UBC-GIF: Capabilities for EM Modelling Modelling and ... · – Conductivity (indicative of water)...
Transcript of UBC-GIF: Capabilities for EM Modelling Modelling and ... · – Conductivity (indicative of water)...
slide 1© UBC-GIF 2003
The University of British ColumbiaThe University of British Columbia Geophysical Inversion FacilityGeophysical Inversion Facility
UBCUBC--GIF: Capabilities for EM GIF: Capabilities for EM ModellingModelling and and Inversion of LSBB dataInversion of LSBB data
http://http://www.eos.ubc.ca/ubcgifwww.eos.ubc.ca/ubcgif
Douglas W. OldenburgDouglas W. Oldenburg
Department of Earth and Ocean SciencesDepartment of Earth and Ocean Sciences
June 12, 2009
slide 2© UBC-GIF 2003LSBB: Karst Aquifer Characterization
• Principle Questions:
– What is the underground matrix– Hydrogeologic properties – Storage capacity – Production capacity– Potential for pollution – Sustainability
• What is the role of geophysics?
slide 3© UBC-GIF 2003
Framework for Applied Geophysics• What is the question to be answered?
• What are the diagnostic physical properties?
DensityMagnetic susceptibilityElectrical conductivity
Electrical permittivityElastic parameters
• Choose survey type• Collect data• Invert data to get physical property model• Interpret model and synthesize with other data
slide 4© UBC-GIF 2003Geophysical Experiment
DC surveyData: ”pseudosection”
Data images cannot be directly interpreted in terms of geologyData must be inverted
• Physical property: Electrical conductivity• Survey: DC Resistivity• Collect data
slide 5© UBC-GIF 2003What is Inversion?
Inversionprocessing
Model Inversion estimates Earth models based upon data and prior knowledge.
??
Data
Measurements over the Earth are data.
slide 6© UBC-GIF 2003
The inverse problem
• Geophysical data are: F[m] +
= d– m: model --- unknown– F: forward mapping operator– : errors– d: observations (data)
• Given:– data, errors, a forward modelling method
• Find:– the model that generated measurements.
• Major Difficulty: Nonuniquenes
??F-1
slide 7© UBC-GIF 2003
Inversion as optimization: 3 parts
Inversion as optimization:
= d
+ m
.
0 < <
is a constant
Choose such that d
< Tolerance
A priori information: reference model, structural detail...
...)()()(2
020
dvxmmdvmmm
Sx
Ssm
Model objective function:
• s , x … constants• m0 : reference model
Misfit: 2
1
][
N
i i
obsii
ddmF
• i : standard deviation
slide 8© UBC-GIF 2003Numerical solution
• Discretize: Divide the earth into M cells of constant physical property (M>>N).
2
0
2)()][( mmWdmFW m
obsd
md
• Minimize
• Use the Gauss-Newton method for solution.
• Solving for :- Discrepancy principle.- GCV.- L-Curve.
1
M
slide 9© UBC-GIF 2003Numerical solution (Gauss-Newton method)
• Minimize
• set
20
2)(][ mmWdmF obs
md
0)()][()()( 0
mmWWdmFmJm
mg TobsT
j
iij m
dJ
J is the sensitivity matrix:
slide 10© UBC-GIF 2003Numerical solution (Gauss-Newton method)
• Minimize
• set
20
2)(][ mmWdmF obs
md
0)()][()()( 0
mmWWdmFmJm
mg TobsT
j
iij m
dJ
J is the sensitivity matrix:
mmJmFmmF )(][][ • Expand forward operator (dropping higher order terms):
slide 11© UBC-GIF 2003Numerical solution (Gauss-Newton method)
)()()( kT
kT
k mgmWWmJmJ • where mk
is the model at the kth iteration.• This is an M x M system of equations.
