### ERTH2020_(Magnetotelluric) - Exploration Geophysics at the

```ERTH2020
Introduction to Geophysics
The Electromagnetic (EM) Method
Magnetotelluric (MT)
ERTH2020
0
Magnetotelluric
combination of magnetic and telluric* methods
(Latin ‘tellūs’ = ‘earth’  “Earth current”)
ERTH2020
1
Magnetotelluric
ERTH2020
…other scientists Tikhonov (1950) and Rikitake (1951), Kato & Kikuchi (1950).
2
Induction
Equivalent Circuits
I
I
•
I
DC Resistivity
•
Induced Polarisation
C
L
R
•
ERTH2020
Inductive EM
3
DC / IP
ERTH2020
4
Magnetotelluric (Passive EM)

ERTH2020

5
Goal
2
− iωμσ = 0
2

1.
1D diffusion equation
2.
→
=
2
≈ 500
ωμσ
Skin Depth (Penetration Depth)
ERTH2020
6
Overview
DC Resistivity
Induced Polarisation
Method
Direct Electrical Connection (galvanic)
Passive EM
Active EM
No direct electrical connection (inductive)
Induced primary
magnetic field via loop
Electrical
potential
Decay of electrical
potential
Ratio of E and H
fields
Secondary magnetic
field
Resistivity
Resistivity &
Chargeability
Conductivity
Conductivity
Measured
Injected DC current via electrodes
Induced primary
magnetic field via
natural EM fields
ERTH2020
(or its decay)
7
Contents
• Introduction
o
o
o
o
Maxwell Equations
Induction
Sources
Example
• EM theory
o
o
o
o
o
Divergence & Curl
Diffusion equation
1D Magnetotelluric
Skin Depth
Apparent Resistivity & Phase
• 2D MT Introduction
o Example
ERTH2020
8
Electromagnetic Induction
Ampere’s Law (1826)
×  = J + t
×  = −t
(magneto) quasi-static approximation , i.e. separation of electrical charges occur
sufficiently slowly that the system can be taken to be in equilibrium at all times
J electric current density (A/m2)
H magnetic field intensity (A/m)
ERTH2020
e.g.
B magnetic induction (Wb/m2 or T)
E magnetic field intensity (V/m)
http://farside.ph.utexas.edu/teaching/302l/lectures/node70.html
http://farside.ph.utexas.edu/teaching/302l/lectures/node85.html
9
Electromagnetic Induction
ERTH2020
Simpson F. and Bahr K, 2005, p.18
10
Electromagnetic Induction
Plane Wave
Source

45∘

×  = −t
Primary field
Ampere’s Law
×=J

Ohm’s Law

J = σ∙
ERTH2020
11
Magnetotelluric
Sources
ERTH2020
12
Magnetotelluric
Sources
Power spectrum: signal's power (energy per unit time) falling within given frequency bins
ERTH2020
Simpson F. and Bahr K, 2005, p.3
13
Magnetotelluric Applications
• Mineral exploration
• Hydrocarbon exploration (oil/gas)
• Deep crustal studies
• Geothermal studies
• Groundwater monitoring
• Earthquake monitoring
ERTH2020
Simpson F. and Bahr K, 2005, p.3
14
Magnetotelluric
Example 2D-MT resistivity model
• Locations of MT measurement
sites, Mount St Helens and nearby
• White and red dots show the locations of
the magnetotelluric measurements;
measurement sites shown in red were used
for 2D inversion.
• The east–west line (red) shows the profile
onto which these measurements were
projected. The coloured area shows the
region of high conductances. (=conductivity
X thickness)
• The green-to-orange transition corresponds
to a conductance of 3000 Siemens.
ERTH2020
Hill et al., 2009
15
Magnetotelluric
Example 2D-MT resistivity model after inversion
the conductivity anomalies are caused by the presence of partial melt
ERTH2020
Hill et al., 2009
16
EM Theory
divergence

∙
curl
×

∙

×

+   +
−
−
−
+    +
(vector)
ERTH2020
(scalar)
(vector)
17
Divergence (Interpretation)
The divergence measures how much a vector field ``spreads out'' or diverges
from a given point, here (0,0):
• Left:
divergence > 0 since the vector field is ‘spreading out’
• Centre: divergence = 0 everywhere since the vectors are not spreading out.
