Overview of Alignment of CORVEG to TERN Plot Ontology - Archival Page

Owned by Edmond Chuc (Unlicensed)
Last updated: 18 Feb 2022 by TERN Data
2 min read

The mapping described here are for old CORVEG system and mapping to new CORVEG is available in https://ternaus.atlassian.net/wiki/spaces/CDI/pages/2058223625 .

This alignment relates the (Queensland) CORVEG logical model to the O&M conceptual model.

CORVEG reports a large set of observations of different properties of a site. The total number of potentially observed properties is large. The description of the site itself is grouped into thematic groups named Geology, Disturbance, Situation etc. The description of the vegetation is organized by strata. The description of the taxa at a site is generally organized by taxon, and also by strata.

Principles of alignment of CORVEG to O&M

Following the standard observation model:

  • every 'value' is recorded as the hasResult of an Observation

  • every Observation is related to a FeatureOfInterest

  • the type of value is recorded as the Observation's observedProperty. It is taken from a controlled-vocabulary of observable properties

  • Observations may be grouped in an ObservationCollection which is a set of observations with (usually) a common hasFeatureOfInterest and phenomenonTime

  • each individual Observation is nested as a ssn-ext:hasMember of the ObservationCollection

The templates on the TERN Plot Ontology description illustrate the application pattern.

Advantage of Using a Generic Model

Since the observation type is indicated by properties of the observation (i.e. the observedProperty and hasFeatureOfInterest), the generic classes Observation and ObservationCollection can be used for all observed data. The controlled vocabulary of ObservableProperties is maintained separately, and can be updated frequently, without having to touch the main data structures.

We could have defined a sub-class of ObservationCollection for each CORVEG table (i.e. thematic group), but this would tie the ontology directly to CORVEG, which is undesirable because

  1. the CORVEG conceptual model does not match all the data submitted by the various providers, while TERN-plot - which is a lightly extended SSN/SOSA model - is flexible enough to cope

  2. too much specialization/customization compromises compatibility with observation data from other sources or providers, or from other discipline areas

Where an ObservationCollection does correspond with a row from a CORVEG table, the name of the table can be indicated using the dct:type property. This may assist discovery and processing by users or applications who are familiar with CORVEG. A list of observation groups is maintained as a TERN-plot 'controlled vocabulary'.

Detailed Implementation

Detailed implementation of CORVEG entities in RDF can be found under Table Mappings.

CORVEG Controlled Vocabularies

CORVEG controlled vocabularies can be viewed at the persistent URI http://linked.data.gov.au/def/corveg-cv.

Datatypes

Observation results are encoded using a set of standard data-Datatypes