(M = # model parameters, or cells)
• Solve:
0)()][()()( 0
mmWWdmFmJm
mg TobsT
mmJmFmmF )(][][ • Expand forward operator (dropping higher order terms):
• Minimize
• set
20
2)(][ mmWdmF obs
md
j
iij m
dJ
J is the sensitivity matrix:
slide 13© UBC-GIF 2003Inversion Capabilities
• Gravity (3D)
• Magnetics (3D)
• DC resistivity and IP (2D and 3D)
• Frequency domain EM (1D and 3D)
• Time domain EM (1D and 3D)
Software for inversion is distributed world wide through UBC and third party vendors
3D UTEM
--2150 -1550
0
200
400
600
800
slide 14© UBC-GIF 2003
“Keel”
Tertiary BrecciaMafic Volcanics
Quartz Rhyolit
e
Massive Sulphide
Mafic Volcanics
Ele
vatio
n (m
) 2000
1600
-2000 -1100Easting (m)
LocationGeologic cross section
Physical properties
Field Example: San Nicolas Deposit
Unit Density Susceptibility Resistivity Chargeability(g/cc) (S.I. x10) (ohmm) (msec)
Qal 2 0 50 5Tv 2.3 0 20-30 10Mst./Lst. 2.4 0 150 20Mafic Vol. 2.7 0 80 30Mafic/IntVol. 2.7 0 80 30Sulphide 3.5 10 20 200Qtz. Rhyolite 2.4 0 100 10Graphitic Mst. 2.4 0 100+ 30
Unit Density Susceptibility Resistivity Chargeability(g/cc) (S.I. x103) (ohm- m) (msec)
Qal 2 0 50 5Tv 2.3 0 20 -30 10Mst./Lst. 2.4 0 150 20Mafic Vol. 2.7 0 80 30Mafic/IntVol. 2.7 0 80 30Sulphide 3.5 10 20 200Qtz. Rhyolite 2.4 0 100 10Graphitic Mst. 2.4 0 100+ 30
- 10(20)
- 5- 5
- 10- 5
- 30- 40- 50- 50
- 20- 70
slide 15© UBC-GIF 2003Gravity
Nor
thin
g (m
)
Observed Data
-2900 -600
600
-120
0
Easting (m)
San Nicolas
2.3
1
-0.4
Predicted Data
-2900 -600
600
-120
0
Easting (m)
Nor
thin
g (m
)
2.3
1
-0.4
Gravity data collected at the San Nicholas deposit.
slide 16© UBC-GIF 2003Gravity Inversion Results
-0.22
0.07
0.35
g/cc
Dep
th (m
)
Easting (m)
Cross-section of Density Contrast Model with Geology
-800-2200
0
200
400
600
800
Density contrast
slide 17© UBC-GIF 2003Magnetics
-100
-400
-700
-2200 -1000
Observed Data
Predicted Data60 22 -16
nT
Easting (m)
Nor
thin
g (m
)-1
00-4
00-7
00
-2200 -1000Easting (m)
Nor
thin
g (m
)
Magnetic data collected at the San Nicholas
deposit.
slide 18© UBC-GIF 2003Magnetic Inversion Results
0
5
10
Dep
th (m
)
O
-2200 -800
800
Easting (m)
Cross-section of Magnetic Susceptibility Model with
Geology
slide 19© UBC-GIF 2003
“Keel”
Tertiary BrecciaMafic Volcanics
Quartz Rhyolit
e
Massive Sulphide
Mafic Volcanics
Ele
vatio
n (m
) 2000
1600
-2000 -1100Easting (m)
Geologic cross section
Field Example: San Nicolas Deposit
Physical propertiesUnit Density Susceptibility Resistivity Chargeability
(g/cc) (S.I. x10) (ohmm) (msec)
Qal 2 0 50 5Tv 2.3 0 20-30 10Mst./Lst. 2.4 0 150 20Mafic Vol. 2.7 0 80 30Mafic/IntVol. 2.7 0 80 30Sulphide 3.5 10 20 200Qtz. Rhyolite 2.4 0 100 10Graphitic Mst. 2.4 0 100+ 30
Unit Density Susceptibility Resistivity Chargeability(g/cc) (S.I. x103) (ohm- m) (msec)
Qal 2 0 50 5Tv 2.3 0 20 -30 10Mst./Lst. 2.4 0 150 20Mafic Vol. 2.7 0 80 30Mafic/IntVol. 2.7 0 80 30Sulphide 3.5 10 20 200Qtz. Rhyolite 2.4 0 100 10Graphitic Mst. 2.4 0 100+ 30
- 10(20)
- 5- 5
- 10- 5
- 30- 40- 50- 50
- 20- 70
Density Magnetic Susceptibility• Electrical Conductivity• Chargeability
slide 20© UBC-GIF 2003Electrical Conductivity: Different Surveys
Sources:• Galvanic (grounded electrodes)• Inductive (current loops)
Waveforms• Sinusoidal (Frequency domain
• Time waveforms (Time domain)
Waveform
Time
I
Time
I
slide 21© UBC-GIF 2003
3D EM: Frequency Domain
Borehole Data
Depth
data
(E, H)
Source: Loop or grounded electrode
Waveform
Time
I
eiωσiω
JEHHE
)(0
Data (E,H)
slide 22© UBC-GIF 2003CSEM Survey
1.6 km
3.5 km
transmitter
receivers
• 15 Frequencies between 0.5Hz, 8192 Hz
• 3 lines, 1.6km long, 200m apart
• 25 meter station spacing
• single transmitter
• data are scalar impedances (Ex/Hy)
• data collected with the goal of MT interpretation
slide 23© UBC-GIF 2003
Iso-surface cutoff 10 Ohm-m
Frequencies0.5, 8, 64, 256 Hz
