Geoscience Reference
In-Depth Information
(detailed in Chapter 7) shows that, for a 10-Ma offset fracture zone, temperatures
anomalous by
100
◦
Cwould be observed within about 10 km of the transform
fault, whereas on the ridge axis close to the intersection with the transform
fault the temperature would be about 300
◦
C too low (Fig. 9.40). From such
a temperature difference, it is fairly simple to calculate the resulting contribution
to the change in depth of the median valley as the transform fault is approached.
These anomalous temperatures provide an explanation for thin crust in large-
offset transform faults, such as the Kane, Oceanographer and Charlie Gibbs
faults, but cannot explain thin crust in small-offset transform faults, the gradual
thinning of the crust adjacent to the transform fault and the deepening of the
median valley towards the nodal basin. These effects are better explained by
a magma 'plumbing' system, in which each ridge segment is fed by a single
subcrustal, centrally located magma-injection zone. In this way, the crust near
large- and small-offset transform faults would be thinner because these faults
are at the far ends of the magma-supply systems. Additionally, the regular 40-
80 km spacing between transform faults may indicate the horizontal distance over
which magma from a single central supply point can feed a slow-spreading ridge
segment. The more infrequent transform faults on fast-spreading ridges could in
this way be an effect of a more plentiful supply of magma.
±
9.5.3 Seismic activity at transform faults
The active part of a transform fault is that portion between the two ridge segments.
Figure 9.33 shows how closely the epicentres follow the axis of the mid-ocean
ridge and transform faults. These epicentres are determined by observations of
the earthquakes made by the WWSSN system, which, being on land, are nec-
essarily far from many of the mid-ocean-ridge epicentres. This means that only
fairly large-magnitude earthquakes (body-wave magnitude greater than about 5
in the case of the Mid-Atlantic Ridge) are detected and located and that location
is subject to some error (to determine the location, one has to assume a velocity
structure, which might not be particularly close to the exact structure at the epi-
centre). Nevertheless, it is clear that the earthquake activity is confined to a very
narrow zone centred on the median valley at the ridge axis and to the transform
valley on the transform faults. To detect smaller-magnitude earthquakes on the
mid-ocean ridges, it is necessary to deploy seismometers directly on the seabed
in an array close to and within the median and transform valleys. These instru-
ments, called ocean-bottom seismographs (OBS), must withstand pressures of
at least 5 km of water and record for many days; they are necessarily expen-
sive. Launching and recovery and the determination of their exact positions on
the seabed are difficult and important procedures. To use these instruments to
locate earthquakes, at least three OBS must be recording simultaneously because
four unknowns - origin time and epicentre location (
x
,
y
,
z
)-must be deter-
mined (see Section 4.2.1). The level of seismic activity along transform faults is
