Acoustic characteristics of Pyroclastic Density Currents : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Earth Science at Massey University, Palmerston North, New Zealand

Abstract

Pyroclastic Density Currents (PDCs) are a deadly and destructive aspect of volcanic eruptions; however, their study has been hampered by a lack of direct observations and measurements of their internal flow dynamics in the field. Although deposits, videos, and photos offer some insight into PDC behavior, there remains a need to interrogate their poorly constrained internal flow structure and dynamics. Owing to their destructive nature, field-based studies are challenging. Remote measurements through infrasound sensors have been proposed as a possible remedy to close this gap. However, there is a lack of understanding of how and where the infrasound signals recorded from PDCs are generated and what properties they are linked to. This thesis covers the use of large-scale analog experimental PDCs at the PELE facility (the Pyroclastic flow Eruption Large-scale Experiment) to constrain PDC infrasound signal generation. In addition, this thesis expands on the case study literature by revisiting the pressure signals recorded and observed by eyewitnesses from the May 18, 1980 lateral blast of Mount St. Helens.

Through the combination of case study data and three dedicated PDC infrasound experiments, three distinct pressure signals were identified. The first is the displacement of the atmosphere by the initial intrusion of the material into the ambient atmosphere and the transition of the material into a flow. This signal has previously been observed and interpreted as an intrusion displacement; however, the use of experimental PDCs allowed for the confirmation of this signal. This signal was scaled with the initial momentum of the flow within the experiments, with increasing acoustic power and decreasing dominant frequency with increasing momentum. Preliminary analysis results suggest that this relationship holds in natural examples. On larger scales, such as the Mount St. Helens blast, this signal may even be a Lamb wave directed in the direction of the blast. There are currently insufficient datasets to develop a physical or empirical relationship for this behavior; however, this may be possible at volcanoes with a repeating source (e.g., a collapsing dome). The results also suggest that the use of pre-flow arrival signals generated by PDCs could differentiate between pyroclastic surge and pyroclastic flow end members.

The second type of pre-flow arrival signal is a dynamic pressure increase that occurs immediately before the arrival of the front flow. This signal suggests that the acceleration of the atmosphere ahead of a flow can produce a previously unrecognized dynamic pressure hazard due to the acceleration of the atmosphere up to the flow front velocity. Given the range of velocities associated with propagating PDCs, this has the potential to be a significant dynamic pressure before the arrival of the flow front.

Finally, the third type of signal generated by the propagating flow is driven by turbulence along the entire surface boundary with the atmosphere. While the initial signal front can be used to calculate propagation velocity of the flow front, subsequent signals may or may not be new flows, and could be reactions of the flow to underlying conditions like the passing of an energetic underflow.

These findings expand our knowledge of the recordable pressure signals produced by PDCs in the field. These findings need to be further validated and benchmarked in the field; in particular, the limited number of case studies needs to be addressed by increasing the deployments of infrasound arrays.