Understanding Seismo-Acoustic Explosion Signals and Their Relation to Dome Building and Destruction at Cleveland Volcano, Alaska

Cleveland is one of the most active volcanoes in the Aleutian arc with over 60 explosions since December 2011, yet there is a lack of a working volcanic model. Recent activity is characterized by lava dome growth, nearly continuous degassing, and elevated surface temperatures punctuated by short-lived ash-rich explosions with little to no precursory activity. Because ash from the frequent explosions pose a hazard to aviation, an effort is underway to develop a model for explosion mechanisms that includes observations such as seismic and acoustic signals, dome size, duration of dome growth, repose time, and post-eruptive deposits. This study focuses on the local seismic and acoustic explosion recordings. We find that both seismic and acoustic waveforms are repeatable in shape at low frequencies (0.25-0.5 Hz); however, the lag time between seismic and acoustic onsets can differ by up to 2 seconds between events when recorded at a station 3.5 km away from the summit. This time lag could be due to a change in explosion depth, nonlinearity effects, or atmospheric conditions such as wind that can enhance or delay acoustic travel times. Over the last 4 years, we see a slight decreasing trend of both infrasound energies and recurrence times, which may suggest deeper explosions and an uptick in magma supply. After model development, we implement a source inversion for eruption parameters to better understand the associated hazard. We follow the methods of Kim et al. [2015] to invert the acoustic waveforms using 3D Greens functions that account for topography and assume a simple volumetric source. Estimating a bulk flow density, the resultant volume flow rate is converted to erupted mass, which is a critical parameter for volcanic plume modeling. To validate our infrasound-derived erupted mass, we utilize independent observations from satellite imagery including eruptive deposits, dome observations, and plume parameters. We plan to implement this infrasound-derived erupted mass calculation in near real time for future explosions of Cleveland. The combination of an updated model for explosion mechanisms with an infrasound-derived eruption mass will allow for better understanding of this remote volcano and its associated hazards.

Kim et al. (2015), Acoustic source inversion to estimate volume flux from volcanic explosions, GRL.