Geoscience Reference
In-Depth Information
Why was Lord Kelvin wrong?
In the second part of the nineteenth century, Lord Kelvin (William Thompson) fiercely opposed
the geologists on the burning issue of the age of the Earth. From observations on modern
sedimentation rates and intuitions based on the time they felt necessary to let the species
evolve to modern living creatures, sedimentologists and paleontologists guessed that the
Earth was older than a billion years. Lord Kelvin elaborated on the theory of conductive
cooling formulated earlier in the century by Joseph Fourier. Although he saw the opportunity,
Fourier was sceptical enough about the actual cooling processes not to work out the solution
for the Earth! Kelvin assumed that the Earth was initially molten at 3850
◦
Ckm
−1
and used
the expression of the conductive heat loss to infer the time it would take to match the modern
terrestrial heat flow (
40 mW m
−2
or a gradient of 30
◦
km
−1
). He concluded that the Earth
was no more than 94 Ma old, which he quickly reduced to a merciless 24 Ma.
Kelvin carefully assessed that he had ignored chemical reactions. It is often stated that, had
he known about radioactive heating by U, Th, and K, he would have found the right answer.
This is not correct. Kelvin's major mistake was actually to have ignored mantle convection,
by far the most efficient mode of heat transfer from the deep Earth to its surface. Mantle
convection had to await Arthur Holmes and another 30 years to be recognized as a major
geodynamic trait of our planet. Kelvin's Earth was rock-solid and the age he calculated was
the time it takes to cool a thermal skin (boundary layer) with a thickness approximately
equivalent to the continental crust.
≈
10
21
)
10
−
2
)
(1.4
×
×
(1.89
×
10
6
y
τ
=
1)
=
91
×
1.035
×
(3.6
×
10
16
)
×
(7.8
×
10
−
6
/
where allowance is made for the density of seawater (1.035 kg l
−
1
) with respect to river
water and of rivers. What we are really seeing here is the residence time of chlorine in
seawater. In addition to the unsupported assumption of an initially fresh-water ocean, the
theory described in
Chapter 5
states that over the course of geological time the ocean has
reached steady state for salt concentration. It therefore lost any memory of its origins and
no significant age can be deduced by such a computation.
As ages of more than three billion years started appearing at the hands of the earli-
est geochronologists, namely those who utilized lead isotopes and the U-He technique,
it became clear that Kelvin's calculations overlooked two fundamental processes, internal
heating by radioactive decay of U, Th, and K, and mantle convection. It was only in 1954,
however, that Claire Patterson produced the first meteorite isochron with chondrites and
iron meteorites on it (
Fig. 12.12
). The age of 4.55 Ga indicated by this isochron has been
challenged many times, but to no avail. Troilite (FeS) from one of these meteorites, Canyon
Diablo, contains no detectable uranium and therefore yields the initial isotopic composition
of lead in the Solar System. The
c
. 4.54 Ga age of the Earth has been repeatedly confirmed
by all known chronometers, which in itself is startling corroboration of the validity of the
principles of radiometric dating.
