[ Pobierz całość w formacie PDF ]

Cape Verde
lithosphere (which we discuss in the
next section) and suggested this mate-
Cameroon Line
rial may occasionally become delami-
nated. Because it is cold, it would also
St Helena
sink to the deep mantle. As in the case
of the Hofmann and White model, it Mangaia
would be stored in a thermal boundary
layer, heated, and rise in the form of
mantle plumes. However, recent stud-
ies have shown that the Os isotope
composition of the subcontinental
Figure 11.30. Three dimension plot of 87Sr/86Sr, Nd/144Nd,
lithosphere is quite distinctive, and
and 206Pb/204Pb. Most oceanic basalt data plot within a tet-
quite different from that of mantle
rahedron defined by the composition of EMI, EMII, HIMU,
plumes, as we shall see in the next sec-
and DMM components. Oceanic islands and island chains
tion. This rather strongly suggests
tend to form elongate isotopic arrays, many of which seem to
that delaminated subcontinental
point toward a focal zone (FOZO) at the base of the tetrahe-
lithosphere does not contribute to man-
dron. Adapted from Hart et al. (1992).
tle plumes. Because mantle plumes
come in several geochemical varieties, it is possible that both mechanisms operate. Indeed, other as
yet unknown processes may be involved as well.
Most oceanic islands show some variability in their isotopic compositions, defining elongated ar-
rays on plots of isotope ratios. Such elongated arrays suggest mixing. This raises the rather obvious
question of what is mixing with what. In a few cases, the Comores are a good example, the elongate
arrays seems to reflect mixing between different plume reservoirs. The Comores data defines a trend
in isotopic space that appears to be the result of mixing between an EMI and a HIMU component. In
other cases, such as the Galapagos, the plume is clearly mixing with the depleted upper mantle.
However, in many cases, the cause of the isotopic variation is less clear.
Hart et al. (1992) plotted oceanic basalt isotope data in three dimensions, with axes of 87Sr/86Sr,
Nd/144Nd, and 206Pb/204Pb (Figure 11.30). Principal component analysis confirmed that 97.5% of the
variance in the oceanic basalt isotope data could be accounted for by these ratios (leaving 2.5% to be
accounted for by 207Pb/204Pb, 208Pb/204Pb, and 176Hf/177Hf). They found that most of the data plotted
within a tetrahedron defined by the hypothetical end members EM1, EM2, HIMU, and DMM. They
also noticed that many arrays were elongated toward the base of this tetrahedron on the DMM-
HIMU join. From this they concluded that in many, if not most cases, mantle plumes appear to mixing
with a previously unidentified component, which they named FOZO (an acronym for Focal Zone),
that has the approximate isotopic composition of 87Sr/86Sr = 0.7025,  = +9, and 206Pb/204Pb = 19.5.
They suggested that FOZO is the isotopic composition of the lower mantle and that plumes rising
from the core mantle boundary entrain and mix with this lower mantle material. It is unclear, how-
ever, whether such a composition for the lower mantle can be fitted to reasonable isotopic mass bal-
ances for the Earth. A rather similar idea was presented by Farley et al. (1992), who point out that
this additional component, which they called PHEM, seems to be associated with high He/4He.
White (1995) concurred with these ideas, but argued that the 87Sr/86Sr of FOZO is higher, and the 
503 November 25, 1997
W. M W hit e Geochemistry
Chapter 11: The Mantle and Core
lower, than estimated by Hart et al. (1992) and probably closer to the values chosen by Farley et al.
If Hart et al. (1992) are correct, this may explain why the isotopic composition of some volcanos,
notably those of Hawaii and the Society Islands, change over time. Hawaiian volcanos commonly go
through several evolutionary stages: a shield-building stage, during which the volcanic edifice is
rapidly constructed and a relatively uniform tholeiitic basalt is erupted, followed by a late stage,
during which eruption rates drop dramatically and compo-
sition shift to alkali basalt, and, following a significant
Shield-building Phase
hiatus in activity, a post-erosional stage in which small
volumes of basanite and nephelinite are erupted. These
latter magmas types are thought to be produced by small
degrees of melting. Through this sequence, isotope ratios
evolve, with Sr/86Sr decreasing and  increasing (e.g.,
Chen and Frey, 1985).
White and Duncan (1996) observed a similar pattern in
volcanos of the Society Islands. The argued that this evo-
lution in isotopic composition reflects the passage of the
volcano over a compositionally zoned mantle plume (Figure
11.31). The most incompatible element enriched material,
which is also the hotest, is located in the center, and
formed the core of the plume. This is surrounded by a Post-erosional Phase
sheath of material that is viscously entrained by the
plume as it rises. It is cooler and also not as enriched in in-
compatible elements. When the volcano is over the core of
the plume, the degree of melting is high, giving rise to
tholeiitic basalt with high Sr/86Sr and high eruption
rates during the shield building stage. When the volcano is
over the sheath, the extent of melting is smaller, producing
small volumes of nephelinite with low 87Sr/86Sr.
There is also a geographic pattern to both the distribu-
tion of mantle plumes and their isotopic compositions.
Mantle plumes appear to be preferentially located within
Figure 11.31. Cartoon of a model that
regions of slow lower mantle seismic velocities, as may be
explains why isotopic signatures of
seen in Figure 11.32. There are two areas where isotopic
magmas become more depleted as
compositions are particularly extreme (e.g., high Sr/86Sr),
volcanos evolve. During the shield-
one in the southeastern Indian Ocean and South Atlantic,
building stage the time of most vigor-
the other in the central South Pacific (Hart, 1984; Castillo,
ous growth, the volcano is located di-
1989). The anomaly in the Indian Ocean is called the
rectly over the plume, and magmas are
DUPAL anomaly, while that in the South Pacific is called
derived from the hot core of the
the SOPITA anomaly. Interestingly enough, both anoma-
plume, which has enriched isotopic
lies close to regions where lower mantle seismic velocities
signatures (high Sr/86Sr, low  ).
are particularly slow (Figure 11.32), which indicate low
Because of lithopsheric plate motion,
densities. The low density in turn implies that these are
the volcano will be located over the
regions of high temperatures in the lower mantle. While
edge of the plume during later stages,
the exact significance of this remains unclear, it does estab-
such as in the post-erosional stage.
lish a connection with oceanic island volcanism and lower
During this stage magmas are derived
mantle properties, strengthening the plume hypothesis,
from the viscously entrained sheath,
and favoring a lower mantle origin for plumes.
which has more depleted isotopic
There is still much to understand about the nature and [ Pobierz całość w formacie PDF ]


Powered by WordPress dla [Nie kocha się ojca ani matki ani żony ani dzieca, lecz kocha się przyjemne uczucia, które w nas wzbudzają]. Design by Free WordPress Themes.