Detection, source location, and analysis of volcano infrasound

Abstract

The study of volcano infrasound focuses on low frequency sound from volcanoes, how volcanic processes produce it, and the path it travels from the source to our receivers. In this dissertation we focus on detecting, locating, and analyzing infrasound from a number of different volcanoes using a variety of analysis techniques. These works will help inform future volcano monitoring using infrasound with respect to infrasonic source location, signal characterization, volatile flux estimation, and back-azimuth to source determination. Source location is an important component of the study of volcano infrasound and in its application to volcano monitoring. Semblance is a forward grid search technique and common source location method in infrasound studies as well as seismology. We evaluated the effectiveness of semblance in the presence of significant topographic features for explosions of Sakurajima Volcano, Japan, while taking into account temperature and wind variations. We show that topographic obstacles at Sakurajima cause a semblance source location offset of ~360-420 m to the northeast of the actual source location. In addition, we found despite the consistent offset in source location semblance can still be a useful tool for determining periods of volcanic activity. Infrasonic signal characterization follows signal detection and source location in volcano monitoring in that it informs us of the type of volcanic activity detected. In large volcanic eruptions the lowermost portion of the eruption column is momentum-driven and termed the volcanic jet or gas-thrust zone. This turbulent fluid-flow perturbs the atmosphere and produces a sound similar to that of jet and rocket engines, known as jet noise. We deployed an array of infrasound sensors near an accessible, less hazardous, fumarolic jet at Aso Volcano, Japan as an analogue to large, violent volcanic eruption jets. We recorded volcanic jet noise at 57.6° from vertical, a recording angle not normally feasible in volcanic environments. The fumarolic jet noise was found to have a sustained, low amplitude signal with a spectral peak between 7-10 Hz. From thermal imagery we measure the jet temperature (~260 °C) and estimate the jet diameter (~2.5 m). From the estimated jet diameter, an assumed Strouhal number of 0.19, and the jet noise peak frequency, we estimated the jet velocity to be ~79 - 132 m/s. We used published gas data to then estimate the volatile flux at ~160 - 270 kg/s (14,000 - 23,000 t/d). These estimates are typically difficult to obtain in volcanic environments, but provide valuable information on the eruption. At regional and global length scales we use infrasound arrays to detect signals and determine their source back-azimuths. A ground-coupled airwave (GCA) occurs when an incident acoustic pressure wave encounters the Earth's surface and part of the energy of the wave is transferred to the ground. GCAs are commonly observed from sources such as volcanic eruptions, bolides, meteors, and explosions. They have been observed to have retrograde particle motion. When recorded on collocated seismo-acoustic sensors, the phase between the infrasound and seismic signals is 90°. If the sensors are separated wind noise is usually incoherent and an additional phase is added due to the sensor separation. We utilized the additional phase and the characteristic particle motion to determine a unique back-azimuth solution to an acoustic source. The additional phase will be different depending on the direction from which a wave arrives. Our technique was tested using synthetic seismo-acoustic data from a coupled Earth-atmosphere 3D finite difference code and then applied to two well-constrained datasets: Mount St. Helens, USA, and Mount Pagan, Commonwealth of the Northern Mariana Islands Volcanoes. The results from our method are within ~<1° - 5° of the actual and traditional infrasound array processing determined back-azimuths. Ours is a new method to detect and determine the back-azimuth to infrasonic signals, which will be useful when financial and spatial resources are limited.