Page 327 - Caldera Volcanism Analysis, Modelling and Response
P. 327

302                                                          Valerio Acocella


             Stage 4 results from a subsidence usually in the order of 1 cm or more (in a
          caldera a few cm wide), corresponding to Z1 km in nature. This develops an
          outer normal ring fault, replacing the former peripheral downsag. The fault
          reaches surface forming an additional rim, or scarp; the caldera rim is inward
          dipping, as the underlying fault. Even though this stage is characterised by the
          development of the normal fault, slip along the inner reverse fault still continues;
          therefore, the overall collapse structure consists of a pair of nested calderas (Figure
          10d). No variation in the deformation pattern is observed in the experiments with
          larger subsidence.
             The above estimates of the amount of collapse required to reach a certain stage
          vary as a function of the diameter of the caldera and the thickness of the brittle crust
          analogue. In particular, it is expected that wider calderas require higher subsidence
          to develop ring faults.



               5. Comparison to Nature: Guidelines

               The main structural features of the experiments are the downsags (Stages 1
          and 3) and the ring faults (Stages 2 and 4). Both are visible at well-studied natural
          calderas (Newhall and Dzurisin, 1988).
             The sagged floor may be the only structure accommodating subsidence, as at
          Bracciano (Italy; Di Filippo, 1993) or Buckhorn (Texas; Henry and Price, 1984), or
          may be accompanied by a ring structure within, as at Bolsena (Italy; Walker, 1984;
          Nappi et al., 1991), Rotorua and Reporoa (Figure 11a; New Zealand; Milner et al.,
          2002; Spinks et al., 2005). Many downsags have later turned into ring faults, as
          Sabaloka (Sudan; Almond, 1977), Ishizuchi (Japan; Yoshida, 1984) and Grizzly Peak
          (Colorado; Fridrich and Mahood, 1984; Fridrich et al., 1991), consistently with the
          experimental evolution.
             Regarding ring faults, while the presence of a ring fault at surface may be
          highlighted by the caldera rim, at depth its location is more uncertain. This has
          generated a debate over the fault’s nature, variously described as normal inward
          dipping, reverse outward dipping or vertical (Kennedy and Styx, 2003). Despite
          the common expectation that calderas are bordered by normal faults (Newhall and
          Dzurisin, 1988, and references therein), there is widespread, scale-invariant
          evidence for outward dipping reverse faults. At the 10–100 m scale, subcircular
          collapses and pit craters are bordered by outward dipping reverse faults, as at
          Taupo Volcanic Zone (New Zealand; Figure 11b) and Masaya volcano (Nicaragua;
          Figure 11c; Roche et al., 2001). At larger scales (several km), outward dipping
          ring faults border Rabaul (Figure 11d; Mori and Mckee, 1987; Saunders, 2001),
          Ishizuchi (Yoshida, 1984) and Glencoe calderas. Peripheral normal ring faults may
          be present as well, as at Miyakejima (Figure 11; Geshi et al., 2002) or Rabaul
          (Saunders, 2001), forming pairs of concentric ring faults, as observed in Stage 4
          experiments. Such nested pairs of concentric calderas are common, as at Campi
          Flegrei (Figure 11e), Latera, Pantelleria (Italy), Tavua (Fiji), Batur (Indonesia),
          Fantale (Ethiopia), Taupo (New Zealand), Suswa (Kenya), Guayabo (Costa Rica),
          Daisetsu (Japan), the Archean Hunter Mine Group (Canada) and Karkar and
   322   323   324   325   326   327   328   329   330   331   332