Page 43 - Caldera Volcanism Analysis, Modelling and Response
P. 43

18                                                              Fidel Costa


          5–6 wt%. The liquidus calculated with MELTS (Ghiorso and Sack, 1995) vary
          between 840 and 9701C.



          3.2.1. Unfolding the building stages of large silicic magma bodies
          The Youngest Toba Tuff was investigated for its residence time using U-series
          disequilibria measured in allanite by SIMS (Vazquez and Reid, 2004). The crystals
          are zoned in age and composition, from cores with ages at 100–236 ka and rims
          with eruption ages. This gives magma residence times of up to 160 ky (Figure 1 of
          Vazquez and Reid, 2004). Combining the allanite’s age and compositional zoning
          Vazquez and Reid (2004) calculated magma fractionation rates of 0.1–0.4% per ky.
          Moreover, they proposed a magmatic history consisting of a quiescent early period
          followed by multiple episodes of mixing of melts differentiated to different degrees.
          The development of the chemical zoning of the deposits seems to have occurred
                                                                     1
          close to eruption. Calculated cooling rates are 0.8–1.1   10  3  Ky , and magma
                                      3  1
          production rates are ca. 0.02 km y  (Table 1).
                                                                              39
             The eruption age of the Youngest Toba Tuff has been determined by  40 Ar/ Ar
          and K–Ar using sanidine and biotite with consistent results between laboratories
                                    39
                               40
          (Table 4). However, the Ar/ Ar data of Gardner et al. (2002) on hornblende and
          plagioclase indicates ages of up to 1.5 Ma. This result, together with experimental
          work, was interpreted evidence for the hornblende and part of the plagioclase being
          xenocrystic. Gardner et al. (2002) calculated that about 25 years is the maximum
          time that plagioclase crystals of 1.5 Ma would be able to retain any old age, and thus
          this is the maximum time since xenocrysts incorporation and eruption. An
          alternative interpretation for the age data is that sanidine and biotite were as old and
          from the same source as the hornblende and plagioclase xenocrysts (perhaps of an
          earlier intrusion?) but their ages were completely reset to that of eruption because
          of their lower closure temperatures (Figure 2).



          3.3. Yellowstone calderas

          The Yellowstone volcanic field encompasses part of the USA states of Wyoming,
          Idaho and Montana and its caldera-related activity about 2 My ago. Volcanism could
          be related to a hotspot but not necessarily a plume (e.g., Christiansen et al., 2002).
          The geologic history defines three caldera cycles, each culminating with the
          eruption of a voluminous rhyolitic ash-flow (e.g., Christiansen, 2001). The three
          major caldera-forming eruptions (Table 5; Figure 5) have been dated at 2.0 Ma
                                    3
          (Huckleberry Tuff; 2,500 km ), 1.3 Ma (Mesa Falls Tuff; Henrys Fork Caldera;
                3                                                       3
          280 km ) and 0.6 Ma (Lava Creek Tuff; Yellowstone caldera; 1,000 km ). Yellow-
          stone caldera-related magmas are stored at shallow levels, have low water contents,
          and pre-eruptive temperatures between 810 and 9401C(Hildreth, 1981; Hildreth
          et al., 1984; Bindeman and Valley, 2001). These are only 10–351C lower than the
          liquidus of the bulk-rocks calculated using MELTS (Ghiorso and Sack, 1995;
          Bindeman and Valley, 2001).
   38   39   40   41   42   43   44   45   46   47   48