Title:
|
Volcano infrasound : from laboratory to global scale
|
Volcanic infrasound is detectable at distances up to 10,000 km from the source and increasingly
commonly, volcanoes are monitored at local distances using infrasound. As a relatively new
area of research, many avenues can be explored to improve techniques for monitoring and
interpreting infrasound and a range of approaches are utilised in this thesis.
A database was compiled of 110 events at 36 globally distributed volcanoes. Infrasound from
these events was searched for and characterised using data from infrasound arrays of the
International Monitoring System (IMS) to assess how the IMS could be used to monitor remote
volcanoes. The distance at which volcanic events were detected increased with eruption plume
height. Amplitude decay rates varied widely, but an amplitude decay rate of l/r (where r is
distance to source) was successfully used to compare eruptions. The amplitude and energy of
volcanic infrasound was found to increase with plume height, whilst frequency decreased.
These characteristics allow a broad estimation of eruption intensity using long-range infrasound
data from the IMS alone, and in combination with simple event location methodologies
demonstrate how the IMS could be used to monitor remote volcanoes.
Infrasound from Mount Erebus, Antarctica, is regularly detected at an infrasound array 25 km
distant. Comparing these data with that recorded near the vent allowed study of infrasound
propagation effects, and investigation of how volcanoes could be monitored at this distance.
The rate of amplitude decay observed between near-vent sensors and an infrasound array 25 km
from the vent was greater than expected by purely hemispherical spreading. Decay rates varied
between just less than l/r and approaching 1/r2 over the course of just a few days, indicating
that varying meteorological conditions are likely to have a strong effect. Of the known
infrasound signals, 75% were detected at IS55 and methodology was successfully developed to
automatically detect further events. This demonstrates that although the true amplitude of
volcanic events would be unknown due to varying amplitude decay rates, volcanoes could still
be successfully monitored at this distance in terms of the frequency of occurrence of events.
A set of analogue experiments were conducted to offer insights into the processes that may be
important during infrasound production at Strombolian eruptions and other similar volcanic
events. Bubbles burst at an air-liquid interface whilst being recorded by a pair of microphones
and a high speed camera to investigate the effect of bubble volume, fluid viscosity and rupture
dynamics. The experimental system exhibited complex behaviour. Sound began concurrently
with bubble rupture and varied greatly with radial position around the source exhibiting dipole
characteristics. Both acoustic and rupture characteristics; amplitude, frequency and bubble
rupture speeds all increase with decreasing viscosity. Audio results are most closely reproduced
by a simple model which assumes that the sound produced is due to excitation of the
atmosphere by directional mass flow from the bubble.
|