Pattern and process in volcano seismology

Abstract

The patterns of occurrence and the underlying processes of two important seismological phenomena at volcanoes, earthquake swarms and volcanic tremor, were investigated. A global database of volcanic earthquake swarm parameters was compiled and was used to evaluate the March 10-14, 1996 seismo-volcanic crisis at Akutan Volcano, Alaska. Earthquake swarm durations and magnitudes were compared with eruptive activity using this database. Trends identified using the database suggest that the Akutan swarm was not precursory and, no eruption occurred. We postulate that a deep instrusion with a large opening component occurred under the flanks of Akutan. The global swarm database has provided an important baseline and has proved to be useful in preparing eruption scenarios for public information releases. The duration-amplitude distribution or frequency-size scaling of volcanic tremor was also examined. The hypothesis tested was that the duration-amplitude distribution may be approximated by an exponential function. The exponential model, implying a scale-bound source process, is found to be a better fit to data then a power-law (scale invariant) model. The exponential model gives a satisfactory description of tremor associated with a wide range of volcanic activity. We propose that exponential scaling of tremor amplitude is due to fixed source geometry driven by a variable excess pressures. This implies that the characteristic amplitude of the duration-amplitude distribution is proportional to a geometric dimension of the source. Broadband seismic data recorded at Arenal volcano, Costa Rica, provide new constrains on tremor source processes. Arenal's tremor contains as many as seven harmonics, whose frequencies vary temporally. This source is inferred to be a shallow, 200-660 m-long resonator, radiating seismic energy from displacement antinodes. We infer that the resonator is gas-charged magma with variable bubble concentration within the conduit and also changes as a function of time, thereby changing the acoustic velocity and the boundary conditions. Polarization analyses for the fundamental mode show particle motion azimuths abruptly rotating, which may be explained by a decrease in incidence angle near the recording site. We suggest that energy for this mode is radiated predominantly from a displacement antinode that is changing position with time.