• Right: divergence < 0 since the vectors are coming closer together instead
 is the extent to which the vector field flow behaves like a source or a sink at a
given point. (If the divergence is nonzero at some point then there must be a
source or sink at that position)
ERTH2020
18
Curl (Interpretation)
The curl of a vector field measures the tendency for the vector field to “swirl
around”. (For example, let the vector field represents the velocity vectors of water in a lake. If the
vector field swirls around, then when we stick a paddle wheel into the water, it will tend to spin.)
• Left:
curl > 0 (right-hand-rule  thumb is up+)
• Centre: curl = 0 everywhere since the field has no ‘swirling’.
• Right: curl ≷ 0 since the vectors produce a torque on a test paddle wheel.
 describes the infinitesimal rotation of a vector field ( p.s. The name "curl" was first
suggested by James Clerk Maxwell in 1871)
ERTH2020
19
EM Theory
Time-Domain Maxwell Equations (magneto-quasi-static)

×=−

1

→ ×=−
μ

(Ampere)
×=
Note the use of the constitutive relations:
= μ
= ε
=
Also note that generally
1
→ ×=

μ = μ , .
=  , .
 first order, coupled PDEs
ERTH2020
20
EM Theory
Time-Domain Maxwell Equations (magneto-quasi-static)
 to uncouple, take the curl
1

× =−
μ

1

→× ×=−
×
μ

(Ampere)
1
× =

1
→× ×= ×

 Second order, uncoupled PDEs
ERTH2020
21
EM Theory
Time-Domain Maxwell Equations (magneto-quasi-static)
 Second order, uncoupled PDEs
1

→  ×  ×  = −
μ

Plane wave source  sinusoidal time variation
= 0
1

→  ×  ×  = −μ

= 0
where  = 2 the angular frequency and
= −1 the imaginary unit
•
Complex numbers arise e.g. from equations such as
2 + 1 = 0 →  2 = −1 →  = −1 ≡ .
•
Complex numbers can also
be written as  =
•
Generally complex numbers have a real and
imaginary part and are written as  =  +  where
is the real part and  the imaginary part.
•
=   = cos  +  sin
•
Compact way to describe
waves
ERTH2020
22
EM Theory
Time-Domain Maxwell Equations (magneto-quasi-static)
 Second order, uncoupled PDEs
Plane wave source  sinusoidal time variation
1

→  ×  ×  = −
= −
μ

= 0
1

→  ×  ×  = −μ
= −μ

= 0
where  = 2 the angular frequency and
= −1 the imaginary unit
•
Complex numbers arise e.g. from equations such as
2 + 1 = 0 →  2 = −1 →  = −1 ≡ .
•
Complex numbers can also
be written as  =
•
Generally complex numbers have a real and
imaginary part and are written as  =  +  where
is the real part and  the imaginary part.