3D CSEM Inversion
slide 24© UBC-GIF 2003
(Loop or grounded electrode)Source
Waveform
Time
I (half sine, step…)
Surface Data
(E, H, dB/dt)
Borehole Data
Depth
data
(E, H, dB/dt)
3D TEM Setup
slide 25© UBC-GIF 2003
Introduction to UTEM data at San Nicolas
• 3 large loop transmitters– 2 km by 1.5 km
• dB/dt receivers– mainly z component
• transmitter waveform– 30 Hz sawtooth wave– dI/dt constant over
half cycle
I(t)
dIdt
dBdt
15 ms
slide 26© UBC-GIF 2003
Loop 1Loop 2
Loop 9
San Nicolas UTEM data
dBz /dt
nT/s
UTEM channel 4 (1.513ms)
easting
1075
-1300-3000 -220
nort
hing
slide 27© UBC-GIF 2003Fitting the data
View from SW
Observed 15 m iso-surface 1000.0
31.0
1.0
observedpredicteddBz /dt
nT/s
log10(t)
One decay curve: Observed and predicted
Observed
Predicted
slide 28© UBC-GIF 2003San Nicolas inversion results:
easting-2500 -500
1000
5
m
Recovered cross section at 450 S
easting-2500 -500
1000
5
m
Resistivity from drilling at 450 S
slide 29© UBC-GIF 2003Density, Magnetic susceptibility, Conductivity
3D CSEM
3D UTEM-950-2150 -1550 -950-2150 -1550
-950-2150 -1550
0
200
400
600
800
0
200
400
600
800
0
200
400
600
800
-800-2200
0
200
400
600
800
Density contrast
-800-2200
0
200
400
600
800
Magnetic susceptibility
slide 30© UBC-GIF 2003
“Keel”
Tertiary BrecciaMafic Volcanics
Quartz Rhyolit
e
Massive Sulphide
Mafic Volcanics
Ele
vatio
n (m
) 2000
1600
-2000 -1100Easting (m)
Geologic cross section
Field Example: San Nicolas Deposit
Physical propertiesUnit Density Susceptibility Resistivity Chargeability
(g/cc) (S.I. x10) (ohmm) (msec)
Qal 2 0 50 5Tv 2.3 0 20-30 10Mst./Lst. 2.4 0 150 20Mafic Vol. 2.7 0 80 30Mafic/IntVol. 2.7 0 80 30Sulphide 3.5 10 20 200Qtz. Rhyolite 2.4 0 100 10Graphitic Mst. 2.4 0 100+ 30
Unit Density Susceptibility Resistivity Chargeability(g/cc) (S.I. x103) (ohm- m) (msec)
Qal 2 0 50 5Tv 2.3 0 20 -30 10Mst./Lst. 2.4 0 150 20Mafic Vol. 2.7 0 80 30Mafic/IntVol. 2.7 0 80 30Sulphide 3.5 10 20 200Qtz. Rhyolite 2.4 0 100 10Graphitic Mst. 2.4 0 100+ 30
- 10(20)
- 5- 5
- 10- 5
- 30- 40- 50- 50
- 20- 70
Density Magnetic Susceptibility Electrical Conductivity• Chargeability
slide 31© UBC-GIF 2003Induced Polarization
Source Current (Amps)
Measured Voltage (Volts)
V Vs
(t)
VIP datum: Vs/Vn
Non-zero area occurs because charges took time to equilibrate.
Collect IP data along with DC resistivity data
slide 32© UBC-GIF 2003DC/IP data at San Nicolas
• Pole-dipole • Real Section
Pseudo-section
-800-2200
0
200
400
600
800
Chargeability
3D Inversion
slide 33© UBC-GIF 2003
Summary of physical property inversion at San Nicolás
Sulfide: dense, chargeable, susceptible, conductive.
-800-2200
0
200
400
600
800
Density contrast
-800-2200
0
200
400
600
800
Chargeability
-800-2200
0
200
400
600
800
Magnetic susceptibility
-950-2150 -1550
0
200
400
600
800
3D UTEM
slide 34© UBC-GIF 2003Back to Karst Aquifers: LSBB
• What scale? • Which physical properties?
•Soil•Epikarst•Unsaturated•Saturate
slide 35© UBC-GIF 2003Back to Karst Aquifers: LSBB
• Tunnel Scale (meters to km):– Conductivity (indicative of water)
(DC resistivity, FEM or TEM)
– IP might be useful for clay layers if they are chargeable
– Time-lapse DC (EM) resistivity can provide information about hydraulic conductivity
– Electrical permittivity (GPR)
slide 36© UBC-GIF 2003Geophysics for large scale aquifer
• Conductivity– Airborne EM for 200 meters (Epikarst delineation)– ZTEM or MT for deeper structure (large voids/conduits,
depth of saturated zone)
slide 37© UBC-GIF 2003Geophysics for large scale aquifer
• Density: (Change in water volume)• Magnetics ??• Self-potential (from fluid motion)• Chargeability ??• Self-potential, MRI, Seismic
Good news from geophysics side: We can invert most types of survey data to recover 3D distribution of physical properties.
slide 38© UBC-GIF 2003
Thank you!