Page 71 - Caldera Volcanism Analysis, Modelling and Response

46 Fidel Costa

and thermal budgets of this type of igneous activity. The most important findings
are:

3
(1) The residence times of large volume caldera-forming eruptions (W100 km )
range from a few ky to up to a few 100 ky, and may be a robust indication for
the duration that magma spends above the solidus. This is indicated by in-situ
age data of crystals from the Whakamaru and Toba ignimbrites which
show progressive age changes from core to rim over a couple of 100 ky. This
time frame is consistent with numerical simulations of conductive cooling
models of igneous bodies, magma production rates, and intrusive to extrusive
ratios. The cooling rates of such large magmas are calculated to be between
5 10 4 and 3 10 3 Ky 1 which are also within the results of the thermal
models.
3
(2) Residence times of smaller (o10 km ) pre-and post-caldera magmas are
comparable or shorter than most caldera-forming magmas. Aside from
genuinely reflecting the times that these magmas were in the crust, other
interpretations come into play. The long residence times (e.g., 50–300 ky) of
some pre-caldera deposits can be explained if they are batches from the same
reservoir as the caldera-forming magma (e.g., Long Valley, Taupo). The long
residences of other small magmas reflect that the dated crystals are recycled from
a caldera-forming deposit of a previous caldera cycle (Yellowstone), from
plutonic rocks of the same caldera cycle with or without erupted equivalents on
the surface (Crater Lake, Taupo, Long Valley), or perhaps from a not completely
solidified caldera-forming magma reservoir (Taupo).
(3) Time scales of specific magmatic processes are in general shorter than overall
magma residence times. The geochemical imprint of assimilation and partial
melting of wall-rocks (e.g., Valles-Toledo) or of previously erupted intracaldera
lavas (e.g., Yellowstone) can occur rather quickly, in less than a thousand years.
Even shorter times are obtained for the time since the last magma replenish-
ment and eruption (ca. 100 years for the Bishop Tuff) and for xenocrysts
incorporation and eruption (oa decade).
(4) The majority of magma production rates from large caldera forming eruptions,
including the tuffs of Bishop, Fish Canyon, Toba, Huckleberry, Lava Creek,
3
3
Oruanui and Rotoiti, (ca. 15 10 km in total) fall within the range of
3 1
272 10 2 km y , values that are comparable to global eruptive fluxes
of basalts. When combined with the high intrusive to extrusive values expected
for silicic magmas of 10:1 they suggest that the mass fluxes related to caldera
forming volcanism are of the same order as that of some flood basalts (e.g.,
Grande Ronde basalts of Columbia River). Such high magma fluxes are
transient phenomena and insights should be gained from thermal models of
the crust that lead to episodic and silicic magma generation. Magma production
rates derived from pre- and post-caldera lavas are much lower and might reflect
the initiation or decline of a caldera cycle, although care has to be exercised as
other factors might bias the residence time estimates towards higher values (e.g.,
recycling of old crystals).