Page 302 - Caldera Volcanism Analysis, Modelling and Response

A Review on Collapse Caldera Modelling 277

conscious that such an image is not a reconstruction of the subsurface prior to the
collapse.

6. Conclusions

Despite the existence of important limitations, the application of experi-
mental, theoretical, and geophysical modelling has represented a significant advance
in the understanding of caldera-collapse processes. In combination with traditional
field studies, the development of experimental and theoretical models has allowed
us to determine the causes and mechanisms controlling caldera collapse. They
provide a clear idea on how and when caldera collapse will occur and on what the
resulting structure will look like. Similarly, geophysical modelling helps to visualise
the internal structure of collapse calderas and can provide clues as to sources of
unrest in active systems.
The combination of field studies with experimental, theoretical, and
geophysical modelling enables identification and quantification of the main
controlling factors of collapse caldera formation. These factors include magma
chamber size and shape, magma chamber depth, host rock rheology, previous
history of deformation, topography, regional tectonics, temperature field around
the magma chamber, and magma composition and rheology. In the same way, the
critical role of the magma chamber shape, roof aspect ratio, and volume fraction of
erupted magma on the resulting caldera structure have also been determined using
experimental and theoretical modelling. It has also been possible to prove that
fractures and faults controlling caldera subsidence may nucleate both at the free
surface and at depth. Conditions for caldera collapse may be achieved in magma
chambers subjected to both overpressure and underpressure.
However, there are some critical aspects that need to be improved in future
models in order to make them more realistic and reliable. For example, new models
should consider host rock mechanical heterogeneities and the pre-existing
deformation history of the volcanic systems, assuming fracture and fatigue of host
rock, as well as considering the effect of a gradual loading of a growing volcanic
edifice. Similarly, future models should be able to include the presence of fluids
(possibility of dyke injection) and to allow coupling between magma chamber
thermodynamics and rock mechanics.
Finally, in order to better understand the dynamic processes at caldera volcanoes,
cross-boundary interaction across many disciplines of Earth sciences is of utmost
importance. Only then can the benefit of each individual technique for providing
answers to the most striking questions on caldera volcanism be addressed.

ACKNOWLEDGEMENTS

This research has been in part funded by the EC project EXPLORIS (EVR1-2001-00047). JG
acknowledges support from a ‘Ramon y Cajal’ fellowship (Spanish Ministry of Education and