Geoscience Reference
In-Depth Information
∼
350 km (Kawakatsu & Yoshioka, 2011). The
phase relation and the calculated thermal struc-
ture (Figures 13.1 and 13.2 of this study, and also
Figures 6 and 7 of Tonegawa
et al
., 2008) indi-
cate that this layer includes phase A, as well as
NAMs, to transport a few wt % H
2
O to the transi-
tion zone. It should be stressed that the presence
of phase A along the slab beneath Central Japan
is exceptional associated with the very cold envi-
ronment, whereas only NAMs are predicted to be
stable along the slabs in most of other subduction
zones.
The subducted HBL is subject to RT instabil-
ity, since hydrous materials are less dense than
the ambient dry mantle materials. Richard and
Iwamori (2010) have demonstrated that the HBL
along the horizontally lying slab (i.e., stagnant
slab) may produce numerous hydrous plumes
in the backarc region, and may explain the ex-
tensive Cenozoic magmatism ranging from the
backarc side of the Japan arc to mainland Eura-
sia (Whitford-Stark, 1987), beneath which a large
stagnant slab is observed (Fukao
et al
., 1992).
It is noted that these processes of instability
and plume ascent, even when ascending across
410 km discontinuity at which the water solu-
bility reduces significantly, accompany no fluid
phase until the plume reaches
plume originating from the HBL) has already been
limited at a shallow depth around the choke point
(Figure 13.2).
In addition, high electrical conductivity region
is observed under the Japan sea (Toh
et al
., 2006),
as well as low
V
P
-
V
S
regions beneath the backarc
region of the Japan arc (Nakajima & Hasegawa,
2007) and beneath the Japan sea and the mainland
Eurasia (Zhao
et al
., 2009), both being rooted
apparently from the subducted deep Pacific plate.
Based on the seismic structure along and within
the slab, including presence of olivine metastable
wedge, Kawakatsu and Yoshioka (2011) infer that
water is transported into the deep mantle mainly
along the top surface of the subducting slab, but
no significant amount by the slab itself. In this
case, slab will not be dehydrated. Neither the
HBL liberates a fluid phase because the maximum
H
2
O content is limited to be low at the choke
point. Therefore, the plume model triggered by
dehydratoin beneath the backarc region (Zhao &
Ohtani, 2009) may not work, whereas the RT
instability of the HBL itself is a feasible scenario.
Water transport from the surface to the mantle
transition zones can be summarized as schemat-
ically illustrated in Figure 13.5. At the wedge
corner shallower than
200 km depth (around the
choke point), extensive dehydration and hydra-
tion through porous and channel flows occur to
∼
50 km depth
since the H
2
O content of the HBL (hence the
∼
Magmatic-
Hydrothermal
Arc
Back-Arc
Intra-Plate
Porous &
Channel Flow
Fig. 13.5
Schematic
illustration showing water
subduction and migration
processes from the surface to
the transition zone when the
slab stagnates at 660 km
depth. Abbreviations are;
serp
Choke Point
Wet Plumes
RT-instability
=
serpentine,
chl
chlorite,
NAMs
=
nominally anhydrous
minerals, wad
=
hydrous wad/ring
=
wadsleyite,
=
ring
ringwoodite.
