Page 377 - Caldera Volcanism Analysis, Modelling and Response

352 Thomas R. Walter

1. Introduction

Large-volume ash-flow eruptions and sporadic cones typically align along the
circumference of a caldera basin (e.g., Walker, 1984; Lipman, 1997; Cole et al.,
2005). Studies of caldera structures show that groups of eruption feeder pathways
form discordant intrusive bodies with near-circular geometries in map view.
Intrusions along a circle around a volcanic center were described in detail in the
early twentieth century (Clough et al., 1909); these intrusions are referred to as ring-
complexes. Ring-complexes are very common for volcanoes with surface expressed
calderas (Richey, 1935; Smith and Bailey, 1968; Lipman, 1984), and different types
are distinguished. The geometry of ring-complexes comprise circular or angular,
inwardly dipping, vertical, and outwardly dipping dikes of variable thickness. These
can be singular intrusion events or multiple dikes, forming basaltic or silicic sheets.
Intrusion dynamics of ring-complexes include those that generate their own
propagating fractures or those that reactivate existing fracture zones. In the latter
case, the dike may follow predefined faults of regional tectonic and/or volcano-
tectonic origin.
In association with collapsed calderas, two main types of ring-complexes can be
distinguished (see Table 1): inwardly dipping (often 30–451) concentric dikes,
referred to as cone sheets (Bailey et al., 1924), and near-vertical or often outwardly
dipping concentric dikes intruded parallel to (or into) the ring-faults, referred to as
ring-dikes (Anderson, 1936; Billings, 1943). In the mid-1930s, E.M. Anderson
developed the first mathematical theory for the development of ring-complexes
(Anderson, 1936); a fluctuating pressure within a deep parabolic magma chamber is
thought to be responsible for the formation of ring-complexes. The two different
types of ring-complexes are therefore defined geometrically as well as genetically;
while cone sheets are thought to form during stages of caldera floor uplift
(inflation), ring-dikes form during stages of caldera floor subsidence (deflation).
This work focuses on the different conditions and geometries of ring-dike
formation related to caldera subsidence.
Ring-dikes ideally intrude along the ring-fault and form a closed ring; more
commonly, they only partially intrude into a ring-fracture to form curved or
segmented dikes (Billings, 1943; Bonin, 1986). Ring-dikes are often only a few
centimeters or meters thick. However, old eroded caldera system ring-dikes can
reach massive dimensions, wider than 10 km and more than 0.5 km thick. Classic
ring-dikes were described for volcanic systems in Scotland; for instance, at Glencoe
caldera, a deeply eroded caldera with inverted relief, caldera subsidence affects an
oval-shaped area and activated boundary faults (Clough et al., 1909). Although
similar mechanisms were applied to most intrusive complexes in the British
Volcanic Tertiary Province (Richey, 1935), newer studies suggest that some of these
ring-complexes are lopolithic intrusions associated with inflation and doming
(O’Driscoll et al., 2006).
A major difficulty in ring-dike studies is that exposures are usually poor and
often obscured by sedimentary caldera infill or other intrusive bodies (O’Driscoll
et al., 2006; Kennedy and Stix, 2007). It has been shown that the geometry of a