Page 379 - Caldera Volcanism Analysis, Modelling and Response
P. 379

354                                                         Thomas R. Walter


          ring-fault is important for ring-dikes. Based on studies using field constraints
          (Newhall and Dzurisin, 1988), laboratory work (Komuro, 1987; Roche et al.,
          2000; Acocella et al., 2001; Walter and Troll, 2001; Kennedy and Stix, 2003; Geyer
          et al., 2006) and numerical modeling (Komuro et al., 1984; Gudmundsson, 1988;
          Burov and Guillou Frottier, 1999; Kusumoto and Takemura, 2003, 2005), it now
          appears that ring-fault location, dip, and slip are mainly controlled by the depth and
          geometry of the magma chamber. The depth of a magma chamber is usually
          between 2 and 15 km. For example, an intrusive body is suspected to be 2 km
          beneath the Yellowstone caldera, 4–7 km beneath the Long Valley caldera, and 5–
          15 km beneath the Valles caldera (Lipman, 1997). Calderas are often elliptical, for
          instance the 35   73 km Garita caldera, and the presently active Rabaul caldera,
          which is 5   10 km. Besides such geometrical difficulties, combinations of stress
          fields, for instance, those due to a second deep magma chamber or regional
          tectonics, may influence the geometries of ring-complexes near the surface (Marti
          et al., 1996; Gudmundsson, 1998). In order to understand the general formation
          and appearance of ring-faults, the reader is referred to other papers within this
          volume detailing the structural development of caldera-related fault structures.
             Ring-faults are often described as pure dip-slip structures; however, as shown
          by the conceptual model of Figure 1 and by the abundance of intrusive dikes along
          these faults, ring-faults (re)activate as opening mode fractures (Anderson, 1936).
          The necessary space for ring-dike intrusion can be generated during caldera floor
          subsidence. In fact, many former ring-faults are used as dike propagation paths,
          such that outcrops of ring-dikes appear to be more common than outcrops of ring-
          faults. A caldera ring-fault may become active during magma pressure changes
          associated with, for example, input of fresh magma from deeper sources (Sparks et
          al., 1984) or magma evacuation (Druitt and Sparks, 1984). Also, caldera ring-faults
          may slip or open due to tectonic activity, a remote stress field, or pre-existing strain
          history (Gudmundsson, 1988; Newhall and Dzurisin, 1988; Marti et al., 2000;








                                        Collapse caldera

                                    Ring-dike


                                        Magma chamber


          Figure 1  Sketch of ring-dike formation. During de£ation of a large magma chamber
          (decrease of magma pressure), the chamber roof subsides (white arrow) to form subvertical or
          outwardly dipping ring-faults. A ring-dike (shown in dark gray) intrudes into such ring-faults.
          The ring-dike may reach the surface, as shown here, or may occur only underground and
          develop a bell-jar geometry and the surface £exure above (this geometry is shown by the
          dashed lines).
   374   375   376   377   378   379   380   381   382   383   384