Machine learning research may help find new tungsten deposits in SW
England

Date:
August 7, 2020
Source:
University of Exeter
Summary:
Geologists have developed a machine learning technique that
highlights the potential for further deposits of the critical
metal tungsten in SW England.



FULL STORY ========================================================================== Geologists have developed a machine learning technique that highlights
the potential for further deposits of the critical metal tungsten in
SW England.


========================================================================== Tungsten is an essential component of high-performance steels but global production is strongly influenced by China and western countries are
keen to develop alternative sources.

The work, published in the leading journal Geoscience Frontiers, has
been led by Dr Chris Yeomans, from the Camborne School of Mines, and
involved geoscientists from the University of Nottingham, Geological
Survey of Finland (GTK) and the British Geological Survey.

The research applies machine learning to multiple existing datasets
to examine the geological factors that have resulted in known tungsten
deposits in SW England.

These findings are then applied across the wider region to predict areas
where tungsten mineralisation is more likely and might have previously
been overlooked. The same methodology could be applied to help in the exploration for other metals around the world.

Dr Yeomans, a Postdoctoral Research Fellow at the Camborne School of
Mines, based at the University of Exeter's Penryn Campus in Cornwall said: "We're really pleased with the methodology developed and the results of
this study.



==========================================================================
"SW England is already the focus of UK mineral exploration for tungsten
but we wanted to demonstrate that new machine learning approaches may
provide additional insights and highlight areas that might otherwise
be overlooked." SW England hosts the fourth biggest tungsten deposit in
the world (Hemerdon, near Plympton), that resulted in the UK being the
sixth biggest global tungsten producer in 2017; the mine is currently
being re-developed by Tungsten West Limited.

The Redmoor tin-tungsten project, being developed by Cornwall Resources Limited, has also been identified as being a potentially globally
significant mineral deposit.

The new study suggests that there may be a wider potential for tungsten deposits and has attracted praise from those currently involved in the development of tungsten resources in SW England.

James McFarlane, from Tungsten West, said: "Tungsten has only been of
economic interest in the last 100 years or so, during which exploration
efforts for this critical metal have generally been short-lived.

"As such is very encouraging to see work that aims to holistically
combine the available data to develop a tungsten prospectivity model
in an area that has world-class potential." Brett Grist, from Cornwall Resources added: "Our own work has shown that applying modern techniques
can reveal world-class deposits in this historic and globally-significant mining district.

"Dr Yeomans' assertion, that the likelihood of new discoveries of
tungsten mineralisation may be enhanced by a high-resolution gravity
survey, is something in which we see great potential.

"Indeed, such a programme could stimulate the new discovery of
economically significant deposits of a suite of critical metals, here
in the southwest of the UK, for years to come."

========================================================================== Story Source: Materials provided by University_of_Exeter. Note: Content
may be edited for style and length.


========================================================================== Journal Reference:
1. Christopher M. Yeomans, Robin K. Shail, Stephen Grebby, Vesa
Nyka"nen,
Maarit Middleton, Paul A.J. Lusty. A machine learning approach to
tungsten prospectivity modelling using knowledge-driven feature
extraction and model confidence. Geoscience Frontiers, 2020; DOI:
10.1016/j.gsf.2020.05.016 ==========================================================================

Link to news story: https://www.sciencedaily.com/releases/2020/08/200807102332.htm

--- up 3 weeks, 2 days, 1 hour, 55 minutes
* Origin: -=> Castle Rock BBS <=- Now Husky HPT Powered! (1337:3/111)