Earth's Magnetic Field
By convention, the "north-seeking" pole
corresponding to that at the north end of a compass needle is called the
positive pole, and the "south-seeking" pole is referred to as the negative pole
The lines of force are directed outward from a positive (i.e., north) pole and
inward to a negative (i.e., south) pole." P. V. Sharma
Geophysical Methods in Geology, 2nd Edition.
Note that this means the Earth's north magnetic pole is
a negative pole, because the positive "north-seeking" end of a compass
needle is attracted toward it. The lines of force referred to would be the force
on a positive "test monopole" of unit strength.
- Gilbert (1500s) recognized Earth's field was similar to
a dipole's
- 1838, Carl Friedrich Gauss proved 95% of Earth's
magnetic field is internal, approx. 5% external
Earth's field is sum of 3 parts:
- External Magnetic Field
- Anomalous Induced Magnetic Field
- Main Magnetic Field
1. External magnetic field
- about 1-5% of total field
- biggest secular variation is diurnal, with an amplitude
of tens of nT
- also a seasonal variation, i.e., diurnal variation greatest in summer
- strongest at equator
- all suggest role of sun
- EM (UV and x-rays) ionizes particles in
ionosphere
- Sun's tidal force produces cyclic wind currents
in ionosphere, which in turn produces electrical currents
- ionosphere directly produces 2/3 of observed
diurnal effect; induced currents in Earth contribute about 1/3
- magnetic storms: external field occasionally variable over minutes, at hundreds of nT
or more (magnetic storm)
- from 100s to 1000s of nT

- last hours to days
- caused by sunspot activity
-
WWV
reports useful to predict when not to conduct
surveys
- 27 day cycles (sunspots rotate with Sun)
- fixed-station recording magnetometers provide data
on field variability (e.g.,
Alaska)
[3/28/20001 propagation forecast
based on sunspot activity]- produced by electrical currents in ionosphere
consisting of particles ionized by solar radiation
- sunspot number strongly affects solar radiation

- current cycle, as of November 2004:

- sunspot cycles back to about 1750

-
Sunspots and Global Warming
2. Anomalous, induced magnetic field
- magnetization induced in crust by Main Field (and
External Field) or remanent magnetization ("permanently induced")
- limited to upper crust (Curie T)
- limited to ferr(o,i)magnetic materials
3. Main magnetic field
- produced by electrical currents in outer core
- steady on time scale of days, but variable over years
- approximately 50,000 nT (0.5 Oe)
- accounts for large regional variations in intensity and
direction
We showed that the potential due to a dipole is

and we obtain magnetic induction, B, or the vector field, by taking gradient of potential,
and find the components of this vector field:

The best-fit dipole approximation of Earth's field, at r =
a (radius of Earth):

- B0 is the equatorial strength of best-fit
dipole
- axis of dipole is currently tilted about 11 degrees to
spin axis
- this gives M (dipole moment) of 7.94*1022 Am2
- because this is a best fit dipole field centered
at Earth's center, the N and S poles are antipodal:
- North best-fit pole approximately at 79o N, 71o W
- South best-fit pole approximately at 79o S, 109o E
- Dip Poles: looking at Earth's actual field, we
(currently) have two points on Earth where field is vertical:
- North
dip pole (I = +90o), approximately at 76 deg N, 101 deg W, field strength 0.6 Oe
(60,000 nT)
- South dip pole (I = -90o), approximately at
66 deg S, 143 deg E, field strength 0.7 Oe (70,000 nT)
- magnetic equator: 0 deg inclination contour line; for
best-fit dipole field, it is a great circle "tilted" 11o from
geographic equator; "real" magnetic equator depends on current field values
but is shown schematically below

Spherical Harmonic Analysis of Earth's Field
- because magnetic field is conservative and, for a
monopole, follows same 1/r2 dependence as gravity, it, too,
obeys Laplace's equation
- thus the potential, V, must be of the form:


- here, q
and l are taken to be the
geographic colatitude and geographic longitude, respectively
- in contrast to gravity, for magnetic field, S' and C'
are not equal to zero (until you are outside Earth's ionosphere, where
external field is sourced)
- but they are small, average out over short times, and
are neglected when discussing "solid Earth" field, i.e., exclusive of external
field
- l = 0 term is zero because no magnetic monopoles
- it is convenient to introduce the Gauss coefficients:


