Tsunami Waveform Stacking and Complex Tsunami Forcings from the Hunga-Tonga Eruption.

On January 15, 2022 the Hunga Tonga Hunga Ha'apai volcano erupted violently, generating a significant tsunami that was observed around the world. This is the first volcanic tsunami that has been observed worldwide by the modern deep- and shallow-water hydrodynamic instrument network. Here we study deep water DART buoy and coastal tide gauges recordings using waveform stacking, amplitude analysis, and spectral analysis. We find that there are potentially as many as three distinct phases to the tsunami each corresponding to different generation mechanisms. Most sites observed an arrival with a propagation speed of ~ 305–310 m/s that we deduced is likely to be an atmospheric Lamb wave with a ~ 40 min period. We find that this feature has a weakly azimuthally-dependent radiation pattern. We also find secondary phases which range in propagation speed from ~ 150 to 205 ms⁻¹. Analysis of their amplitudes and variability in propagation speeds in the waveforms tacks, their decay, and their frequency domain behavior leads us to posit that they represent a complex amalgam of traditional tsunami and atmospheric effects such as internal gravity waves. Overall, from this analysis we find no evidence that processes at the volcanic edifice, such as caldera collapse, land sliding, or pyroclastic density currents, are the dominant tsunamigenic source in the far field. Finally, we also argue that amplification of tsunami wave heights around the coastal areas of the Pacific basin may be due to the combined effects of processes like those seen in meteotsunamis such as Proudman resonance induced by the Lamb wave front, with various coastal areas seeing also effects from Greenspan and shelf resonance. We show a comparison of the event to the last great basin tsunami the M9 Tohoku-oki earthquake, and discuss how, while the amplitudes of the Hunga Tonga tsunami are smaller than other basin-wide events, the worldwide nature, and atmospherically driven forcing of it pose unique challenges to hazards forecasts and warning. [ABSTRACT FROM AUTHOR]