Geology Reference
In-Depth Information
T
C
4
1573
3
1428
923
2
598
1
0
273
(b)
(a)
Figure 10.20. Grey scale versions of cut-away images of (a) temperature and (b)
tracer concentration in a three-dimensional spherical model of mantle stirring with
heavy tracers. After Huang and Davies [216]. Copyright American Geophysical
Union.
100
3
(a)
(b)
80
2
60
40
1
20
0
0
0
1
2
3
4
0
2
4
6
Earth time [Gyr]
Mantle masses processed (
t
/
τ
)
Figure 10.21. (a) Residence times from MORB and 'OIB' samples of three-
dimensional models (solid curves). Dashed curves are estimates based on the
sampling theory (Sections 10.7.3 and 10.7.4). 'OIB' 50 is for an excess density of
50 kg/m
3
, and 'OIB' 100 is for 100 kg/m
3
. (The 'OIB' interpretation is discussed
in Section 10.7.4 below.) (b) Percentage of mantle processed through melting
zones versus time, normalised by the processing time, from a 3D numerical model
(dotted) and the sampling theory (dashed; Section 10.4.3). After Huang and Davies
[215, 216]. Copyright American Geophysical Union.
the equivalent of 18 Gyr of processing at present rates. The amount of man-
tle processed, the complement of the amount of primitive mantle remaining
(Eq. (10.5)), must also be run for several processing times to approach a real-
istic value (Figure 10.21(b)). Because convection runs nearly 10 times faster early
in Earth history (Figure 10.10), models that are run at steady rates for only 4.5 Gyr
or less will have accounted for less than the first billion years of evolution, and will
seriously underestimate the degree of processing and the residence times.
