"Seegrid will be due for a migration to confluence on the 1st of August. Any update on or after the 1st of August will NOT be migrated"

Material Description

XML implementation of NADM Earth Material model

The sketches developed below in the XMML project are likely to be superceded by the more comprehensive model that has been developed in the NADM project. The encoding of this in XML has been advanced significantly and is recorded in the NADM subsection of the CGI International Collaboration TWiki. Note that this implementation follows GML/XMML patterns closely.

-- SimonCox - 21 Jun 2004

XMML Implementation

  • XMML Material model:

-- SimonCox - 09 Sep 2003

NADM model for Earth Materials

BoyanBrodaric has pointed to the NADM C1 Conceptual Model which includes a model for "Earth Material" as Plate 2 (page 43).

  • NADM C1 Earth Material Model:

Note that the UML idiom used here deviates slightly from the idiom used in ISO/TC 211 specifications. The main issue is that several associations carry the same name - "has". This could be problematic to systems that see class-attributes and association-names as being essentially equivalent (they certainly are at a logical level, and particularly when the the association is a "composition"). Repeating a name, but with a different type, could lead to clashes. -- SimonCox - 13 Mar 2003

You are right about role names; they're missing because we simply have not got to it as we've been struggling to identify the main concepts (classes).

In implementation the role name would become the name of the slot in the class (i.e. an attribute or an association). we do need them, but our concerns so far have been mainly conceptual... -- BoyanBrodaric - 14 Mar 2003

The NADM model is quite clean and adheres to some good basic principles. It includes both "EarthMaterial" and "CompoundMaterial".

  1. pure species Mineral and Fluid are specialised from "EarthMaterial", and thus inherit just the generic material properties (Color, Chemistry, Age, Physical Property (!) etc)
  2. Rock and Unconsolidated [material] are explicitly derived from Compound Material, and thus may also carry certain properties which only apply to composites, including Constituents, Fabric (deformation related?), Texture (particle size and shape descriptions) and properties related to Voids and Consolidation

Of course Compound Material may be composed of both simple and compound materials (e.g. a rock may contain both minerals and rocks as components). Some other aspects of the model:

  • the Constituent relationship is modelled as an "association class" which allows it to carry a number of secondary properties in turn, concerning the relationship of the component to the whole
  • an (optional) "genesis" property is provided for any material, which has input and output materials. The cardinality of these is not indicated in the diagram. But not clear to me if/how the "present" material and the "output" of the Genesis property are distinguished.

This all looks fairly powerful, but also very generic, which can be a challenge in determining best practice. Looking further through the documentation, there a lot of multiple-inheritance, and some use of "association classes". Although these are respectable UML, these aspects are not so easily or directly mapped onto XML schema structures.

Nevertheless, I've made a start on encoding this model in XMML, plus some additions from the discussion at ARRC and the WMC model. See material.xsd in the CVS.

-- SimonCox - 17 Mar 2003 -- SimonCox - 16 Jun 2004

A comment, re: multiple inheritence in NADM: the use of multiple inheritence in NADM is restricted to a relatively small fragment of the model dealing with fossils, and was adopted because it really fit with how some fossil concepts are used in practice.

-- BoyanBrodaric - 18 Mar 2003

Views of rock properties

There are various legitimate views of rock/mineral properties. For example:

  1. the "definitive" values associated with named materials in a catalogue or database - this is the view most useful when constructing an input deck for forward-modelling (e.g. pmd*CRC)
  2. the "observation" of a set of values on a target specimen by a fieldy or downhole tool (for example) - this is the view of most interest to the companies while capturing the results in the field

I'm trying to figure out how to support both of these with maximum re-use. Some thoughts

  • An XML Schema trick I intend to explore is bundling some of the pieces into named content-model groups, which makes them more easily re-usable in a pseudo-multi-inheritance manner.
  • A model idea I'm thinking about is to lighten up the "Observations and Measurmeents" model so that the properties associated with the act of observation may be attached to other properties in a more flexible manner.

-- SimonCox - 18 Mar 2003

Other Related work

For discussion of borehole models, see XMML BoreHole. -- SimonCox - 14 Mar 2003

See NADM model above. -- SimonCox - 14 Mar 2003

Notes from meeting held at CSIRO, 2003-03-13.

Here are the two visio images I (Craig) captured during our meeting:



(Craig 14 March 2003)

  • This paper shows the sort of data exploration people want to do with lithological data. It's dated 2001 - I wonder how much work GOCAD has done since then on data transfer (see page 7), given their market acceptance. http://www.geo.tu-freiberg.de/~apelm/gocad/cancun.pdf

I had lunch with Marcus in Cancun. At that time they had a rather rigid view of what "GIS" was. No contact since. -- SimonCox - 14 Mar 2003

  • Here's another lithological legend - not interesting in itself, but more of a check to see that there's consistency in our thinking. (Consistency with what others do in practice, that is). There does seem to be one "description" we omitted at our workshop, or gave a different name - "Rock composition". What is the geo's usual terminology? http://geoinfo.usc.edu/gvdc/Workshop_104501/Papers/WS1-5_final.pdf (Rob F)

From my quick read (there's 20 pages) I believe we have covered the 'Rock Composition' with our mineralogy component section (Nige)
Topic attachments
I Attachment Action Size Date Who Comment
NADM_earth_material.pngpng NADM_earth_material.png manage 26.9 K 14 Mar 2003 - 13:17 SimonCox Earth Material model from NADM
RockDescriptionImage1.gifgif RockDescriptionImage1.gif manage 9.8 K 14 Mar 2003 - 14:02 CraigMorley General diagram showing rock description model
RockDescriptionImage2.gifgif RockDescriptionImage2.gif manage 12.8 K 14 Mar 2003 - 14:12 CraigMorley This is the content of the 2nd white board
RockDescriptionImage3.gifgif RockDescriptionImage3.gif manage 14.8 K 14 Mar 2003 - 14:17 CraigMorley A slightly larger version of the white board image
material.pngpng material.png manage 12.2 K 09 Sep 2003 - 20:09 SimonCox XMML Material model
Topic revision: r23 - 15 Oct 2010, UnknownUser

Current license: All material on this collaboration platform is licensed under a Creative Commons Attribution 3.0 Australia Licence (CC BY 3.0).