Page 41 - Caldera Volcanism Analysis, Modelling and Response
P. 41

16                                                              Fidel Costa


          older (e.g., Taupo units D and E). If we neglect crystals older than 0.5 My, most
          zircons have maximum model ages of up to 200 ka. Moreover, when all the data are
          taken together, there are two peaks, at about 35 and 100 ka. These peaks do not
          correspond to any eruption and thus if some of the old zircons are inherited from
          previous magmas these were never erupted. The exceptions are the 100 ka ages
          found in the deposits of three consecutive eruptions of Tihoi, Okaia, and Oruanui.
          This might reflect that they share a similar (common?) reservoir or petrogenetic
          history (see Section 4).
             The residence times were calculated using the oldest and the youngest zircon
          peak ages determined by Charlier et al. (2005) either by SIMS or TIMS data, and
          they vary between 13 and 100 ky (Table 3). The residence of the Oruanui magma
          can be up to ~70 ky, but can also be as short as 4–26 ky if the youngest peak ages as
          used (Figure 4). Charlier et al. (2005) argue that the zircon model ages at ca. 100 ka
          are from inherited crystals from previous but young intrusive episodes, and they
          propose residence times that are somewhat intermediate between the two extremes
          noted above. Moreover, they suggest that bimodal (or multi-modal) zircon model
          age spectra indicate remobilization of zircons from a crystal mush, rather than a
          long-lived magma body. Perhaps the complexities of the data partly reflect the
          current limitations in spatial resolution and precision of the analytical techniques.
          For example, the detailed in situ zircon age data for the Whakamaru group
          ignimbrites (Brown and Fletcher, 1999; their Figure 3) shows crystals with ages
          that progressively decrease from core to rim (e.g., with no jumps). If this is the case,
          one would expect no modality in the age distribution, and the origin of the
          bimodality found by Charlier et al. (2005) in the Whakamaru data is unclear (their
          Figure 22a).
             The cooling rate of the Oruanui magma obtained from the difference between
          the pre-eruptive temperature of 7601C (water content of 4.5 wt% and pressure of
          150 MPa; Wilson et al., 2006) and the liquidus (8301C, as calculated by MELTS;
          Ghiorso and Sack, 1995) over the range of residence times varies between 2   10  2
                       1
          and 10  3  Ky . This is almost two orders of magnitude higher than for the
          Whakamaru ignimbrites discussed above (Table 1). The calculated magma produc-
                                                                              3  1
          tion rates (Figure 4) range from o10  5  for small eruptions up to 0.1 km y
          for the Oruani, the highest found in all caldera related magmas (e.g., Table 1).



          3.2. Toba Caldera complex

          The Toba Caldera complex is located in Sumatra, Indonesia, in a subduction-zone
                                                     3
          setting. During the past 1.2 Ma, at least 3,400 km of magma have been erupted in
          four ash-flow units, three from the Toba caldera (Chesner and Rose, 1991;
          Chesner, et al., 1991; Chesner, 1998). The tuffs are compositionally zoned from
          minor andesite to predominant rhyolite with relatively high crystal content
          (12–40 wt%; Chesner, 1998). Toba is the largest resurgent quaternary caldera on
          Earth, measuring 100 km by 30 km, and was the source of the voluminous
                   3
          (2,800 km ) Youngest Toba Tuff at 73 ka (Table 4). Chesner (1998) reports pre-
          eruptive conditions of 701–7801C, 300 MPa, and magmatic water content of
   36   37   38   39   40   41   42   43   44   45   46