Page 224 - Fundamentals of Physical Volcanology

9780632054435_4_013.qxd 12/10/2007 12:34PM Page 201

VOLCANISM ON OTHER PLANETS 201

rocks are essentially the same as those on Earth, and about 400 times less than the current rate at which
this leads to some interesting consequences. Some magma is supplied to the Hawai’I hot spot on Earth.
of these can be seen in Table 13.2, where the depth This low supply rate could not possibly maintain a
at which magmatic gases start to exsolve and the magma reservoir of the size inferred above from the
depth at which magma undergoes fragmentation caldera sizes against heat loss by conduction. Even
are both seen to be about three times greater on worse, this supply rate could not initiate the accu-
Mars than Earth. Other consequences are that the mulation of such a reservoir – each batch of new
depths at which magmas reach neutral buoyancy magma would freeze long before it was joined by a
are also about three times greater on Mars and so subsequent batch. For example, if the first magma
are the vertical extents of magma reservoirs. It is batch formed a sill 30 km in diameter (the full
hard to predict how the gravity should affect the width of the eventual reservoir), the elastic prop-
horizontal extent of a magma reservoir; this may erties of the host rocks would require the sill to
have more to do with the rate of supply of magma be about 30 m thick. This represents a volume of
3
from the mantle in the early stages of reservoir for- about 20 km . Thus at an average supply rate of
3 −1
mation. However, we do have observations to help: 0.03 m s the next magma batch would be
the widths of the calderas at the summits of martian expected to arrive 23,500 years later. But the time
volcanoes are typically ten times as large as those taken to cool a sill 30 m thick to the point where
on Earth, meaning that the reservoir widths are also it is all solid is only about 25 years! Worse yet, the
about 10 times greater, thereby making the typical martian shield volcanoes typically have several
reservoir volumes 300 times greater. The widths of summit calderas, and the way these overlap and
dikes are likely to be about twice as great on Mars intersect one another implies that each new one
but the excess pressures in magma chambers are formed from the collapse of the roof of a magma
likely to be similar to those on Earth, and this leads reservoir that was created in a new location after an
to typically greater magma flow speeds in dikes (by earlier reservoir had ceased to be supplied with
a factor of at least two), and hence the potential magma and had frozen. Thus the creation of new
for greater magma effusion rates (by a factor of up reservoirs in a given volcano has to occur not once
to about ten). This in turn implies greater lengths but many times. Also the time taken for an old reser-
of lava flows (also by a factor of about ten, which voir of the size seen on Mars to cool to the point
is consistent with what is observed) and, coupled where the density and stress distributions force any
with the much larger volumes of magma stored in new reservoir to form at a different location is two

reservoirs, gives the potential for extremely large to three million years.
compound lava flow fields, which are indeed what The only simple way to reconcile these observa-
are seen to dominate most of the shield volcanoes. tions and calculations is to assume that the magma
A second indicator of long-lived high-effusion-rate supply to any one volcano from the underlying man-
eruptions exists in the form of sinuous rilles of the tle hot spot is episodic rather than continuous.
kind first identified on the Moon. These are not To make all of the steps in the cycle work correctly
found everywhere, but are particularly common we have to assume that the volcanoes each have
on the volcano Elysium Mons. active periods lasting about 1 million years alter-
A final aspect of the sizes and ages of martian vol- nating with dormant periods of about 100 million
canoes deserves mention. The total volume of any years. There is an intriguing consequence. The most
one volcano divided by the length of time for which recent crater counts made on the highest resolu-
it was active gives the average rate at which magma tion images available from current spacecraft in
has been supplied to it from the mantle. Typical orbit around Mars show that some lava flows may
3
6
figures for the Tharsis shields are 10 km (from be at most a few million to a few tens of million
topographic data) and about 2 billion years (from years old. In the past this has been taken to mean
impact crater counts, and subject to an error that that the martian volcanoes are dead, and that
could be as large as a factor of two). This implies humans have started exploring Mars too late to see
3 −1
an average supply rate of 0.015 m s , which is any volcanic activity. But given the time scale of the