Soil classifications structure

[dr-shorthair]

I've had a go at encoding part of the 3rd Ed. ASC in SKOS+:

https://github.com/ANZSoilData/def-au-asc/blob/master/rdf/ASC.ttl

Just the Tenosols, in order to figure out the structure.

First we can describe asc:TE using the text from ASC:

asc:TE
  a asc:Order ;
  skos:prefLabel "Tenosols" ;

  dcterms:description "This order is designed to embrace soils with generally only weak pedologic organisation apart from the A horizons, excluding soils that have deep sandy profiles with a field texture of sand, loamy sand or clayey sand in 80% or more of the upper 1.0 m. It encompasses a rather diverse range of soils which occur in many parts of Australia." ;

  skos:definition """Soils that do not fit the requirements of any other soil orders and generally with one or more of the following:
1. A peaty horizon.
2. A humose, melacic or melanic horizon, or conspicuously bleached A2 horizon, which overlies a calcrete pan, hard unweathered rock or other hard materials; or partially weathered or decomposed rock or saprolite, or unconsolidated mineral materials.
3. A horizons which meet all the conditions for a peaty, humose, melacic or melanic horizon except the depth requirement, and directly overlie a calcrete pan, hard unweathered rock or other hard materials; or partially weathered or decomposed rock or saprolite, or unconsolidated mineral materials.
4. A1 horizons which have more than a weak development of structure and directly overlie a calcrete pan, hard unweathered rock or other hard materials; or partially weathered or decomposed rock or saprolite, or unconsolidated mineral materials.
5. An A2 horizon which overlies a calcrete pan, hard unweathered rock or other hard materials; or partially weathered or decomposed rock or saprolite, or unconsolidated mineral materials.
6. B2 horizon with 15% clay (SL) or less1 , or a transitional horizon (C/B) occurring in fissures in the parent rock or saprolite which contains between 10 and 50% of B horizon material (including pedogenic carbonate).
7. A ferric or bauxitic horizon >0.2 m thick.
8. A calcareous horizon >0.2 m thick.""" ;

  rdfs:comment """It may be desirable to specify a minimum thickness for those A1 horizons which do not meet the requirements for a peaty, humose, melacic or melanic horizon. The inclusion of certain soils with conspicuously bleached A2 horizons may be questioned by some, but it is difficult to find a more appropriate place for them.

The Tenosols will differ from Rudosols by virtue of having either a more than weakly developed A1 horizon, an A2, or a weakly developed B horizon. As B horizons are difficult to identify consistently in some Tenosols, specific mention of a B horizon is omitted from some Suborders. They will obviously grade to Kandosols, and some difficulty may be experienced in separating medium-textured Tenosols from Kandosols. Here again, B horizon development is the key; Kandosols must have a clearly distinguishable, well-developed B2 horizon with more than 15% clay. Tenosols will grade to Podosols, but the latter must have a Podosol diagnostic B horizon. Tenosols also grade to Arenosols but lack the deep sandy horizons characteristic of this order. In cold, wet environments, some Tenosols with peaty A horizons will grade to Organosols.

This third edition of the Australian Soil Classification removes the Orthic term from the bleached and colour suborders.""" ;

  skos:notation "TE" ;
.

We also encode definitions of the classes, e.g. here's three of them:

asc:AT
  a asc:Class ;
  skos:definition "Soils with a conspicuously bleached A2 horizon." ;
  skos:notation "AT" ;
  skos:prefLabel "Bleached " ;
.

asc:IM
  a asc:Class ;
  skos:definition "Soils with a horizon containing more than 20% of fine earth carbonate that is at least 0.2 m thick." ;
  skos:notation "IM" ;
  skos:prefLabel "Calcenic" ;
.

asc:AA
  a asc:ColourClass ;
  skos:notation "AA" ;
  skos:prefLabel "Red" ;
.

Then we can add links (cross-references) from the Order to the applicable Suborders and Colour-classes, along with the comments about the great groups and sub-groups:

asc:TE
  a asc:Order ;
  skos:notation "TE" ;
  skos:prefLabel "Tenosols" ;
  asc:suborder asc:AT ;
  asc:suborder asc:AW ;
  asc:suborder asc:BE ;
  asc:suborder asc:BF ;
  asc:suborder asc:CY ;
  asc:suborder asc:IL ;
  asc:suborder asc:IM ;
  asc:colourclass-comment "The dominant colour class in the major part of the upper 0.5 m of the profile (or the entire profile if it is less than 0.5 m thick) is:" ;
  asc:colourclass asc:AA ;
  asc:colourclass asc:AB ;
  asc:colourclass asc:AC ;
  asc:colourclass asc:AD ;
  asc:colourclass asc:AE ;
  asc:greatgroup-comment "If a diagnostic feature in the key begins more than 1.5 m from the soil surface it may not have a significant impact on the performance of the soil. Refer to diagnostic features in the glossary for guidance on the use of such features in the classification." ;
  asc:subgroup-comment """These have been grouped into the various suborders, but not all subgroups will be appropriate for each great group of a particular suborder.

If a diagnostic feature in the key begins more than 1.5 m from the soil surface it may not have a significant impact on the performance of the soil. Refer to diagnostic features in the glossary for guidance on the use of such features in the classification.""" ;
.

However, it looks like the applicable Great Groups and Subgroups are specified at the Suborder level.
So in order to list these, the Suborders must be enumerated.
A Suborder is defined by adding a classifier to an Order like this:

asc:TE-AT
  a asc:Suborder ;
  skos:prefLabel "Bleached Tenosol" ;
  skos:broader asc:TE ;
  asc:classifier asc:AT ;
.