Dipole and non-Dipole Components
- Earth's Main Field can be thought of as consisting of dipolar and
non-dipolar components
Dipole Field
- if Earth's field were purely dipolar, and
aligned with geographic north (i.e., with coordinate system for spherical
harmonic analysis), then g10
would be only non-zero term
- Q: Why not just use, say a Fourier series to represent
Earth's field? I.e., what is advantage of using spherical harmonic analysis,
Legendre polynomials, yada yada...
- if purely dipolar, but tilted, g10,
g11, h11
would be only non-zero terms

In other words the best-fit (tilted) dipole field, at r = a (radius of
Earth), as discussed earlier, can be analyzed in terms of spherical harmonics:
- B0 is strength of best-fit dipole at
magnetic equator
- as mentioned above axis of dipole is tilted about 11 degrees to spin
(geographic) axis, dipole moment is 8 x 1022 Am2, N pole approximately at 79o N, 71o W, S pole approximately at 79o S, 109o E
Non-Dipole Field
- Non-dipole field is just Earth's total (main) field
minus tilted dipole field
- shows four major anomalies (quadrupolar)
- can be constructed from spherical harmonic coefficients
by setting g10, g11, h11
= 0
Secular variation
- one of the few "solid Earth" phenomena that change
significantly over a human lifetime
- data go back to at least 1500s! Rregular
observations of magnetic field made since 1540 at London, later Paris
(figure from McElhenny)
- data go back to at least 1500s (London Observatory)
- change about 50 - 150 nT/year most places (see gs,
hs) mostly related to non-dipole field
- westward drift: non-dipole field drifts about
0.2 deg longitude per year
- produces decreases and increases at some localities
- indicates that core rotating more slowly than
mantle/crust
- dipole field changing also; decreasing 1/2000 per year
International Geomagnetic Reference Field (IGRF)
- IGRF is just the main field as defined by spherical harmonic coefficients
(external field excluded)
- it is a reference field in the sense that in exploration work, which is
concerned with the shallow crustal induced field, we subtract IGRF from
measured field values in a region, to remove a regional field
- to express the secular variation in the Earth's
main field, a new set of constants in the spherical harmonic
expansion of the field are made from the latest magnetic data, at 5
year intervals, and time-rate-of-change in field at end of the 5
year interval is computed for extrapolating into the future.
- for exploration field work, it is important to subtract the IGRF for the
time period the data were collected
9th
Generation IGRF
Geomagnetic Models and Software
My FORTRAN IGRF Program
- example: to find field in June, 2003 (2003.5),
extrapolate by using IGRF2000 values for 2000.0 plus secular terms for 2000.0
times 3.5; to find field in 1994, interpolate between 1990.0 and 1995.0
coefficients
- example:
Model: USGS90 Latitude : 35 N
Date : 10/6/93 Longitude: 97 W Elevation: 300.000 m
D I H X Y Z F
deg min deg min nT nT nT nT nT
--- --- --- ---- ----- ----- ---- ----- -----
6 3.0 64 14.1 22914 22787 2414 47474 52715
Model: USGS90 Latitude : 35 N
Date : 10/6/83 Longitude: 97 W Elevation: 300.000 m
D I H X Y Z F
deg min deg min nT nT nT nT nT
--- --- --- ---- ----- ----- ---- ----- -----
6 56.3 64 16.5 23199 23029 2802 48149 53447
Field Elements
- Earth's field can be separated into vector components:
- where F is the total field intensity, and X, Y, Z and H
are the north, east, vertical and horizontal components, respectively
- inclination: angle field makes with horizontal ("dip")

where qm is
magnetic colatitude; fm
is magnetic latitude
- declination: angle horizontal field makes with true
north
- X, Y and Z can be found from |F| (magnitude of total
field) and declination, D, and inclination, I
- and vice versa:

Magnetic field of Earth based on IGRF 1990 (Blakely):
- (a) Isodynamic map
showing total intensity, contour interval 2,500 nT
- (b) Isoclinic map
showing constant inclination, contour interval 10o
- (c) Isogonic map showing
constant declination, contour interval 10o
Present-day total magnetic field
- total field
- declination
- inclination
Non-dipole field
Secular variation in
vertical field
Location of Source of Magnetic Field
- we assume that, at the depth where the field is
created, all wavelengths would be of equal strength ("white" spectrum)
- note that terms in spherical harmonic expansion of
internal part of Earth's field drops of like:
- hence, higher degree terms drop off faster with
distance from source
- by examining spectral content (relative amplitude of
different wavelengths (orders) of Earth's field, one can estimate depth to
source
- define a quantity (analogous to power in seismology)
called mean square field per harmonic degree:
- gotten by adding up square of terms of all orders, m, for each
degree l.
- for lower order terms (e.g., 10 or less, as in IGRF), use IGRF
coefficients
- for very high order (short wavelength) terms (
),
use Fourier analysis of field strength from aeromagnetic data (great-circle airborne
magnetic survey at 3 km altitude)


