Geoscience Reference
In-Depth Information
Heat flow shows little correlation to crustal
age (Figure 26.2), ocean depth at the sampling
site, or the inferred thickness of the plate, it is
nearly constant through the Atlantic and Indian
ocean crust and Pacific ocean crust older than
about 40 million years. Young Pacific crust has
a slightly elevated heat flow but much less than
predicted from plate tectonic models of cooling
with constant thermal conductivity. Heat flow is
not significantly elevated over hotspots, swells or
superswells, compared with the flux for normal
oceanic lithosphere, implying that (1) the under-
lying mantle is not hotter than average, (2) the
associated volcanism does not significantly warm
the lithosphere or (3) underplating, intrusion
and heating are common and that conditions for
extrusion involve stress in the lithosphere rather
than excess temperatures or unusual mantle at
places that are called hotspots. Even midocean
ridges are not always distinguished from 160-Ma
lithosphere on the basis of heat flow. A significant
problem with interpreting heat-flow data is that
hydrothermal circulation suppresses the conduc-
tive gradient. Thus, the conduction gradient can
be affected by fluid circulation from above or by
magma injection from below. These processes are
controlled by physical and environmental effects
that are not simply functions of age.
The conduction gradient is steep compared to
melting curves and the lower part of the TBL can
be above the solidus. Melts can be extracted from
various depths in the TBL and can therefore yield
different potential temperatures. Rapidly ascend-
ing melts rise adiabatically and one speaks of the
potential temperature
inferred from these melts.
This may not be the same as the potential temper-
ature of the deeper mantle. The mantle beneath
the plate is referred to as 'the convecting man-
tle' with implications about an adiabatic gradi-
ent and chemical homogeneity. At high tempera-
tures, the base of the TBL may be weak and fall
off, or delaminate, if it is denser than the under-
lying mantle. It may also deform or flow laterally,
if the necessary forces exist, or experience small-
scale convection, if the viscosity is low enough.
Departures of bathymetry and heat flow from
the theoretical square-root of age relations, are
often attributed to thermal perturbations due to
deep mantle plumes. These have been used to
estimate the buoyancy flux of plumes and as an
estimate of core heat. However, the mantle is not
homogenous or isothermal, plates are not uni-
form or impermeable, and thermal properties are
not independent of temperature. The upper man-
tle is close to or above the solidus and the melt-
ing point of the mantle is variable. Regions that
are shallow for their age do not imply a deep
source of heat; there is no reason to believe that
the mantle has a single potential temperature or
a uniform composition or melting temperature.
Mechanisms for seafloor flattening
Ocean depth and heat flow are usually inter-
preted in terms of boundary layer and plate the-
ories. In both, the initial condition is an isother-
mal (or adiabatic) homogenous bath of fluid of
zero viscosity. The surface is suddenly dropped
to a low temperature. A cold boundary layer
grows with time. In the plate model, the thermal
boundary layer is taken to be constant thickness
and the bottom is at the same temperature as the
vertical ridge axis. Heat is conducted through the
plate and its average temperature increases with
time. After a long time, the heat flow and the
temperatures in the plate reach equilibrium val-
ues. The constant thickness of the TBL and the
constant temperature at its base are not natural
boundary conditions, even if the plate is of dif-
ferent material than the underlying mantle. A
variant of the plate model is to change the lower
boundary condition to one of constant heat flux.
Explanations for seafloor flattening at old age
range from the mundane to the exotic; some
explanations affect estimates of the global heat
budget. Flattening sets in at different times for
different sections of the seafloor and sometimes
does not occur at all. The most obvious expla-
nation is to accept that the theoretical mod-
els are highly artificial and they should not be
expected to correspond to reality. The mantle
is not isothermal or adiabatic, in the absence
of plates, and the plate is subjected to pro-
cesses other than monotonic conductive cool-
ing. If the plate has a constant heat flow lower
boundary condition, rather than a constant tem-
perature one, then one automatically obtains a
background heat flux, and a constant heat flux
in old ocean basins. The inferred thickness of
the outer conduction layer is often much greater
than can form by cooling in 200 Myr, consistent
