This presentation was delivered at the June 10 (2014) 3D Interest Group Meeting at the Centre for Exploration Targeting, UWA.

1. School of Geosciences
20 years of 3D structural modelling.
Where are we at? Where next?
In alphabetical order:
Laurent Ailleres, Thomas Carmichael, Eric deKemp, Lachlan Grose, Mark Jessell, Vitaly Kolin,
Gautier Laurent, Mark Lindsay, Cheng Yu Chui
et al.,

2. 1994
2
 My first voxet! Attempting to interpolate orientations of finite strain
ellipse in the French Alps and finite strain parameters

3. Seriously… a lot of explicit modelling was
done… and still is and still will be.
3
at multiple scale: from plate
scale on spheroid to the mine scale.
for different purposes:
green field exploration, near mine
exploration, hydrogeology, oil &
gas, research.
3D Model of Mt Isa Inlier, Qld,
Australia - After Murphy et al. (2008).
3D Model of Barwon Head aquifers, Nury et al., 2010

4. The Workflow

5. Example: Agnew Wiluna Belt
TMI AGC
Field Work
Geophysical Processing

6. Example: Agnew Wiluna Belt

7. Example: Agnew Wiluna Belt

8. Structural 3D Modelling: the Falun Mine
•Model built from 35+ level maps (average spacing 25m)
and 50+ sections from historical underground mapping
•Aim was to understand the structural framework of the
deposit

9. Structural 3D Modelling: the Falun Mine

10. 3D Structural Analysis
Late Fault
Main late shear zone
(SZ3)
Early, // mineralisation,
folded shear zone
(SZ1a)
Splaying off SZ1
probably during D3:
SZ3b

11. En-echelon
tension
gashes?
3D Structural Analysis

12. 3D Structural Analysis - Reconstruction
S
2

13. Explicit Modelling
•Non reproducible
•Labour intensive (only one model is built – no update please)
•Rarely utilised for anything more than visualisation
•Near impossible to play “what if” scenarios
•Highly uncertain and uncertainty cannot be characterised

14. Knowledge Data Data + Knowledge Data + Knowledge + Uncertainty
(My) Short history of 3D modelling
(strangely similar to Mark Jessell’s)
Perrouty et al., (2012; 2014)
Armit et al., (2012; 2014)

15. Implicit Modelling
• Reproducible
• Labour intensive - data need cleaning before modelling (nothing new!)
• Geological simulations are possible
• Allows uncertainty and geodiversity work
•Highly uncertain and uncertainty CAN be characterised
HOWEVER:
•Still does not use the full range of structural elements measured in
the field

16. Better use of field data Alsop et al., 1996
(Near Moine Thrust)

17. Why should we care about folding… and
more importantly ALL foliations
Next slide detail

18. Why should we care about folding… and
more importantly ALL foliations
Looking South into
Bermagui Heads,
stratigraphy dipping
steeply West
Next slide detail

20. Topography

21. Topography

22. Topography
Unless drawn or over determined by lots of structural measurements
(strike and dip), it is impossible to interpolate scalar field looking like
below, and this is with only 2 deformation events!!!

23.  Multiple generations (f1 + f2)
– strategy:
• find f2 fold parameters
• interpolate s1
• find f1 fold parameters
• interpolate bedding
f1 f1+f2f2
+ =
time
modelling
Implicit folding!!
S2
S1S1 S0

24. Implicit Modelling of Folds
 Motivations:
– What the implicit packages can constrain:
• Value/orientation at a location
• Smooth variations (gradient orientation and norm)
few constraints on fold structural parameters
• In StructuralLab:
– Axial Surface: grad f axial surface
– Fold Axis : grad f fold axis direction
 Yes, because we actually measure more in the field than
what is used
A good start
But can we use more ?

25. Axial surface field Foliation field
 Foliation field :
– // axial surface
– Relatively consistent over the whole area
– Vergence helps a lot
Next antiform
is this way!S1
S0

26.  General idea:
• Interpolate foliation field
• Interpolate vergence intensity (fold intensity ~ fold amplitude
and interlimb angle)
• Constrain “folded surface” orientation
• Is there a need for upscaling of bedding and structural data?

27. Intelligent upscaling
Tom Carmichael
Data: Limousin, France
Dispersed
Clustered
’
Kent Distribution

28. Model Comparison
Eastern Cross Section
Northern Cross Section
Modified model
Modified model
Initial Model
Initial Model
1389 Structural measurements 198 Structural “measurements”
15% of original data

29. Conclusion
 How far have we come? Not that far until implicit modelling packages
appeared. Since then, amazing progress has been made especially for
layer cake type geometries (basins… mainly).
 Complicated structures still have to be drawn…
 However:
 Where are we at? A very interesting spot! We could be the first group
of people to implement complete implementation of full structural
elements into implicit schemes.
 Where next? Uncertainty!!! Geodiversity, test concepts on topologically
varying models
 Fully constrained inversions including all input geological/structural
parameters as part of geology penalty function.
29

30. The Monash Group
 Laurent Ailleres - concepts development & mentoring
 Gautier Laurent – implicit folding & REED; including implementations
and concept development
 Thomas Carmichael – PhD candidate, upscaling of structural data
 Lachlan Grose – PhD candidate, uncertainty & geodiversity of poly-
deformed terranes
 Yu Ching Chui (Ching) – Honours student, uncertainty & geodiversity of
poly-deformed terranes using simple Noddy
models
 Vitaly Kolin – Honours student (to start in Aug), Uncertainty &
geodiversity of poly-deformed terranes,
applications to the Davenport terrane
30

31. Funding… ? Thanks for asking…
31