Page 211 - Caldera Volcanism Analysis, Modelling and Response

186 W.U. Mueller et al.

3. Notion of Calderas

Stratovolcanoes, composite volcanoes or shield volcanoes are the precursor
structures to calderas. The former two favour large-volume silicic-dominated,
paroxysmal ash-flow calderas (e.g. Valles caldera; Smith and Bailey, 1968; Long
Valley caldera, Bailey, 1989) whereas the latter is generally associated with mafic
summit calderas (e.g. Kilauea, Hawaii; Tilling and Dvorak, 1993). Caldera
geometry varies from central circular depressions such as the 6 7 km in diameter,
submarine Myojin Knoll, (Fiske et al., 2001) to the small 9–16 km ellipsoidal Las
Can ˜adas caldera (Martı ´ and Gudmundsson, 2000), the medium-scale 17 32 km
Long Valley Caldera (Bailey, 1989) and the large 30 100 km, Toba caldera
(Lipman, 2000). Calderas are rarely isolated and occur (1) in clusters, as indicated by
the Taupo volcanic zone (Cole et al., 2005) or the Minami-Aizu field (Miura and
Tamai, 1998), (2) nested, as exemplified by the Campi Flegrei field (Orsi et al.,
1996) or (3) as overlapping collapse calderas such as Las Can ˜adas caldera (Martı ´ and
Gudmundsson, 2000). In addition, complex mega-nested calderas of the Olympic
Mons type on Mars may develop, as indicated by the Abitibi Blake River caldera
complex (Pearson, 2005). Volcano-depressions are caused by numerous processes
3
including: (1) the rapid explosive evacuation of magma producing hundreds of km
of pyroclastic debris (e.g. subaerial silicic arc, Taupo calderas; continental rift
systems, Valles caldera), (2) continuous outpouring of magma via extensive lava
flows and fountaining eruptions (some Archean arc calderas), (3) draining of magma
into satellite chambers along rift zones (oceanic hot spot, e.g. Hawaii) and (4)
magma migration causing a shift and overlap of calderas (oceanic hot spot, e.g. Las
Can ˜adas caldera).
Our understanding of caldera development has advanced significantly based on
recent analogue experiments (Acocella et al., 2000, 2001; Acocella, 2008; Roche
et al., 2000; Walter and Troll, 2001; Kennedy et al., 2004), numerical modelling of
stress regimes associated with caldera-forming eruptions (Gudmundsson, 1998,
2008), studies of collapse processes and subsidence geometry (Lipman, 1997) and
direct monitoring of caldera collapse (Geshi et al., 2002). The importance of early
reverse faults was recognised by Mueller and Mortensen (2002), and Stix et al.
(2003) in their role during the circulation of hydrothermal fluids and formation of
VMS deposits structures preceded normal ring faults. Modelling of Acocella et al.
(2000, 2001) and Roche et al. (2000) confirmed the inference of Branney (1995)
that during draining of a shallow magma chamber, an early set of outward-dipping
faults with a reverse sense of displacement develop and encircle the subsidence
centre. Ring faulting is a rapid, syneruptive phenomenon, and is associated with the
formation of chaotic breccias or talus scree breccias, as newly formed caldera walls
suffer large landslides. The intracaldera depression contains further syneruptive
debris as a result of synthetic and antithetic fault formation, which forms complex
horst and graben structures. Gudmundsson (1998) argues that the encircling normal
faults are used for magma venting. The greatest displacements typically occur along
the outermost set of encircling ring faults (Walker, 1984). The collapse mechanism
of calderas varies with depth of magma chamber (e.g. Roche et al., 2000), its size,