20 years of 3D structural modelling - Laurent Ailleres (Monash University)
-
Upload
3d-modelling-interest-group -
Category
Science
-
view
117 -
download
1
description
Transcript of 20 years of 3D structural modelling - Laurent Ailleres (Monash University)
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 Chuiet al.,
2
1994 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.
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
The Workflow
Example: Agnew Wiluna Belt
TMI AGC
Field Work
Geophysical Processing
Example: Agnew Wiluna Belt
Example: Agnew Wiluna Belt
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
Structural 3D Modelling: the Falun Mine
3D Structural Analysis
Late Fault
Main late shear zone
(SZ3)
Early, // mineralisation, folded shear zone
(SZ1a)
Splaying off SZ1 probably during D3:
SZ3b
En-echelon tension gashes?
3D Structural Analysis
3D Structural Analysis - Reconstruction
S2
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
1981
Nod
dy
1982
Arc
Info
Ver
sion
1
1989
Goc
ad E
xplic
it Mod
eller
1993
Ear
thvis
ion Im
plicit
Mod
eller
2008
1997
Geo
mod
eller
Jess
ell, 1
981
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)
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
Better use of field data Alsop et al., 1996
(Near Moine Thrust)
Why should we care about folding… and more importantly ALL foliations
Next slide detail
Why should we care about folding… and more importantly ALL foliations
Looking South into Bermagui Heads,
stratigraphy dipping steeply West
Next slide detail
Topography
Topography
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!!!
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
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 ?
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
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?
Intelligent upscaling
Tom Carmichael
Data: Limousin, France
Dispersed
Clustered
’ Kent Distribution
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.
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
31
Funding… ? Thanks for asking…