Then we can add cross-references to the applicable Great Groups and Subgroups

asc:TE-AT
  a asc:Suborder ;
  skos:prefLabel "Bleached Tenosol" ;
  skos:broader asc:TE ;
  asc:classifier asc:AT ;
  asc:greatgroup asc:AK ;
  asc:greatgroup asc:AO ;
  asc:greatgroup asc:AP ;
  asc:greatgroup asc:AS ;
  asc:greatgroup asc:BJ ;
  asc:greatgroup asc:BU ;
  asc:greatgroup asc:CZ ;
  asc:greatgroup asc:DD ;
  asc:greatgroup asc:DU ;
  asc:greatgroup asc:DZ ;
  asc:greatgroup asc:EA ;
  asc:greatgroup asc:EF ;
  asc:greatgroup asc:EM ;
  asc:greatgroup asc:FK ;
  asc:greatgroup asc:GE ;
  asc:greatgroup asc:GF ;
  asc:greatgroup asc:HF ;
  asc:greatgroup asc:IA ;
  asc:greatgroup asc:IS ;
  asc:greatgroup asc:JQ ;
  asc:greatgroup asc:JW ;
  asc:greatgroup asc:JX ;
  asc:subgroup asc:AI ;
  asc:subgroup asc:AR ;
  asc:subgroup asc:BC ;
  asc:subgroup asc:CK ;
  asc:subgroup asc:DC ;
  asc:subgroup asc:DG ;
  asc:subgroup asc:DK ;
  asc:subgroup asc:DW ;
  asc:subgroup asc:FC ;
  asc:subgroup asc:FU ;
  asc:subgroup asc:FV ;
  asc:subgroup asc:GU ;
  asc:subgroup asc:GY ;
.

At this point we have essentially reproduced the description of the Soil Order from the ASC book, but using cross-references to the Class descriptions, rather than by copying the text of the Class over multiple times.
So far so good.

The question now is: should we continue to build all the valid combos?
e.g. Great Groups in each Suborder, such as

asc:TE-AT
  a asc:Suborder ;
  skos:prefLabel "Bleached Tenosol" ;
  skos:narrower asc:TE-AT-AK ;
  skos:narrower asc:TE-AT-AO ;
  skos:narrower asc:TE-AT-AP ;
  skos:narrower asc:TE-AT-AS ;
  skos:narrower asc:TE-AT-BJ ;
  skos:narrower asc:TE-AT-BU ;
  skos:narrower asc:TE-AT-CZ ;
  skos:narrower asc:TE-AT-DD ;
  skos:narrower asc:TE-AT-DU ;
  skos:narrower asc:TE-AT-DZ ;
  skos:narrower asc:TE-AT-EA ;
  skos:narrower asc:TE-AT-EF ;
  skos:narrower asc:TE-AT-EM ;
  skos:narrower asc:TE-AT-FK ;
  skos:narrower asc:TE-AT-GE ;
  skos:narrower asc:TE-AT-GF ;
  skos:narrower asc:TE-AT-HF ;
  skos:narrower asc:TE-AT-IA ;
  skos:narrower asc:TE-AT-IS ;
  skos:narrower asc:TE-AT-JQ ;
  skos:narrower asc:TE-AT-JW ;
  skos:narrower asc:TE-AT-JX ;
.

where, each Great Group is defined by adding an additional classifier to a Suborder. For example

asc:TE-AT-AK
  a asc:Greatgroup ;
  skos:prefLabel "Andic Bleached Tenosol" ;
  skos:broader asc:TE-AT ;
  asc:classifier asc:AK ;
.

and how far we can go with that?
Does it makes sense to 'pre-cache' all allowed combos, at least down to Subgroup level.
For example, asc:TE-AW-BU-AI would be Acidic Ferric Bleached-Leptic Tenosol, with the code position within the sequence indicating its role - here Order=TE, SubOrder=AW, Great Group=BU, Subgroup=AI.

This wouldn't work if Subgroups can be used without a Great Group, or either of those without as Suborder.

Or am I barking completely up the wrong tree here?

[meganrwong]

Great! May be useful for you to know that soil profile mapping we've used 'Austalian Soil Classificaion Order' and 'Australian Soil Classificaion Suborder' as Observable Properties

[dr-shorthair]

Makes sense.
Haven't built any collections or higher-level things yet, just asc:Order and asc:Suborder etc as Classes.
But it is quite OK for those URIs to be used for observed-properties.

[dr-shorthair]

@ljgregory can we have a chat about this?

[meganrwong]

Simon - in response to your query re what FedUni have encountered, to help consider whether to create concepts for each combination of Order, Suborder, Great Group, Sub Group etc in ASC

We've mapped only 2 soil profile datasets -

In our DB we are storing Order, Suborder, Great Group, Sub Group as each as different sets of results in database (ie imported as separate columns of data for each of Order, Suborder, Great Group, Sub Group)

FYI, the soil scientists in their field sheet (excel) and thesis (pdf) as one result -
One reported Order and Sub Orders together (as one result)
One reported Order, SOrder, GGroup, Sgroup together as one result eg "SO AE ES AH "

[meganrwong]

In one dataset I see also results for "ASC-CONF" - the results all look to be a numeric classifier eg "4".
No idea what this is.
@bsimons14 ?