•
=   = cos  +  sin
•
Compact way to describe
waves
ERTH2020
23
EM Theory
Frequency Domain Diffusion Equations
 Second order, uncoupled PDEs
1
→  ×  ×  +  = 0
μ
1
→  ×  ×  + μ = 0

 General equations for inductive EM
ERTH2020
24
EM Theory
1D solution 
= (piecewise) constant,  ≡ constant
with vector identity  ×  ×  =   ∙  −  ∙
=0
→   ∙  −  ∙   = −iωμσ
→  2  − iωμσ = 0
→   ∙  −  ∙   = −iωμσ
→  2  − iωμσ = 0
=0
 Diffusion Equations (Frequency Domain)
ERTH2020
25
EM Theory
1D solution 
= (piecewise) constant,  ≡ constant
∙=
∙=
Divergence of Ampere’s law
→  ∙  ×  = − ∙
→  ∙  ×  =  ∙  =  ∙ σ = 0
∙ σ = σ ∙  +  ∙ σ = 0
→∙=0

=−
∙ =0

(Gauss law for magnetism,
i.e. no magnetic monopoles)
→ σ ∙  = − ∙ σ
→∙=0
σ = 0
Proof  ∙  ×  = 0
via Cartesian coordinates
ERTH2020

∙  ×  =  ∙  ×  =

=    −    +    −
−

∙   −
−

+    −    = 0
26
EM Theory
1D solution 
= (piecewise) constant,  ≡ constant
→  2  − iωμσ = 0 ⇔
2
− iωμσ = 0
2
General solution for second-order PDE:
= 1  ωt− + 2  ωt+
decreases in
amplitude with z
<0
increases in
amplitude with z
 unphysical
>0
ERTH2020
Simpson F. and Bahr K, 2005, p.21
27
EM Theory
1D solution 
= (piecewise) constant,  ≡ constant
= 1  ωt−
Taking the second derivative with respect to z
2
= 2 1  ωt− = 2
2

→
=
ωμσ =  ωμσ = 1 +
↔
2
− iωμσ = 0
2
ωμσ 2 =
ωμσ/2 +  ωμσ/2
Real part
→
ERTH2020
1
=
=
ℜ
2
ωμσ
Imaginary part
Skin Depth (Penetration Depth)
Simpson F. and Bahr K, 2005, p.22
28
EM Theory
1D solution 
= (piecewise) constant,  ≡ constant
= 1  ωt−
Taking the second derivative with respect to z
2
= 2 1  ωt− = 2
2

→
=
ωμσ =  ωμσ = 1 +
↔
2
− iωμσ = 0
2
ωμσ 2 =
ωμσ/2 +  ωμσ/2
Real part
→
1
=
=
ℜ
2
ωμσ
Imaginary part
Skin Depth (Penetration Depth)
For angular frequency  for a half-space with conductivity
ERTH2020
Simpson F. and Bahr K, 2005, p.22
29
EM Theory
1D solution 
1
=
=
ℜ
→
=
ERTH2020
≡ constant → 0 = 4 ∙ 10−7
= (piecewise) constant,
2
ωμσ
2
=
4 ∙ 10−7 ωσ
Skin Depth (Penetration Depth)
107 2
=
4 ∙ 2
107
≈
40
107
≈ 500
4 2
Simpson F. and Bahr K, 2005, p.22 & http://userpage.fu-berlin.de/~mtag/MT-principles.html
30
EM Theory
1D solution 
= (piecewise) constant,  ≡ constant
Real part
=
Imaginary part
ωμσ/2 + ωμσ/2
and
1
=
=
ℜ
2/ωμσ
The inverse of q is the Schmucker-Weidelt Transfer Function
1

= = +
2
2
..has dimensions of length but is complex
The Transfer Function C establishes a linear relationship between
the physical properties that are measured in the field.
ERTH2020
Simpson F. and Bahr K, 2005, p.22
31
EM Theory
1D solution 
= (piecewise) constant,  ≡ constant
Schmucker-Weidelt Transfer Function
1

= = +
with  = 2/ωμσ
2
2
We had with the general solution earlier
= 1  ωt− →

= −

×
Therefore
ERTH2020

=
= −ωμ

−ωμ = − →  =
1
1
1
=
=−
ωμ
ωμ
Simpson F. and Bahr K, 2005, p.22
32
EM Theory
1D solution 
= (piecewise) constant,  ≡ constant
Schmucker-Weidelt Transfer Function
1
1
1
= =
=−
ωμ
ωμ
•  is calculated from measured  and fields  (or  and  ) .
• from  the apparent resistivity can be calculated:
with q =
ωμσ →
apparent resistivity
ERTH2020
2
2
1
= ωμσ → σ =
or ρ = 2 ωμ
ωμ

→ ρ =  2 ωμ
Simpson F. and Bahr K, 2005, p.22
33
EM Theory
Apparent Resistivity and Phase
apparent resistivity
phase
=  2 ωμ
=
tan−1
ℑ
ℜ
The phase is the lag between the E and H field and together with
apparent resistivity one of the most important parameters in MT
ERTH2020
Simpson F. and Bahr K, 2005, p.22
34
EM Theory
Apparent Resistivity and Phase
=
2 ωμ
−1
= tan
=
ℑ
ℜ
with

+
2
2
=
2/ωμσ
For a homogeneous half space:
ℑ  = ℜ  →  = tan−1 1 = 45°
•  < 45° → diagnostic of substrata in which resistivity increases with depth
•  > 45° → diagnostic of substrata in which resistivity decreases with depth
ERTH2020
Simpson F. and Bahr K, 2005, p.26
35
EM Theory
ERTH2020
Simpson F. and Bahr K, 2005, p.27
36
2D-MT Introduction
For this 2-D case, EM fields can be decoupled into two independent modes:
• E-fields parallel to strike with induced B-fields perpendicular to strike and in
the vertical plane (E-polarisation or TE mode).
