Geology Reference
In-Depth Information
having primarily frequencies of 24, 12, 8, and 6 h
and amplitudes of only a few tens nT. However,
we have seen that the external contribution to
the geomagnetic field can reach 1,000 nT during
magnetic storms. Finally, the third important con-
tribution to the total magnetic field is represented
by an “anomalous” field F D F ( r , t ) (intended
as perturbation of the main core field) associated
with the remnant and induced magnetizations of
crustal rocks. This field can be considered as a
time independent field when the main component
of magnetization is the remnant magnetization,
a condition which is generally met by oceanic
basalts (see Table 3.1 ) . Therefore, the total
magnetic field vector that is observed at the
Earth's surface can be written as follows:
Fig. 5.1 Relationship between main ( core )field F , ob-
served field T , and anomalous field F in the definition
of magnetic anomalies
observed data. Now we want to give a physical
significance to the expression ( 5.2 ). To this pur-
pose, we note that the field F in ( 5.1 ) can be
considered as a small perturbation to the main
reference field, caused by the magnetization of
crustal rocks. In fact, ignoring the external con-
tribution, the average magnitude of the observed
field is 45,000 nT, whereas crustal field mag-
nitudes in the oceans generally do not exceed
500 nT.
Following Blakely ( 1996 ), we also observe
that the total field anomaly T defined in ( 5.2 )
is not equivalent to the magnitude of the anoma-
lous field, F , because T Dj F C F jj F j
¤ F , as illustrated in Fig. 5.1 .However,for
F << j F j we can write:
T .r;t/ D F .r;t/ C S .r;t/ C F .r/ (5.1)
A total field magnetic anomaly is calculated
from scalar field measurements by subtracting
the reference core field, usually an IGRF, and
eventually applying a diurnal correction, which
removes those components of the measured field
associated with solar and ionospheric activity.
Let T D T ( r , t ) be the observed magnitude of
total field at location r and time t , which can be
obtained by a scalar magnetometer survey. Let
F D F ( r , t ) be the IGRF field at the same point
and time. Finally, let us assume that an estimate
of the external contribution to the magnitude of
the observed field, that is a diurnal correction
S D S ( r , t ), is available. Then the total field
anomaly is defined as:
T Dj F C F jj F j
Š p F F C 2F F j F j
D F r 1 C
2 F F
F F F Š
Š F 1 C F F
F F
T .r;t/ D T.r;t/ F.r;t/ S .r;t/
(5.2)
F D F F F
F F D F F
(5.3)
In the next section, we shall see that an esti-
mation of the external components in Eq. ( 5.2 )
can be performed using nearby magnetic obser-
vatory data and/or a special design of the survey
tracks. Unfortunately, most oceanic surveys are
performed far away from magnetic observatories,
and the ship-track design generally must satisfy
the requirements of other kinds of geophysi-
cal measurement. Therefore, the calculation of
marine magnetic anomalies is often performed by
simple subtraction of the reference field from the
Therefore, a total field anomaly T
approximately coincides with the projection
of the anomalous field F onto the reference
field axis. In other words, T approximates
the component of the field generated by the
crustal sources in the direction of the regional
field. Typical total field oceanic anomalies range
from a few nT to thousands of nT, with an rms
value of 200-300 nT. Therefore, the condition
j F j >> F is usually met. Note that in general
 
Search WWH ::




Custom Search