Geology Reference
In-Depth Information
from the presence of small amounts of partial
melt (e.g., Lambert and Wyllie
1970
; Anderson
and Spetzler
1970
), while others proposed that its
existence can be explained by intrinsic properties
of peridotite close to its solidus. For example,
Karato and Jung (
1998
) observed that experi-
mental studies failed to prove significant effects
of partial melting on the physical properties of
peridotite for a range of melt fractions expected
over most upper mantle conditions (<1-3 %).
Conversely, they pointed out that in presence of
partial melting seismic wave velocities should
increase
. In fact, water dissolves in melts much
more than in crystal lattices, thereby in pres-
ence of partial melting it would be removed
from minerals such as olivine determining an in-
crease of mechanical strength. These authors sug-
gested that no significant partial melting occurs in
the asthenospheric LVZ, so that the decrease of
seismic velocities would result exclusively from
the high water content of this layer. Similarly,
Stixrude and Lithgow-Bertelloni (
2005
) built an
upper mantle elastic isotropic and homogeneous
model in conditions of thermodynamic equilib-
rium. These authors proved that an LVZ could
be explained by the model even excluding the
presence of melts. However, Hirschmann (
2010
)
has recently showed that melts should exist any-
way in the LVZ. Therefore, the open problem is
not understanding whether or not partial melting
may occur in the LVZ, but if this melting would
effectively influence or determine the observed
decrease of seismic velocities.
and frequency spectrum. Seismic stations usually
record three components of ground velocity or
acceleration: a vertical component,
Z
,andN-S
and E-W components. The horizontal recordings
are then rotated to have the
x
axis along the radial
direction to the earthquake epicenter. Therefore,
if ” is the azimuth to the source, then the radial
and transverse components of velocity are calcu-
lated by the following transformation:
P
u
R
P
u
T
cos
§
sin
§
sin § cos §
P
u
EW
P
u
NS
D
(9.74)
where §
D
3 /2
”. Figure
9.19
shows an exam-
ple of three-component seismogram, recorded at
a station in the range of distances known as
near
field
(<20
ı
). The three traces illustrate the
main features of a seismogram in the near field,
which includes
P
and
S
phases and surface wave
arrivals. The identification of seismic phases on
seismograms is a difficult task that can be per-
formed either manually by a skilled seismologist
or automatically by specialized computer algo-
rithms. In general, a seismic phase is identified
by a change of both amplitude and dominant fre-
quency. For seismograms in the near field, phase
identification is usually difficult in the sub-range
of angular distances associated with triplication
from the Moho discontinuity (see Fig.
9.13
), but
the correct
picking
of the arrival time of any
seismic phase can be hindered anyway by back-
ground noise or by the complexity of the field of
seismic velocities along the raypath. This is why
the interpretation of broad-band seismograms at
local or regional scale is often preceded by high-
pass filtering (with typical low cut-off frequency
of 1 Hz) to remove low-frequency noise, occa-
sionally by low-pass or band-pass filtering to cut
high-frequency noise.
Sometimes, filtering can be useful for
identifying the precise onset of some phases,
for example
S
waves. In this instance, it is
possible to apply a filter that simulates the
Wood-Anderson torsion seismometer (high-pass
2-poles Butterworth filtering with low-corner
frequency of 2 Hz, followed by integration from
velocity to displacement). The effect of this kind
of filtering on the seismogram in Fig.
9.19
is
9.7
Seismic Phases
Nomenclature
The variations of seismic velocity within the
Earth and the presence of discontinuities deter-
mine the formation of several classes of ray paths
after an earthquake. The corresponding arrivals at
recording devices are named
seismic phases
.In
general,
seismograms
result from the superposi-
tion of distinct seismic phases associated with a
unique event. The possibility to observe a seismic
phase at a particular station depends not only
from its amplitude, but also from its polarization
