structure' is a subclass of 'geosphere', I would argue that 'geologic structure' should be a subclass of 'geologic feature', consistent with GeoSciML usage and your text above. It seems to me that most of the subclasses of 'geologic feature' are defined in tectonic terms, so labeling the class 'tectonic feature' seems more appropriate to me. The conflation of partOf and kind of relations in the hierarchy is a problem, as well as duplication of concepts ('volcanic arc' vs. 'arc').

What's the benefit of defining more subclasses under 'geologic feature'?

[smrgeoinfo]

There is no definition of 'realm', but I'd guess the intention is something like definition 3 in Merriam Webster: 'a primary marine or terrestrial biogeographic division of the earth's surface', generalized to apply to any planet, in which case the I suppose subclassing 'geologic feature' like caldera, diapir, craton makes sense.

Changing the parent class from PlantaryRealm to 'GeometricObject' is a pretty big modification of the wordnet. Its a very 'GIS' kind of view--that the geometry is primary, rather than a feature oriented view in which location/shape are properties of the feature.

In short on that point, I don't thing changing the parent class makes sense.

[dr-shorthair]

I would also caution against seeing Geologic Feature as primarily a kind of GeometricalObject. That is not how classical (historical) geologists think - the geometry is really a secondary property, compared to the primary genetic conceptualization.

[nicholascar]

@lewismc A pull request would be really helpful @nicholascar :)

I will put in a PR once the issues raised here are addressed

@smrgeoinfo all of the other classes have labels in lower case, so it would be odd if the label were 'Geologic Feature'.

OK, I will rescind that proposal and stick to lower case. I'll add an upper case altLabel too

@smrgeoinfo I note that 'geologic structure' is a subclass of 'geosphere', I would argue that 'geologic structure' should be a subclass of 'geologic feature',

I agree but this is out of scope for this change request. When this is done, I propose to work on that and a range of others subclasses of Geologic Feature

@smrgeoinfo It seems to me that most of the subclasses of 'geologic feature' are defined in tectonic terms, so labeling the class 'tectonic feature' seems more appropriate to me.

Yes, I think you are correct. I'm working with our (Queensland's) geologists to make a whole tectonic and other class hierarchy but I want to insert it into SWEET here at Geologic Feature.

I think the way forward here is to generalise the text of Geologic Feature - make it not just about tectonics - and then to reveal the considered class hierarchy we've been working on.

@smrgeoinfo What's the benefit of defining more subclasses under 'geologic feature'?

Subclasses are actually used by the geologists?

@smrgeoinfo & @dr-shorthair "Changing the parent class from PlantaryRealm to 'GeometricObject' is a pretty big modification of the wordnet. Its a very 'GIS' kind of view..." & "I would also caution against seeing Geologic Feature as primarily a kind of GeometricalObject"

OK, I hear you both and will remove the link. I'm most concerned about having a definition for Geometric Object so propose to adopt GeoSciML's as per @smrgeoinfo's suggestion above.

In short on that point, I don't thing changing the parent class makes sense.

As above, I'll drop the geometry association but I still don't like the subClassOf relation to Realm. Sure, GeologicFeatures are within a realm but are not sub classes of it. The hasRealm property does, I think cater for the association without the need for a subclassing.

Some of the other Realm subclassing also seems to suffer from the same issue: the CriticalZone is within the PlanetaryRealm, not a subclass of it.

So I support the small hierarchy of:

• Realm
  • PlanetaryRealm
  • StellarRealm

But not the children of PlanetaryRealm, StellarRealm or that OuterSpace should be a child of Realm. All isPartOf in some way, not subClassOf.

Amended proposal

The propsal now really only adds a definition, mostly from GeoSciML. I've speculatively put in PlanetaryStructure as the parent IFF PlanetaryStructure can be split away from Realm.

@prefix sdo: <https://schema.org/> .

soreag:GeologicFeature 
    a owl:Class ;
    rdfs:subClassOf soreag:PlanetaryStructure ;
    [ 
        a owl:Restriction ;
        owl:onProperty sorel:hasRealm ;
        owl:allValuesFrom sorea:Geosphere
    ] ;
    rdfs:label "geologic feature"@en ;
    skos:altLabel "Geologic Feature"@en ;
    skos:altLabel "Geological Feature"@en ;
    skos:definition "A geologic feature is a conceptual feature a that is hypothesized 
                     to exist coherently in the Earth that results from geological 
                     processes"@en ;
    skos:scopeNote "Geologic Features include sedimentary basins, stratigraphic 
                    units, non-stratigraphic (lithodemic) units, stratigraphic 
                    event features, provinces, tectonic and structural features, 
                    georesource accumulations, and geologically significant sites,  
                    among others"@en ;
    dcterms:contributor [
        a foaf:Organization , sdo:Organization ;
        foaf:name "Geological Survey of Queensland" ;
        sdo:name "Geological Survey of Queensland" ;
        sdo:identifier <http://linked.data.gov.au/org/gsq> .
    ] ;

[dr-shorthair]

Regarding the arguments around sub-classing vs partition: it depends how the concept is to be used - individual vs class. The realms are mostly individuals (there is one terrestrial realm, one oceanic realm, one interstellar realm, etc), whereas feature/structure/geometricObject are clearly classes as they have many members. Different meta-levels.

So I agree that subsumption is almost certainly not the right relationship between realms and structures - i.e. this hierarchy is wrong.

(Visualization from WebProtege )

[dr-shorthair]

I suspect that the underlying cause is that subsumption was almost the only form of relationship between classes within SWEET from the very beginning. It allows views like the one above to be presented, but within which the semantics of the nesting slides rather misleadingly between sub-classing, instantiation, and partonymy.

i.e. I doubt this is the only part of SWEET that needs examining to check that subsumption is actually what is required by the science model.

[nicholascar]

Since the discussions above seems not to have issue with the annotation properties, I've created PR #182.

I'll close this Issue if/when that PR is merged.

[lewismc]

I updated https://github.com/ESIPFed/sweet/wiki/SWEET-Annotation-Convention... any comments?

[rduerr]

On the subject of multiple definitions, I am totally opposed to this! Entities with different definitions need different URI's and axioms! We need ontological mechanisms to say that thermokarst means blah, blah, blah to the permafrost community; while thermokarst means something entirely different to the geology community. In ENVO we did this by labeling the terms differently - thermokarst process vs thermokarst landform. However, it isn't at all clear that this will work in all cases. It would be convenient to be able to call thermokarst a synonym for both versions.

[lewismc]

Hi @rduerr can you please post your comment at https://github.com/ESIPFed/sweet/issues/183 where we can sort this out? Thank you. Have a great weekend :)