Geoscience Reference
In-Depth Information
the
M
w
7.9 earthquake that occurred 626 km beneath Spain in 1954 and is
associated with a seismic velocity anomaly may be the remnant of a slab detached
following Africa-Eurasia subduction. Likewise the zone of seismicity at 600-
670 km beneath the North Fiji Basin, which is separate from the seismicity of the
Tonga and Vanuatu subduction zones, may be due to transformational faulting
within a slab that was subducted at the fossil Vitayz trench, but then subsequently
detached and foundered at the base of the upper mantle.
Deep earthquakes do not have many aftershocks compared with shallow earth-
quakes. Temperature is an important control on aftershocks following deep earth-
quakes but not on aftershocks following shallow earthquakes. Subduction zones
with high thermal parameters (cold) have far more deep aftershocks than do sub-
duction zones with low thermal parameters (hot). The
b
-value (Section 4.2.5) for
deep aftershocks is also related to the thermal parameter: high-thermal-parameter
slabs have high
b
-values and low-thermal-parameter slabs have low
b
-values. This
is consistent with a transformational-faulting origin for deep earthquakes.
Two large deep earthquakes that occured in 1994 promoted much new research
on deep earthquakes. The
M
w
=
7.6 Tonga event at 564 km depth was unprece-
dented in that 82 aftershocks with magnitudes between 3.6 and 6.0 were recorded
during the following six weeks. The main event and many of the aftershocks
occurred on a near-vertical plane consistent with one of the nodal planes from
the main event. The rupture zone was 50 km
=
65 km in extent and extended
beyond the expected metastable olivine wedge and out of the known seismic
zone (Fig. 9.56). It has been proposed that ductile faulting, triggered by tranfor-
mational faulting in the cool slab immediately outside the cold metastable wedge,
caused the two outlying aftershocks. The
M
w
×
8.3 earthquake on the Nazca
subduction zone beneath Bolivia was rather different in that it occurred in a region
with no previous recorded seismicity and had only three aftershocks with
m
b
≥
4.5. The geometry of the Nazca slab is therefore not known, but, depending upon
its shape and extent, this earthquake with a 30-50-km sub-horizontal fault plane
may have ruptured beyond any metastable wedge. Alternatively, this earthquake
might not have been the result of transformational faulting but may instead have
resulted from ductile faulting or plastic instabilities in a warmer spinel slab.
This might account for the very high stress drop (
=
110 MPa), which was over an
order of magnitude greater than that for the Tonga event and for normal shallower
events (Fig. 4.12). Thus present knowledge of the kinetics of mantle reactions, the
behaviour of minerals at very high pressures and the details of the assumptions
made in thermal calculations may need to be re-examined in order to establish
the cause of all deep earthquakes - transformational faulting might not be the
only process taking place. Deep earthquakes remain something of a puzzle.
∼
9.6.4 Gravity across subduction zones
Ve ry large gravity anomalies occur over subduction zones. The anomalies across
the Aleutian Trench, the Japan Arc and the Andes are shown in Figs. 9.57-9.59.
