envo
envo copied to clipboard
EnvironmentOntology
Coordinate soil classifications
There are several systems out there, and they may be talking about the same soil using somewhat different criteria. To avoid representing the classification rather than the soil, we'll try to be conservative and use synonyms until there's good reason to believe the systems are referencing truly different entities.
I'm merging #331 #332 and #330 into this issue:
@celineaubert provided the following in the merged issues:
USDA orders of soil taxonomy are used in USA to classify soil. Could you please add the 12 orders of soil taxonomy :
AlfisolAndisolAridisolEntisolGelisolHistosolInceptisolMollisolOxisolSpodosolUltisolVertisol. Definitions are here or here. COuld you also add the in-subsetUSDA order of soil taxonomyfor these classes? Thanks.
Two of the FAO soil classes are not listed in ENVO: greyzem and podzoluvisol (different from podzol). link to the definition: [http://www.fao.org/fileadmin/user_upload/soils/docs/isricu_i9264_001.pdf]
The World Reference Base (WRB) is the international standard for soil classification system that replaced the FAO soil classification. The WBR soil groups are already in ENVO. Could you please add an in-subset World Reference Base soil groupfor the following soil groups: Acrisol Albeluvisol Alisol Andosol Anthrosol Arenosol Calcisol Cambisol Chernozem Cryosol Durisol Ferralsol Fluvisol Gleysol Gypsisol Histosol Kastanozem Leptosol Lixisol Luvisol Nitisol Phaeozem Planosol Plinthosol Podzol Regosol Solonchak Solonetz Stagnosol Technosol Umbrisol Vertisol?
The Australian Soil Classification [1] has the following top-level orders:
AN Anthroposols CA Calcarosols CH Chromosols DE Dermosols FE Ferrosols HY Hydrosols KA Kandosols KU Kurosols OR Organosols PO Podosols RU Rudosols SO Sodosols TE Tenosols VE Vertosols
Within each of these there is a detailed sub-classification system based on 'Families' and 'Classes' leading to hundreds (probably more than 1000) potential classifications. A turtle file containing the definitions expressed as skos:Concepts is attached [2]
[1] http://www.publish.csiro.au/book/7428
[2] orders.ttl.txt
Thanks! Would it be possible to include a dc:license/dc:rights on the ttl? Ideally we could do something like either bring definitions across (with provenance recorded as axiom annotation), or import this as a module
ah - that is an extract from a much bigger RDF which is still in prep. I wanted to make the names of the orders and the text in the skos:definition fields* available, because I think that is the likely source of a formal axiomatization similar to what has been done for the USDA orders.
The 'White Book' [1] is the ultimate authority. I sought and received permission from CSIRO Publishing to make the content available as an online resource. We haven't finished that yet, but I will take this conversation as a trigger to clarify the licensing.
*which is copied from the White Book.
[1] http://www.publish.csiro.au/book/7428
I've received permission from CSIRO Publishing and have added some metadata, including a license, in the attached. ASC-orders.ttl.txt
What is process to integrate these. Do you need a pull-request, or do you do the integration?
@cmungall can we do this en masse or do you think we should manually curate the alignment? It's well structured already, but alignment can be a tripping point.
The definition structure may need some restructuring (keeping the meaning and transferring verbatim descriptions to comments).
Addition of Australian Soil Classification appears to have stalled? Is there anything to do to allow it to progress?
Hi @dr-shorthair, I'm setting up some background semantics (solum, horizons, etc) to allow the teasing apart of the classifications. I'll unstall this.
?possible to unstall again And perhaps spin this out to a discrete issue?
Oof soil classification is a huge can of worms to open up. Encyclopedia of Soils in the Environment Volume 1 (which happens to be available on LibGen 👀) has a really great section on soil classification systems. The US, Australia, and Russia historically have been the most vocal about their own classification systems being the dominant ones. The FAO developed one to be used internationally and a lot of ~~mistakes~~decisions were made to appease the strange politics of international soil classification
The encyclopedia I mentioned goes into depth about all this as well as the pitfalls and all the factors considered in these systems. I think the main takeaway is that it really does seem impossible to come up with a truly fair and representative way to divide up all the different types of soils that exist. Especially now given how they're increasingly shaped by human interaction
It also doesn't help that soil science in general is one of the most slept-on fields of earth systems science and/or biology. A few years ago they disproved the myth of "humus".[0] An astounding finding that should have huge (mostly negative) ramifications for the IPCC and climate change forecasts. Yet there hasn't been a peep about it in the media and the IPCC and other climate model. It seems even most other scientists have trouble keeping up with soil sciences. GOOD LUCK
[0] https://www.mofga.org/resources/soil/humus-is-dead-long-live-humus/
Indeed. Which is why ENVO would be improved by including more than one classification. Currently it only reflects the North American system.
Then we might look at axiomatizing the relationships between the different classifications.
FWIW this SKOS-based service represents the latest Australian Soil Classification (ASC) (2021 edition). The hierarchies of the Orders and Sub-orders are correct, and the textual definitions of the various classifiers are taken straight from the reference book. But there is no attempt at formal axiomatization (yet).
Note that a new Order - RE (Arenosols) - was introduced, which replaces parts of the CA (Calcarosols), TE (Tenosols) and RU (Rudosols).
FWIW I suspect the different systems partly reflect the differences in distributions of different soil types in different environmental contexts. Soils are strongly affected by geological activity, relief, and climate. In Australia most of the geological activity underlying soil development was a loooong time ago, compared with North America, relief is generally low, and we haven't had any glaciers recently. Nutritionally, many Australian soils are like a tea-bag that has been used too many times ;-)
Its a bit like the old myth about the Inuit having many names for different kinds of snow (and the Welsh having names for many kinds of rain - joke.) which the rest of us don't have a need for.