• B-fields parallel to strike with induced E-fields perpendicular to strike and in
the vertical plane (B-polarisation or TM mode).
ERTH2020
Simpson F. and Bahr K, 2005, p.27
37
2D-MT Introduction
ERTH2020
Simpson F. and Bahr K, 2005, p.30
38
Numerical Modelling in 2D
2D solution 
= (, ),  ≡ constant
→ TE-mode (E-Polarisation)
∙  −  = 0
Numerical schemes, e.g.:
• Finite Differences
• Finite Elements
Escript Finite Element Solver
(Geocomp UQ)
ERTH2020
39
Numerical Modelling in 2D
σ = 10-14 S/m
σ = 0.1 S/m
σ = 0.01 S/m
Dirichlet boundary conditions via a single analytical 1D solution
applied Left and Right; Top & Bottom via interpolation
ERTH2020
40
Numerical Modelling in 2D
Electric Field (Real)
ERTH2020
Electric Field (Imaginary)
41
Numerical Modelling in 2D
Apparent Resistivity at selected station (all frequencies)
ERTH2020
42
Numerical Modelling in 2D
σ = 10-14 S/m
σ = 0.04 S/m
σ = 0.1 S/m
σ = 0.4 S/m
σ = 0.001 S/m
# Zones = 71389
σ = 0.2 S/m
ERTH2020
# Nodes = 36343
43
Numerical Modelling in 2D
Real Part
ERTH2020
Imaginary Part
44
Numerical Modelling in 2D
Apparent
Resistivity
f = 1 Hz
r = 25 Ωm
Skin-depth
≃ 500  ≃ 500 12 ≃ 1700
r = 10 Ωm
r = 2.5 Ωm
r = 1000 Ωm
r = 2 Ωm
ERTH2020
45
References
Simpson F. and Bahr K.: “Practical magnetotellurics”, 2005, Cambridge University Press
Cagniard, L. (1953) Basic theory of the magneto-telluric method of geophysical
prospecting, Geophysics, 18, 605–635
Hill G J., Caldwell T.G, Heise W., Chertkoff D.G., Bibby H.M., Burgess M.K., Cull J.P., Cas
R.A.F.: "Distribution of melt beneath Mount St Helens and Mount Adams inferred from
magnetotelluric data", Nature Geosci., 2009, V2, pp.785
ERTH2020
46
Unused slides
ERTH2020
47
EM Theory

×=−

→  ∙  ×  = − ∙
→∙ = 0

=−
∙ =0

(Gauss law for magnetism)
via Cartesian coordinates
−

∙  ×  =  ∙  ×  =  ∙   −   =
−

=    −    +    −    +    −    = 0
ERTH2020
48
EM Theory

×=+

→∙+∙
=∙+
∙ =0
t

→∙=−
∙

(Ampere)
however, the rate of change of the charge density ρ equals the
divergence of the current density J  Continuity equation

→∙=−
∙ =− ρ

ERTH2020
→∙=
(Gauss law)
49
2D-MT Introduction
−
×  =   −
−
−

=
= −
−

(Ampere)
−
×  =   −
−
−

=
−
→ TE-mode (E-Polarisation)
→ TM-mode (B-Polarisation)
= −
= σ
=
= −σ
σ =   −
ERTH2020

= σ

− =   −
Simpson F. and Bahr K, 2005, p.28
50
Numerical Modelling in 2D
2D solution 
= (, ),  ≡ constant
→ TE-mode (E-Polarisation)
= −
=
σ =   −
= −
=
+    =    −   =
∙  −  = 0
ERTH2020
Scalar PDE of
51
```