Page 114 - Caldera Volcanism Analysis, Modelling and Response

Pyroclastic Density Currents 89

unification of the current two major models for PDC deposition: progressive
aggradation and en masse freezing. The second is that a PDC does not move as a
single body but comprises different pulses that move at different velocities down
valley.
Nevertheless, a considerable gap in understanding remains between documen-
ted deposit features, such as sorting, texture and sedimentary structures and the
conceptual models of PDCs. The new approaches, descriptive schemes and models
presented here represent a tentative attempt to illustrate both modern research and
to address some key concepts on PDC behaviour. However, high-quality research
urges us to enhance our understanding of the physical behaviour of PDCs and other
gravity-driven currents. This is needed to ameliorate both our comprehension of
the natural phenomenon and our forecasting ability.
Being that PDCs are physical phenomenon, their nature and behaviour can be
investigated by future research in three primary ways: observation in nature,
reproduction in laboratory and theoretical prediction (numerical modelling). These
three ways are not independent, as any progress in one of these has to be validated
by the other two. Because PDCs are very hazardous phenomena, real time
observations are limited to small or intermediate-scale events, although in the last
25 years, the observation of some PDC-producing eruptions have greatly pushed
ahead the research (e.g. St Helens, 1980; Arenal, 1987–2001; Unzen, 1990–1995;
Colima, 1991; Pinatubo, 1991; Montserrat, 1995-present day; Lipman and
Mullineaux, 1981; Yamamoto et al., 1993; Newhall and Punongbayan, 1996;
Ui et al., 1999; Druitt and Kokelaar, 2002; Saucedo et al., 2004; Cole et al., 2005).
This implies that only real time observation and measurements of physical
parameters of PDCs generated in future explosive eruptions will significantly
enhance our knowledge.
To partially get around this lack of information from direct observation, a
number of laboratory experiments have been performed, ranging from small
laboratory flumes (Gladstone et al., 1998; Choux and Druitt, 2002; Choux et al.,
2004; Roche et al., 2004) to open air experiments (Dellino et al., 2007). Despite
some problems due to the use of synthetic material, limited mass and scaling
processes, laboratory experiments have highlighted important mechanisms in both
transport and depositional mechanisms of PDCs, which have been used to interpret
natural deposits. Nevertheless, further experimentation is required to learn about
the physics of density-stratified currents that have polydispersed clast populations
and a gaseous fluid phase (dusty gas). The reproduction of small PDCs in large-scale
experiments is the main research direction for future. To date, this approach has
allowed the observation of PDC generation and the measurements of some
important parameters (work, rate of turbulence, potential to kinetic energy
transformation, velocity) over flat topography (Dellino et al., 2007). In the near
future, additions of complexities in topography and enhancement of the ability to
record different parameters are very promising avenues to shed new light on PDC
behaviour.
At the same time, numerical modelling has undergone tumultuous growth
during the last 15 years, with complexities that have constantly been added to
numerical codes to obtain a better fit with natural phenomena (e.g. Dobran et al.,