Geoscience Reference
In-Depth Information
the velocity and density jump can then be used
as a probe for local mantle temperature and
composition (e.g. Ritsema
et al
., 2009).
Mineralogical phase transitions are governed by
their Clapeyron slope, i.e.
water may also influence the sharpness of the
phase transitions and cause further complications
(Wood, 1995; Frost & Dolejs, 2007, Karato, 2011).
Alternatively, some seismic discontinuities in
the mantle could also be explained in terms of a
compositional boundary (e.g. Anderson, 1989) or
rheological boundary (e.g. Karato, 1992). If a dis-
continuity is caused by a phase transition, then
its characteristics, like topography and magni-
tude, are determined by the thermodynamical
properties of the phase transition. If a disconti-
nuity is governed by a change in composition,
then the discontinuity will move in the direc-
tion of the mass flux (e.g. Schubert
et al
., 2001).
So in a subducting slab, for example, the move-
ment of both the 410 and 660 km discontinuities
will be downward. Again, this can be tested by
studying seismic observations of the 410 and
660 km discontinuities in subduction zone areas.
Thus a link can be made with convection pro-
cesses, which will influence and leave their signs
in the structure of seismic discontinuities. The
660 km discontinuity is of particular importance
to geodynamical models as it forms the boundary
between the upper and lower mantle (Hofmann,
1997). Flow may either be impeded at this depth,
resulting in subducting slabs staying above this
discontinuity and leading to 'two-layer' mantle
convection (McKenzie & Richter, 1981). Alter-
natively, unobstructed flow may be allowed in
so-called 'one-layer' mantle convection (Davies &
Richards, 1992), or material may only occasion-
ally 'avalanche' into the lower mantle (Tackley
et al
., 1993). The nature of the topography of the
660 km discontinuity has been a key parameter
in this argument (Honda
et al
., 1993), because
in one-layer convection the 660 km discontinuity
is most likely due to a phase transition, while
in two-layer convection the 660 km discontinu-
ity might be a compositional boundary. Again, by
comparing seismic observations of the '660' with
predictions for mineral phase transitions, we can
test if the discontinuity is indeed caused by a
phase transition or not.
In this chapter, we seek to determine which
discontinuities have been observed consistently
in seismic studies and to what extent there
dT
dP
=
V
S
(10.1)
where
T
is temperature,
P
is pressure,
V
is volume change and
S
is entropy change
determining the slope of the phase boundary
(e.g. Bina & Helffrich, 1994). These slopes can be
determined in mineral physical experiments and
ab initio
calculations. The main mineralogical
mantle component is olivine (40-60%), which
changes from the
α
phase to the
β
phase (also
called wadsleyite) at pressure and temperature
similar to 410 km depth, with a positive Clapey-
ron slope (Akaogi
et al
., 1989; Katsura & Ito,
1989). Around 520 km depth, the
β
phase changes
to
γ
olivine (also called ringwoodite or spinel),
again with a positive Clapeyron slope (Rigden
et al.
, 1991). Around 660 km depth
γ
olivine
transforms to perovskite and magnesiow ustite
(also called ferropericlase), which has a negative
Clapeyron slope (Ito & Takahashi, 1989). The
opposite Clapeyron slopes predict that the '410'
moves up and the '660' moves down in cold
areas (resulting in a thick transition zone), while
the '410' moves down and the '660' moves up
in hot areas (a thin transition zone). Thus, we
can use observed depths of the 410 and 660 km
discontinuities in subduction zone and mantle
plume areas and compare with Clapeyron slope
predictions to test if the olivine phase transitions
can indeed explain the discontinuity observa-
tions. Many seismological studies just use olivine
phase transitions to interpret their discontinuity
observations. This hypothesis would be true
if the mantle consisted for 100% of olivine;
however, the mantle is only 40-60% olivine
(Ringwood, 1975). Additional minerals in the
mantle are garnet and pyroxene; phase transitions
in these minerals will complicate the story of
interpreting seismic discontinuity observations
as mineralogical phase transitions (Vacher
et al
.,
1998; Weidner & Wang, 2000). Volatiles such as
