Page 53 - Caldera Volcanism Analysis, Modelling and Response

28 Fidel Costa

40 39
Ar/ Ar study on the same material (Winick et al., 2001) found even older ages
at 4 Ma, but since the data were obtained using laser step heating, they were
interpreted as the result of excess 40 Ar presumably derived from degassing
of old rocks. Why the Ar isotope compositions of the glass inclusion did not
homogenize during residence times of several hundreds of ky is discussed in the
next section.
Christensen and Halliday (1996) and Davies and Halliday (1998) measured the Nd
86
and Sr isotopes of glass inclusions in quartz and assuming an initial 87 Sr/ Sr they
obtained ages of 1.4–2.5 Ma for the early and ca. 1 Ma for the intermediate and
late erupted units. This gave residence times of 250–450 ky (Table 6, Figure 8).
They were also able to determine age differences between feldspar cores and rims of
up to 250 ky.
Reid and Coath (2000) found a collective mean age (including rims and centre
of crystals from different units) for zircon at ca. 820 ka, which gives a residence
time of ca. 40 ky. This is about 10 times shorter than the time obtained from
feldspars and the Rb–Sr system (Figure 8). Later, Simon and Reid (2005) reported
mean ages for zircons that are 50–90 ky older than the eruption age, implying
also residence times for the Bishop magma of o100 ky. Simon and Reid (2005) cast
doubts on the validity of the Rb–Sr isochrons and propose that the Sr isotope
characteristics of the Bishop Tuff which were used to support inheritance from
the Glass Mountain can be explained by radiogenic in growth and crystal growth
from isotopically heterogeneous domains.
3.4.2.2. Diffusion studies applied to the time scales of processes of the Bishop
magma. Hervig and Dunbar (1992) calculated the time for complete
2
homogenisation of Sr zoning in sanidine (e.g., using the relation: x ¼ D t)to
be between 2 and 13 ky. Although it is a maximum time, it is about a factor of four
smaller than the shortest residence time obtained from zircon U–Pb data (40–50 ky;
Reid and Coath, 2000; Simon and Reid, 2005). Bindeman and Valley (2002)
calculated that between 0.1 and 10 ky occurred since the last replenishment plus a
18
fractionation event using CSD’s of quartz and zircons and the presence of d O
zoning in quartz. Anderson et al. (2000) used the trace element zoning patterns in
sanidine to calculate complete equilibration times of 140 ky–2.8 My, whereas
Morgan and Blake (2006) used the same crystals to obtain times of ca. 100 ky from
a more detailed diffusion model. This last estimate agrees with the longer residence
times obtained from zircon. The new diffusion model of Wark et al. (2007) based
on Ti in quartz indicates that less than 100 years occurred since the last magma
replenishment and eruption.
40 39
A remaining issue related to Bishop magma is why the Ar/ Ar data on glass
inclusion in quartz by Van den Bogaard and Schirnick (1995) and Winick et al.
(2001) did not homogenise within the residence times on the order of 100 ky’s
2
(this should be the case if a simple x ¼ D t relation is used for the calculations).
This can be evaluated using the Ar solubility (crystal/melt) and volume diffu-
sion data in quartz (Watson and Cherniak, 2003) and an analytical solution of
diffusion of glass inclusions in a spherical crystal (Qin et al., 1992). Using a
temperature of 7501C and a ratio of 0.1–0.05 for the inclusion to the crystal radius,