Page 63 - Caldera Volcanism Analysis, Modelling and Response

38 Fidel Costa

intrusions in the model would produce longer equilibration times, but the
endothermic reactions associated with the wall-rock and magma or hydrothermal
convection might partly balance this effect (e.g., Norton and Knight, 1977; Bowers
et al., 1990). For example, detailed studies of contact areoles around exposed
plutons do not deviate significantly from the results of the simple approach used
here (e.g., Bowers et al., 1990; Furlong et al., 1991). The results of the modelling
can be seen first in Figure 9. The region above or to the left of the two curves are
those where magmas would have reached the solidus. The majority of eruptions
3
3
o10 km lie in this field, whereas most eruptions W100 km lie below or to the
right of the solidification curves. Thus, the long residence times calculated for small
magmas could be unrealistic because they would have long solidified before
eruption. Note that the difference is large, with a factor of ten or more of difference
between the magma solidification and residence times. However, it has been
pointed out (Mahood, 1990) that small eruptions might be only the ‘mobile’ parts
of a larger magmatic system that would take longer to solidify. In other words, the
intrusive to extrusive ratios are also an important factor. This has been tested by
calculating the time that a system reaches a given volume of still eruptible magma
3
(Figure 10). For erupted volumes between 1 and 10 km , residence times W ca.
2 3
100 ky imply that the magma at depth is 10 –10 times larger, which is much higher
than the highest reported values of 10 for intrusion to extrusion ratios (e.g., Wadge,
1982; Crisp, 1984; Pritchard and Simons, 2004; White et al., 2006). Thus, it seems
4 3
very unlikely that the residences W10 y of small eruptions (o10 km ) reflect times
the magmas stayed above their solidus. Nonetheless, it will be shown in the next
section that it is useful to consider the data in context of the caldera cycle (pre- and
post-collapse activity) of individual systems before reaching conclusions.
3
In contrast, almost all residence times for deposits W100 km fall to the right or
below the solidification curve (Figure 9). This indicates that their residence times
are not in contradiction with the time-temperature path of such a cooling model.
Moreover, the data for these large eruptions have intrusive to extrusive ratios
from 1:1 to 10:1 in accord with other estimates listed above (Figure 10). These
thermal arguments together with the in situ age variation in single crystals in the
Whakamaru ignimbrites and the Toba Tuff indicate that the residence times of
3
ca. 100 ky or more for magmas W100 km are probably a good indication of the
amount of time that they stayed above the solidus.
4.1.1.1. The caldera cycle, the time evolution of a single reservoir and crystal
ageing. The discussion above considers a scenario where the erupted magmas are
extracted from individual and isolated magma reservoirs. However, caldera collapse
is commonly pre- and post-dated by volcanic activity (e.g., Lipman, 2000a) and
there is the possibility that the long residence times of some small magmas reflect
that they were extracted from the same reservoir as the large caldera collapse
magma. This can be tested with a diagram where the residence times are plotted
against the eruption ages recalculated to a ‘caldera-collapse reference time frame’.
The eruption age of the pre- and post-caldera deposits are calculated with respect to
a ‘zero time’ of caldera collapse (Figure 11). If the pre- or post-caldera volcanics
were extracted from the same reservoir as the caldera-forming magma, their