Geology Reference
In-Depth Information
the continents were not, and had never been, sub-
ject to horizontal motion. Curiously, even some
famous seismologists, like Sir Harold Jeffreys,
were proud opponents of
mobilism
and the the-
ory of Wegener (
1912
) about continental drift
(Frankel
2012
). However, by the end of 1950s, a
large amount of evidence for continental drift had
accumulated, and most of this information came
from paleomagnetic data collected on continental
crust.
Paleomagnetic sampling on continents is usu-
ally structured on a hierarchical basis. At the top
level, we have
rock units
, which often coincide
with geologic formations or with one of their
members. Sampling is performed at several
sites
across a rock unit to give a unique estimation of
the NRM direction associated with the rock for-
mation. In general, at least ten sites are required
to achieve acceptable confidence limits in the
estimated paleomagnetic direction. A site can be
a lava flow, a dike, a pluton, or any other igneous
body unit. In sedimentary rocks, sites usually
coincide with specific layers in the stratigraphic
succession. For each site, paleomagnetists collect
sixtoeight
samples
using a portable drilling
device. The original orientation of the individual
specimens is accurately annotated for the subse-
quent reconstruction of the direction of magneti-
zation after laboratory treatment.
Once the specimens have been brought to a
paleomagnetic lab, they are processed to isolate
their primary remnant magnetization, a procedure
that is referred to as
cleaning
. Usually, NRM of
the order of 10
-5
A/m are sufficient to obtain
meaningful results. The modern measuring de-
vice for the determination of the NRM of a sam-
ple is the
cryogenic magnetometer
.Thisdevice
uses a complex sensor based on superconduc-
tivity, which is called SQUID (Superconducting
QUantum Interference Device). It can measure
NRM of rock specimens with total magnetic
moment
MV
10
10
Am
2
. The NRM vectors
measured on a set of
N
samples from
B
sites
generally include two components: a primary
NRM acquired during rock formation (TRM,
CRM, or DRM), and a secondary component that
is gradually acquired during the geological time
(the
viscous
remnant magnetization, or VRM)
secondary NRM must be considered as a form
of noise in the applications of paleomagnetism
to plate kinematics, thereby, a series of labora-
tory procedures have been designed to remove
this component of magnetization. To eliminate
a secondary NRM, a rock sample is subject to
partial demagnetization
, which operates on the
component of magnetization with lower stability.
This low-stability component usually coincides
with the secondary NRM that we want to remove,
whereas the high-stability component that is iso-
lated through this procedure often coincides with
the primary NRM. However, exceptions exist,
so that the high-stability component is usually
referred to as the
characteristic component
of
remnant magnetization (ChRM), which does not
necessarily coincide with the primary NRM.
There are two basic techniques for accom-
plishing partial demagnetization of a sample.
In the
alternate field
(
AF
)
demagnetization
,
a specimen is exposed for about 1 min to
an alternating magnetic field with linearly
decaying intensity from an initial amplitude
H
AF
100 mT. The typical frequency of this
signal is 400 Hz, and the device allows to rotate
automatically the sample in order to align in
turn all the specimen axes with the applied field.
At the time
t
when a positive peak
j
H
C
(
t
)
j
is
reached, only the domains with micro-coercivity
H
c
j
H
C
(
t
)
j
align their spins with the external
field. Because the amplitude of each half-cycle
is smaller than the previous one, after a half-
period
T
/2 a slightly smaller set of domains with
micro-coercivity
H
c
j
H
-
(
t
C
T
/2)
j
<
j
H
C
(
t
)
j
will align their spins in the opposite (“down”)
direction. Therefore, if the rate of decrease of
the AF is not large, only a small fraction of
spins will be frozen in the “up” direction, so
that the total magnetic moment of the grains
in these two intervals of micro-coercivity
H
c
will be approximately zero. Similarly, at the
next step, only a small fraction of spins will be
frozen in the “down” direction, thereby, after
several cycles the net magnetization along each
axis will be removed from the sample for all
