Page 340 - Caldera Volcanism Analysis, Modelling and Response
P. 340

Magma-Chamber Geometry, Fluid Transport, Local Stresses and Rock Behaviour  315





























             Figure 1  Collapse caldera is normally associated with a magma chamber of a cross-sectional
             geometry similar to the caldera fault itself.While most collapse caldera faults are dip slip,
             some may be outward-dipping reverse faults whereas others are inward-dipping normal faults,
             like the one shown here. Many calderas occur in volcanic ¢elds located above large
             magma-accumulation zones.The crustal segment hosting the ¢eld may, occasionally, be subject
             to magmatic pressure that gives rise to slight doming (uplift) of the ¢eld.

             to be significant mechanical and geometric differences between ring faults and most
             earthquake faults. The first difference is that the calderas faults normally form closed —
             or near to closed — rings or ellipses. By contrast, earthquake faults are comparatively
             straight, if often composed of offset segments, and do not form closed loops.
                A second difference is that the displacement during a single caldera-collapse
             event is normally much larger than the displacement during a single earthquake. In
             both cases, the fracture is primarily a shear fracture and the displacement is driven
             by shear stress concentration at the fault plane. In a very intense earthquake, the
             displacement (slip) may reach a maximum of 10–20 m. By contrast, many ring-fault
             displacements during a single caldera collapse reach hundreds of metres and some
             several kilometres.
                The third difference concerns the fault dip. Ring faults are dip-slip faults
             (Figures 1 and 2). Dip-slip earthquake faults normally dip 45–751 (and less for
             thrust faults). By contrast, most ring faults are close to vertical (Figure 2). Since the
             mean angle between the direction of the maximum principal stress, s 1 , and the fault
             plane is unlikely to be different between caldera faults and earthquake faults, these
             observations indicate that there is a fundamental difference between the local stress
             fields controlling earthquake faulting and caldera faulting as regards orientation of
             the principal stresses at the time of fault formation or slip.
                Many models and mechanisms have been proposed to explain ring-fault
             formation and caldera collapses (Anderson, 1936; Williams, 1941; Druitt and
             Sparks, 1984; Walker, 1984; Scandone, 1990; Chery et al., 1991; Marti et al., 1994;
   335   336   337   338   339   340   341   342   343   344   345