Geology Reference
In-Depth Information
1700.
1700.
400.
400.
400.
400.
(a)
(b)
Parameterised
Heat loss
Heat
generation
Mean mantle
Upper mantle
Plate velocity (cm/yr)
1200.
1200.
0.
0.
0.
0.
0
0
4500
4500
0
0
0
0
4500
4500
4500
4500
Time (Myr)
Time (Myr)
Figure 9.2. Summary of the thermal evolution of a numerical model of mantle con-
vection with decaying heat sources. (a) Upper-mantle temperature (solid black),
mean mantle temperature (solid grey) and the temperature from the parametrised
calculation in Figure 9.1 (dashed grey). (b) Surface heat loss (thick solid) and
radioactive heat generation (thick dashed). The plate velocity (thin solid) is
included.
was hotter and thermal boundary layers thinner, and to run the models for the age
of the Earth, at least for two-dimensional calculations. Results from a model in
which radioactive heat generation decays appropriately are shown in Figures 9.2
and 9.3. Details and parameters of this model are given in Section B.3.
Figure 9.2(a) includes the temperature curve from the parametrised calculation
in Figure 9.1. They are quite similar once the effect of the different starting tem-
peratures passes, and apart from a small vertical offset that reflects the difficulty
of predicting exactly the temperature at which the numerical model will run. This
reflects the similar radioactive decay and similar temperature dependence of vis-
cosity used in the two models. This similarity gives some confidence in both kinds
of calculation.
As before, the surface heat loss (Figure 9.2(b)) tracks the heat generation
closely - more closely in this case because there is no core heat included in
this model, so radioactivity provides the only heat input.
Figure 9.2(b) includes the velocity of the surface plates in this model. Their
present velocities are 6.2 cm/yr but they were much higher in the past. In the
Archaean (0.7-2 Gyr in the plots) they were 18-60 cm/yr. The thermal boundary
layer and the subducting plates are correspondingly thinner at the earlier times
(Figure 9.3). Subduction is not modelled with great accuracy in this model, but the
impression of slower and thicker plates and subducted lithosphere is well conveyed
in Figure 9.3. The slowing plates and the corresponding slowing of convection
velocities within the mantle are also evident in this figure because the spacing of
the streamlines is inversely proportional to the local velocity.
Returning to Figure 9.2, mean temperatures for both the upper mantle and the
whole mantle are included in panel (a). In this model the lower mantle is 30 times
more viscous than the upper mantle, as indicated by observations (Chapter 4). There