- note that above illustration based on
Magsat data
-
Another illustration of spectral
roll-off
- strong roll-off of low-degree terms indicates a deep
source
- downward continuation of low degree terms (1-8)
indicates the spectrum would be white (no roll-off) at 0.47a
- therefore, long wavelength terms generated in outer
core
- further evidence for outer core source:
- dipole and non-dipole components change in direction
and magnitude with time
- motion is of order 0.2o per year (~20
km/yr)
- major non-dipole features move westward, hence called
westward drift
- this is ~105 times rate of mantle
convection
- thus implies origin in fluid core
- less rapid roll-off indicates a much shallower source,
out ~10 km of crust
Physics of Magnetism
Three sources of magnetism:
- Current loops

- Permanent magnets; remanent magnetization (primarily
magnetite and related minerals; hematite)
- TRM: thermal remanent magnetization
- CRM: chemical remanent magnetization
- DRM: detrital remanent magnetization
"detritus: material produced by disintegration and weathering of rock
that has been moved from its site of origin"
- Induced magnetization (same minerals)

It is empirically found that the intensity of
magnetization, I, is proportional to ambient field, H:

The total magnetic induction, B, is

Remanence (2.) and induced magnetization (3.) result from
these phenomena:
- paramagnetism: weak susceptibility possessed by
most materials
- diamagnetism: weak negative
susceptibility (e.g., halite, anhydrite, H2O)
- ferro- and ferrimagnetism: magnetite,
pyrrhotite, ilmenite, hematite, etc.)
Curie temperature
- above about 500-700 oC, minerals cease being
ferromagnetic (about what depth is this?)
- since para- and diamagnetism are weak, must attribute
low degree components of field to electric currents
- mantle is poor electrical (and thermal) conductor, so
currents are in Fe/Ni outer core
| Mineral |
Formula |
Curie Temperature |
| magnetite |
Fe3O4 |
578oC |
| maghemite |
gFe2O3 |
675oC |
| hematite |
aFe2O3 |
680oC |
Source of High Degree/Order Components of Magnetic Field

- example of application of geomagnetism to "exploration"
- remanent and induced magnetism resides in crust
- concentration of magnetic minerals in igneous and
metamorphic rocks, especially mafic rocks
- "basement" usually higher in magnetic minerals than
sedimentary rocks (sed rx are "transparent")
|
Typical Rock Susceptibilities |
| Sedimentary rocks |
0.00005 cgs emu |
| Metamorphic rocks |
0.0003 cgs emu |
| Granites and rhyolites |
0.0005 cgs emu |
| Gabbros and basalts |
0.006 cgs emu |
| Ultrabasic rocks |
0.012 cgs emu |
- since higher order terms tend to be sourced in the
basement rocks in crust, often depth to basement can be found
- involves measuring slopes or widths of first and and
second derivatives of anomalies (e.g., Peter's half-slope method), or spectral
methods

Core Magnetism: Self-Exciting Dynamo
Faraday Induction Law: time-varying magnetic field
induces currents in a conductor (e.g., car's alternator)

Biot-Savart Law: moving charges (current) create a
magnetic field

- must have had external field to kick this off (Sun,
especially in T-Tauri phase), but now self-sustaining
- magnetohydrodynamics: especially difficult because
flow of conducting core in presence of its own field gives rise to the
currents causing the field, but then we have:
Lenz's Law: the induced current will appear in
such a direction that it opposes the change that produces it
- this means the core must do work to maintain the
field
- the work turns into Joule heating in core (I2R)
==> P = IE, E = IR, ==> P = I2R
- where does work (energy) come from?
- estimated 109 - 1011 W
required; for comparison:
- earthquakes: 1012 W
- Earth's total heat flux: 4x1013 W
- precessional torque exerted on core by mantle
probably insignificant
- latent heat of fusion from solidifying core:
crystallization drives convection
- radioactivity drives thermal convection
- thermal convection must provide significantly more heat
than

in order to be efficient
- 40K is only radioactive isotope compatible
with core materials
- would require 0.1% K in core (equivalent to 67% of K in
a chondritic Earth)
Modeling the Self-Exciting Dynamo by Lee J. Harper
One of Maxwell's Equations is

Since

then

or
