Spatial Patterns in Gold Deposits for Southeastern Australia

All images and text are Copyright © 2005 by Larry Robinson

PowerPoint 6-2

Prepared as part of:

The Spatial and Temporal Distribution of theMetal Mineralisation in Eastern Australia

and the Relationship of the Observed Patternsto Giant Ore Deposits

byLarry Robinson

A thesis submitted for the degree of

PhD

2006

The Geology of Southeastern Australia, The Lachlan Fold Belt1

In the Lachlan Fold Belt the rock types are predominantly Ordovician-Silurian granites, volcanics and sediments.

Pattern Recognition using a Clustering Algorithm

The AUTOCLUST AlgorithmThe approach automatically extracts boundaries based on Voronoi modellingand Delaunay Diagrams.

Parameters are not specified by users in the automatic clustering.

All clustering operations takes place on the Delaunay Diagram where data points become vertices and edges connect pairs of points to model spatial proximity.

Estivill-Castro & Lee (2000)2

The AUTOCLUST Algorithm

The m value shown in the following slides designates the lengths of edges, in units of the standard deviation away from the Global Mean, which are included in a single cluster. The Global Mean is determined from edges in the entire data set.

The noise index of a point (p) is the ratio - Local Mean / Global Mean - where, Local Mean is determined from the mean of points with edges incident to that point (p). The noise index is a measure of the deviation of any single point from proximal points. The mean of edges incident to noise tends to be significantly greater than the Global Mean.

The m value can be used as an exploration tool.Larger values of m causes relatively close clusters to merge into the same cluster.

Smaller values of m result in clusters with more homogeneous lengths of edges.

Smaller values of m identifies relatively high-density clusters and homogeneous clusters, but declares relatively sparse clusters asnoise and heterogeneous clusters as potential outliers.

The user may reduce the value of m to find breakable or vulnerable regions in clusters where they are about to split.

Southeastern Australia’s 22,240 Gold Deposits 3 & 4

The following images show Polygonization of gold deposits in southeastern Australia at various m values with 0% noise.

These images progress at a predetermined timing from m = 0.5 to m = 4.0.

m = 0.5

m = 0.6

m = 0.7

m = 0.8

m = 0.9

m = 1.0

m = 1.1

m = 1.2

m = 1.3

m = 1.4

m = 1.5

m = 1.6

m = 1.7

m = 1.8

m = 1.9

m = 2.0

m = 2.1

m = 2.2

m = 2.3

m = 2.4

m = 2.5

m = 2.6

m = 2.7

m = 2.8

m = 2.9

m = 2.9

m = 3.0

m = 3.0

m = 3.1

m = 3.2

m = 3.3

m = 3.4

m = 3.5

m = 3.6

m = 3.7

m = 3.8

m = 3.9

m = 4.0

m = 4.0

By combining the images at various m values (1.6 and 2.9 in the following INSET) it is possible to discern the fine scale clustering in southeastern Australia.

Each cluster is revealed at some optimum m value, after which information is actually lost.

The ~80 km clustering is revealed for the Ophir, Trunkey Creek & Parkes-Forbes radial patterns.

The following images show clustering of gold deposits in southeastern Australia at various m values with 1% noise.

These images are not timed.

Clustering of Gold Deposits SE Australia, m=3.5, noise=1%

Clustering of Gold Deposits SE Australia, m=3.5, noise=1%

Polygonization of Gold Deposits SE Australia, m=3.5, noise=1%

The mean distance between these clusters is ~230 km.

Note the bulge in the synclinorial zones.

This bulge is ~240 km across.

This is the northeastern part of the Lachlan Fold Belt located in New South Wales.5

The Captains Flat - Goulburn Synclinorial Zone is concave to the east.

The Tumut Synclinorial Zone is concave to the west.

This bulge incorporates the Ophir, Trunkey Creek, Gundagai and other clusters.

The clusters and bulges at Ophir, Trunkey Creek, and Gundagai are all ~80 km across.

See PowerPoint presentation 6-1 for Ophir & Trunkey Creek and 6-3 for Gundagai.

It is proposed that there are, at least, two scales of self-similar patterning (~80 km and ~230 km) in the Lachlan Fold Belt of southeastern Australia.

Thank YouThe End

References

1. AGSO Geology Map of Australia, 1982

2. Estivill-Castro, V. and I. Lee. AUTOCLUST: Automatic Clustering via Boundary Extraction for Massive Point-data Sets. In Proceedings of the 5th International Conference on Geocomputation, 2000. to appear. Extended version is available at http://www.cs.newcastle.edu.au/Dept/techrep.html as a technical report.

3. New South Wales Mineral Exploration Data Package, 2004, Sydney, New South Wales Department of Mineral Resources, Australia, 1 CD-ROM.

4. Victoria Geoscientific Data CD, April 2003, Geological Survey of Victoria, Australia, 2 CD-ROM.

5. Suppel, D. W., and Scheibner, E., 1990, Lachlan Fold Belt in New South Wales - regional geology and mineral deposits, in Hughes, F., ed., BOOK (2 Volumes), Geology of the Mineral Deposits of Australia and Papua New Guinea, AusIMM Monograph No. 14, p. 1321-1328.