Page 185 - Fundamentals of Physical Volcanology

9780632054435_4_010.qxd 12/10/2007 12:32PM Page 162

162 CHAPTER 10

of the eruption is probably controlled primarily to the surface through feeder dike systems. There is
by the size of the underlying magma chamber, with currently disagreement, however, about the exact
larger chambers simply having more magma to eruption mechanisms. One view is that, once the
erupt. Why, though, are these eruptions always dike system reaches the surface, magma is able
ignimbrite-forming? Commonly, the eruptions start to erupt from it at very high rates (of the order
9
−1
with Plinian activity but evolve into ignimbrite- of 10 kg s ) in eruptions which last a matter of
forming events, the majority of the erupted volume days. This style of eruption can be seen as analo-
being generated during the ignimbrite-forming gous to the “elastic” eruptions described above, and
phase. We saw in section 6.7 that Plinian eruptions could be possible because the great width of a dike
can evolve into ignimbrite-forming eruptions if the extending completely through the crust would
gas content of the magma declines or if the mass minimize the frictional energy losses of magma
flux increases significantly during the eruption. rising through it. Eruption would cease when the
It seems likely that the key issue here is that the overpressure driving the eruption is relieved. The
magma is unusually gas-rich. This not only ensures large volumes of individual eruptions would then
that the initial phase of the eruption is Plinian, but reflect the large volumes of magma produced in
also guarantees that gas exsolution from the magma the plume head and stored at the base of the
continues beyond the critical point where the lithosphere.
initial excess pressure in the magma chamber is Another explanation is also possible, however.
relieved. This ensures that fractures begin to form It has been suggested that flood basalt eruptions
in the roof rocks and incipient caldera collapse may represent much longer duration eruptions in
starts. The fractures grow in a very inelastic way which lava is erupted much more slowly than pre-
and, depending on their exact locations and ori- viously thought. This type of activity can be likened
entations, can become much wider than the dikes to that seen during certain Hawaiian eruptions.
which fed the initial Plinian phase. The consequent Most eruptions at Kilauea volcano, for example, pro-
increase in mass eruption rate as the fractures widen, duce small volumes of lava in eruptions of short
coupled with the eventual decline in volatile con- duration (a few hours or days). Sometimes, though,
tent of the erupting magma as deeper levels in the an eruption occurs in which magma is erupted
chamber are tapped, leads to the change to ignim- continuously for years or decades at a fairly con-
brite formation and ensures that this happens at stant, slow rate. These eruptions differ from the small
a very high mass flux. The term “super-volcano” has cyclic eruptions described in section 4.4. In these

recently been coined to describe volcanoes with long-duration eruptions a dike system produces
large-volume magma reservoirs that can erupt in a continuous link from the mantle through the
this way, but there is nothing fundamentally dif- magma chamber to the surface, and magma can be
ferent between their eruption mechanisms and erupted steadily through it without periods of stor-
those of less devastating events. age and inactivity. During such eruptions the erup-
tion rate is determined by the rate at which magma
is transferred out of the mantle, and the erupted vol-
10.8.3 Flood basalt eruptions
ume is not limited by the usual constraints based on
Flood basalts differ from the eruptions we have dis- chamber size. The eruption can continue as long
cussed thus far in that they appear to erupt magma as the dike system remains open. It is possible that
directly from the base of the lithosphere without during flood basalt eruptions a similar circumstance
any significant crustal storage occurring. It is known prevails: once emplaced, a dike system is able to
that flood basalt eruptions are associated with supply magma steadily to the surface at a rate which
mantle plumes; they appear to be generated when is dictated by the rate at which magma is produced
a mantle plume first impinges on the lithosphere. in, or is able to segregate from, the mantle plume,
Current ideas suggest that large volumes of magma rather than simply the rate at which it can flow up
generated in the plume head are accumulated at the the dike. The emplacement of the flood basalt lava
base of the lithosphere and then erupted directly flow field is then seen as an extremely large-scale