Page 51 - Caldera Volcanism Analysis, Modelling and Response

26 Fidel Costa

Davies and Halliday (1998) found differences in crystallisation times between
and within minerals of o10 ky (within error of most individual glass-mineral
isochrons). Differences of 20–70 ky (up to ca. 300 ky) between feldspar centres
and rims were also found. They calculated crystal growth rates between 7 10 13
1
and 9 10 14 cm s .
More recent data of Simon and Reid (2005) show that ages of zircons rims
overlap with eruption, but centres are up to 150 ky older. The oldest zircon ages are
significantly younger than those previously obtained by whole-rock-glass-minerals
Rb–Sr isochrons (Figure 8), and thus reduce the residence times of the Glass
Mountain magmas to a maximum of 190 ky (Table 6 and Figure 8). The difference
between the Rb–Sr and U–Pb ages could reflect that the minerals defining the
Rb–Sr isochrons crystallised earlier than the zircons. However, recent interpreta-
tions (e.g., Simon and Reid, 2005) propose that the Rb–Sr model isochrons are
not dating the crystal fractionation. They probably are the result of a complex
87
history involving open system, including in situ Rb decay in the reservoir and the
presence xenocrysts.

3.4.2. Caldera magma: Bishop Tuff petrological attributes and time scale
information
The Bishop magma was thermally (ca. 720–7901C) and compositionally zoned in
87 86
major and trace elements and also in Sr/ Sr values (Hildreth, 1979; Christensen
and DePaolo, 1993; Hildreth and Wilson, 2007) prior to eruption. The first erupted
units are crystal poor, colder and more evolved than the later ones (Hildreth, 1979).
Water contents vary from ca. 6 to 4 wt% (Anderson et al., 1989), and a pressure of ca.
200 MPa was determined for the roof of the reservoir (Wallace et al., 1999). The
liquidus temperatures of the early and late magmas calculated with MELTS (Ghiorso
and Sack, 1995) are ca. 795 and 8551C, respectively. The origin of the zoning of the
Bishop magma has been highly debated and investigated. Hildreth (1979) proposed
that it resulted from convective circulation driven by thermogravitational diffusion,
complexation and wall-rock exchange. Other interpretations include crystal liquid
fractionation (Michael, 1983), side-wall crystallization (Wolff et al., 1990; Bindeman
and Valley, 2002), wall-rock assimilation (Duffield et al., 1995) or mafic intrusion at
the base of the reservoir (Hervig and Dunbar, 1992). Hildreth and Wilson (2007)
have recently proposed that the zoning was created by numerous batches of crystal-
poor melt released from a mush zone at the floor of the accumulating rhyolitic
magma body. Crystal-melt fractionation was the dominant zoning process. A major
role for a magma intrusion at the base of the reservoir for triggering the eruption of
the Bishop Tuff has been recently proposed by Wark et al. (2007).

3.4.2.1. Residence times of the Bishop magma. The study of residence times in
the Bishop magma also started with a long estimate from Christensen and DePaolo
(1993) who proposed a model of the Bishop magma being mainly liquid for
long periods of time (e.g., 0.5 My) based on Rb–Sr isotope systematics. A later
40 39
Ar/ Ar study by Van den Bogaard and Schirnick (1995) reported ages of glass
inclusions in quartz at 1.9–2.3 Ma, and thus residence times W1 My. A later