Mantle Mixing: Processes and Modeling - Physics and Chemistry of the Deep Earth - page 351

dense, warm regions could be the last remnants
of a magma ocean (e.g., Labrosse et al ., 2007),
or could be actively maintained by the recy-
cling of dense oceanic crust (e.g., Christensen &
Hofmann, 1994).
Recent studies on helium have further sug-
gested that significant layering is necessary to
maintain the primordial signature (Porcelli &
Elliott, 2008; Gonnermann & Mukhopadhyay,
2009). Most studies allow the recharge of He from
a deep reservoir (either the very deep mantle or
the core). This scenario is by the fact that the high
He reservoir is nonprimitive by any other mea-
sure (Hart et al ., 1992; Hanan & Graham, 1996),
which may indicate an efficient recharge of He to
the depleted Earth through a boundary layer (Stu-
art et al ., 2003). In recent years the suggestion that
noble gases cannot be recycled into the deep Earth
because of a ''subduction barrier'' has increas-
ingly been questioned and various observations
suggest that noble gases and halogens can be in-
corporated in seawater and subduct to large depth
(Holland & Ballentine, 2006; Sumino et al ., 2010;
Kendrick et al ., 2011). Even a small proportion of
fluids recycled to the deep Earth can explain the
heavy noble gas composition of the Earth (Sum-
ino et al ., 2010). Interestingly, this recycling is
not sufficient to explain He isotopic systemat-
ics and elemental abundances. This suggests that
He may be fundamentally decoupled from the
heavier noble gases and lithophile elements.
Increased instrument precision has led in
recent years also to a fundamental rethinking
about the bulk composition of the Earth's mantle.
For example, analysis of samarium-neodymium
isotopes in chondrites suggests that the Earth is
either not as strictly chondritic as is commonly
assumed (Caro & Bourdon, 2010), or that a very
early differentiation event caused the permanent
sequestration of enriched materials in the deep
Earth (Boyet & Carlson, 2005). This putative
reservoir

of mantle mixing. The mantle appears to be
heterogeneous at very small scale with large
variations (e.g., Warren et al ., 2009; Kooper et al .,
2009; Paulick et al ., 2010). Magmatic mixing
itself cannot necessarily homogenize such het-
erogeneity (Bryce & DePaolo, 2004). Fine scale
heterogeneity is also seen as scattering of seismic
waves, which suggests the presence of fine dis-
tributed differences in bulk chemistry (Margerin
& Nolet, 2003; Kaneshima & Helffrich, 2010),
which is likely linked to the presence of finely
distributed oceanic crust that has been recycled
and stirred, but not yet fully homogenized.
In summary, there is a rich and diverse set
of observations from geophysics and geochem-
istry. While the preservation of heterogeneity is
often most easily understood by invoking nearly-
fixed physical reservoirs that poorly mix between
each other, it is an interesting question whether
the heterogeneity may be better understood by
focusing on the processes that lead to the for-
mation, mixing and sampling of the heterogene-
ity. For example, the preservation of primitive
reservoirs indicated by Nd and He isotopes may
be caused by the incomplete processing of the
Earth's mantle and isolation of such heterogene-
ity in small, distributed regions rather than in
a permanently fixed and convectively isolated
reservoir. Similarly, the recycling of oceanic crust
imparts a continuing introduction of heterogene-
ity which explains at least in part the OIB and
MORB isotope systematics (Hofmann & White,
1982; Christensen & Hofmann, 1994; Branden-
burg et al ., 2008). In the next few sections I will
discuss the addition, stirring and sampling of het-
erogeneity with a view point of such ''dynamical
geochemistry.''

12.3 Adding Heterogeneity

has been linked to extremely high There is overwhelming evidence for the impor-

tance of recycled oceanic crust and sediments
in the MORB and OIB source. A fundamental
observation is the presence of eclogite or pyrox-
enite in the source of OIBs (Hofmann & White,
1982; Lassiter & Hauri, 1998; Ren et al ., 2009;

3 He / 4 He as well (Jackson et al ., 2010).
In addition to the large scale heterogeneity
suggested by the geochemistry and seismic
tomography we can use observations of fine
scale heterogeneity to constrain the efficiency