[auscope-geosciml] [Auscope-geosciml] Particle geometry [SEC=UNCLASSIFIED]

stephen richard steve.richard at azgs.az.gov
Fri Sep 17 10:59:37 EDT 2010


Skewness and kurtosis are properties of a frequency distribution. In the 
case of particle geometry, the 'frequency' is the frequency of particles 
in a given size-range. So one could argue they are more an aspect of 
sorting (which has to do with grain-size distribution) that a grain size 
property that attempts to characterize the particle size range.

I agree that we can't/don't want to try and hard type all the possible 
descriptive properties that might be used to describe particle 
shape/size/size distribution, or geophysical characteristics, structural 
morphology, chemical character etc. Isn't the philosophy to hard type 
the lowest common denominator collection of properties to facilitate an 
80/20 solution. Using this approach, we have to make a seat of the pants 
call on what properties should be hard vs. soft typed. To do this we 
also need a generic property hook for soft typed properties that is 
associated with GeologicFeature (or GeologicUnit, CompoundMaterial, 
GeologicStructure), and a profile that dictates that properties that are 
hard typed may not be encoded using the soft-typed property (i.e. SWE).  
Interoperability for the soft typed properties would require conventions 
on which property vocabulary is used and how values are specified (term, 
term range, number, number range....)

On 6/4/2010 1:52 AM, Laxton, John L wrote:
>
> Thanks for all the thought on this! This looks like a workable 
> solution but I have a couple of reservations:
>
> 1. Are skewness and kurtosis actually 'size' as opposed to statistical 
> measures derived from a range of size measuements? Sorting is derived 
> from size measurements after all. We would have to redefine what we 
> mean by size if we adopt this approach.
>
> 2. This doesn't answer Ollie's point that these parameters, and most 
> of the other properties in ParticleGeometryDescription, can be 
> described numerically or with terms (which is why they were originally 
> CGI_Value….). In fact the scope note for size states 'Values may be 
> reported using absolute measurements (eg: range, mean, median, mode, 
> maximum) or as descriptive terms from a schema appropriate to the type 
> of Compound Material'.
>
> 3. While skewness and kurosis are statistical measures that can be 
> applied to any type of data in my experience they are most commonly 
> used in geoscience to describe particle geometry, so having these as 
> specific properties in ParticleGeometryDescription might be desirable 
> - although that would mean a change to the model.
>
> John
>
> *From:* auscope-geosciml-bounces at lists.arcs.org.au 
> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] *On Behalf Of 
> *Bruce.Simons at dpi.vic.gov.au
> *Sent:* 04 June 2010 03:42
> *To:* auscope-geosciml at lists.arcs.org.au
> *Subject:* Re: [Auscope-geosciml] Particle geometry [SEC=UNCLASSIFIED]
>
>

snip...
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengeospatial.org/pipermail/geosciml/attachments/20100917/c7708c37/attachment.htm>


More information about the GeoSciML mailing list