Page 68 - Caldera Volcanism Analysis, Modelling and Response

Residence Times of Silicic Magmas Associated with Calderas 43

Figure 14 Calculated magma production rates vs. erupted volume for all available data.
Dashed lines are the global estimates of eruption rates for di¡erent rock types fromWhite
1
3
et al. (2006) and grey band is the estimate of this publication (272 10 2 km y ) for most
3
eruptions W500 km . Note that the magma production rates of large eruptions overlap with
magma eruption rates of basaltic compositions excluding £ood basalts. Data sources for this
¢gure are numerous and cited inTables 3--8 and in Section 3.7.
3
3
of ca. 15 10 km ). This is five times higher than the global eruption rate
3 1
estimates of ca. 471 10 3 km y for rhyolitic compositions (White et al.,
3 1
2006), but overlaps with the estimates of ca. 371 10 2 km y for basaltic
eruptions. The higher rates obtained in this manuscript show that residence times
are shorter than repose time between successive eruptions, and thus are better
estimates of magma production rates. However, they are still low values because not
all intruded magma is erupted. Using a nominally high value of intrusion to
extrusion rates of 10 (e.g., Wadge, 1982; Crisp, 1984; Pritchard and Simons, 2004;
White et al., 2006), more realistic magma production rates for rhyolitic magmas are
3 1
close to 0.1 km y . This is about an order of magnitude lower than the highest
3 1
eruption rate estimates of flood basalts (ca. 1 km y ; White et al., 2006), but
overlaps with some of them (e.g., Grande Ronde basalts of Columbia River), and
3 1
also with active oceanic hotspots like Hawaii in the last 100 ky (0.15 km y ;
Lipman, 1995). Note that for this comparison to be appropriate (i.e., mafic to silicic
magma production rates), intrusive to extrusive ratios of mafic magmas are
assumed much lower and closer to 1 (e.g., Wadge, 1982). The lifetimes of caldera-
related magmatic cycles span a few million years or less (except perhaps for
the caldera complexes of the San Juan Mountains; Lipman, 2000b) which
suggest that the high magma production rates related to calderas are transient
phenomena. How such high and transient magma production rates can be achieved