Page 55 - Caldera Volcanism Analysis, Modelling and Response

30 Fidel Costa

3.4.4. Summary of residence times for the Long Valley system
The chronology of the studies of residence times for the Long Valley system
described above indicates that the residence times have been somehow getting
shorter with publication year. The initially long estimates of half a million years or
more for the pre-caldera and Bishop magmas obtained with the Rb–Sr system seem
now to be probably not correct. They probably reflect the effects of in situ isotopic
in growth of 87 Sr and also the presence of xenocrysts and/or assimilation (e.g.,
Simon and Reid, 2005). Likewise, the long times of millions of years suggested by
40
39
40
some Ar/ Ar data appear to be the result of excess Ar (e.g., Winick et al., 2001)
probably due to contamination from old rocks. The zircon age data indicates
residence times for some pre-caldera and Bishop magmas of o100 ky (e.g, Simon
and Reid, 2005), and the longer estimates that have also been found in some zircons
are probably reflecting recycling of crystals from previous intrusive episodes (see
discussion Section 4).

3.5. Valles-Toledo complex

This system is located in the Jemez Mountains (New Mexico) on the margin of
the Rio Grande rift. Here we are concerned only with the two major eruptions
(Table 7; Smith and Bailey, 1966; Self et al., 1986, 1996; Spell et al., 1990, 1996).
3
The lower Bandelier Tuff or Otowi Member (1.61 Ma, 400 km ) is related to the
Toledo caldera, and the upper Bandelier Tuff or the Tshigere Member (1.21 Ma,
3
250 km ) erupted from the present-day Valles caldera that coincides with the older
Toledo. Both tuffs are high-silica rhyolites rather homogenous in major elements
but chemically zoned in trace elements, with incompatible elements always being
more abundant in the early erupted rocks (e.g., Smith, 1979). Between the two
caldera collapses, a series of rhyolitic lava flows, domes and pyroclastic units were
erupted (Cerro Toledo Rhyolite; e.g., Stix et al., 1988; Table 7).

3.5.1. Time scale information: disruption of silicic carapace and mixing
The available Rb–Sr data for this system do not permit residence times to be
calculated with confidence (Table 7). Wolff and Ramos (2003) note that the Sr
87
isotope data are consistent with Sr in growth over 270 ky prior to eruption of the
Otowi magma. This approximately coincides with the time lag since the previous
ignimbrite eruption from the system (at 1.85 Ma). However, the clear evidence of
open system magmatic processes means that no confidence can be attached to this
residence time, and parallels the problems discussed for the Long Valley estimates
( J. Wolff, personnel communication). A maximum residence time of 380 ky for the
Tshigere magma can be obtained if we assume that the two members were erupted
from a common reservoir. Other time information has been obtained using the
elemental and isotopic disequilibrium found between different phases of the
Bandelier rocks and diffusion models. Wolff et al. (2002) found oxygen isotope
disequilibrium between feldspars and quartz of the Tshigere Member which could
have only survived for 1–200 years prior to eruption (Tables 2 and 7). This is the
time since disruption of a silicic carapace, assimilation and eruption. Such time