Page 290 - Caldera Volcanism Analysis, Modelling and Response

A Review on Collapse Caldera Modelling 265

4.1. Results from geophysical imaging of caldera structures
Despite the significant differences in imaging techniques, measurement protocols,
and resulting image resolution, in addition to different tectonic settings, ages, and
evolution of calderas, a number of striking similarities in their subsurface structural
arrangement are evident:
(i) A central negative gravity anomaly inside the collapse structure indicating the
presence of less dense material compared to its surroundings.
(ii) Isolated gravity highs as well as positive magnetic anomalies within the periphery
of the collapse structure, suggesting shallow high-density crystalline bodies.
(iii) A complex arrangement of subsurface structures including:
– Substantial changes in the subsurface relief.
– Vertical and horizontal asymmetries, including structural discontinuities.
– Large- and small-scale anomalous bodies.
– Seismic anomalies occurring at different lithospheric levels (there is usually a
central anomalous body within 4–10 km below ground surface).
– Some degree of magmatic underplating at the interface of the subcalderan
lithosphere and asthenosphere, as well as indications of deep (mantle-based)
roots of central plumbing systems.
(iv) Structural complexities induced by caldera resurgence.

4.1.1. The size, shape, and depth of magma reservoirs
In recent years, seismic imaging in particular has provided a number of clues on the
shape and geometries of magmatic reservoirs beneath active calderas. Seismic waves
are reflected, refracted, or diffracted by subsurface heterogeneities enabling the
mapping of such bodies. P-wave velocity contrasts can also be evaluated as a mean
to assess the liquid fraction in a subsurface reservoir. The resolvable wavelength of
an anomaly is directly proportional to the spacing of seismometers operational
during the survey (Finlayson et al., 2003).

Shallow-seated reservoirs. Shallow (ground surface to a depth of few kilometers)
low-velocity zones were imaged, for example, beneath Campi Flegrei, Rabaul,
Valles, and Long Valley. Their low v p /v s ratios are generally interpreted to result
from a high proportion of pressurised fluid-bearing rock formations. At Campi
Flegrei, the combination of seismic attenuation tomography and borehole data
enabled the discrimination of the reduction in P-wave velocities by hydrothermal
activity from that of conductive cooling of a deeper magma reservoir (de Lorenzo
et al., 2001a, 2001b). Marked changes in v p /v s velocity ratios (with values up to
2.5 km/s) in the first 2–3 km, and lower values (1.6–1.7 km/s) at larger depth were
interpreted as highly fractured layers saturated with water (high v p /v s ) at shallow
depth, and with gas (low v p /v s ) at larger depth (Vanorio et al., 2005). At Long
Valley, a low v p /v s area extending from the surface to ca. 3 km depth is interpreted
as a CO 2 reservoir that is supplying CO 2 -rich springs, venting at the surface, and
killing trees (Foulger et al., 2003). Shallow hydrothermal reservoirs were detected
by magnetotelluric studies at the Las Can ˜adas caldera which also coincide with