Volcanic tsunamis, while less frequent than earthquake-triggered ones, can cause devastating damage to coastal communities. The mechanisms behind their generation remain poorly understood, hindering the development of effective early warning systems. The 2022 Hunga-Tonga-Hunga-Ha'apai (HTHH) eruption generated a tsunami that caused significant destruction in Tonga, highlighting the urgent need for improved understanding and prediction capabilities. Previous research proposed various mechanisms for tsunami generation from volcanic eruptions, including subaerial and submarine landslides, pyroclastic flows, caldera or flank collapses, eruptive column collapses, deep-ocean explosions, volcano-tectonic earthquakes, and atmospheric air-pressure waves. However, the dominant mechanism for the 2022 Tonga event remained unclear, with studies offering various explanations based on different aspects of the eruption. This study aimed to investigate the source mechanism of the Tonga tsunami, focusing on the relationship between the atmospheric air-pressure wave and the tsunami itself, and to explore the potential of using air-pressure data for improved tsunami early warning.

The 1883 Krakatau eruption and the 2018 Anak Krakatau eruption serve as stark examples of the destructive potential of volcanic tsunamis. Both events generated large atmospheric waves and significant tsunamis, resulting in substantial loss of life. The generation mechanisms of these tsunamis have been the subject of extensive research, with at least eight different mechanisms proposed. Following the 2022 Tonga eruption, several studies attempted to model the tsunami using different hypotheses, including the atmospheric Proudman resonance effect and various submarine explosion scenarios. These studies revealed inconsistencies and highlighted the complexities of the phenomenon. The lack of a unified understanding and the difficulties in obtaining near-field data during eruptions have hampered progress in developing effective predictive models. This paper directly addresses this gap, focusing on the previously unexplored link between atmospheric pressure waves and tsunamis at the source region.

The authors used an integrated atmosphere-ocean model based on the Regional Ocean Modeling System (ROMS) to simulate the eruption and subsequent tsunami generation. The model incorporated atmospheric and oceanic components coupled through the initial conditions. The atmospheric model utilized the shallow-water equations with atmospheric depth derived using the ideal gas law, while the ocean model employed the same equations but with ocean bathymetry. The initial conditions for both models were derived from seismic data to represent the eruption's effects on the surrounding water and air. Three eruption cycles, reflecting the Tonga air-pressure data, were simulated with decreasing intensity. The model predicted both the air pressure and tsunami waveforms, and the results were validated against various datasets. Air pressure data from Tonga Meteorological Office and tide gauge measurements were used for near-field validation. Far-field validation involved comparisons with Deep-ocean Assessment and Reporting of Tsunamis (DART) data, land-based air-pressure measurements, and satellite altimetry data (from AltiKa, Jason-3, and Sentinel-6 Michael Freilich satellites). A cycle-differentiating method was used for satellite altimetry data to isolate tsunami signals from ocean dynamic features. Empirical equations were used to convert air pressure to equivalent sea level change and predict tsunami runup.

The study found a strong correlation between the air-pressure waves and the tsunami waveforms in Tonga. The three pressure lows in the air-pressure data corresponded closely to the peaks of the tsunami recorded by tide gauges. The model successfully reproduced the observed air-pressure waves and tsunami, both in strength and waveform. This supports the hypothesis of a coupled atmosphere-ocean source mechanism, where lateral air blasts from the eruption generated air-pressure waves and the simultaneous influx of water to fill the newly formed crater generated the tsunami. The mass of ejecta into the atmosphere is equal to the mass of water lost in the ocean, providing a direct link between air pressure and tsunami generation. Validation with DART data and satellite altimetry data confirmed the model's accuracy. Satellite altimetry measurements captured the near-field and far-field tsunami characteristics, supporting the proposed mechanism. The analysis also indicated a lead time of approximately 5-10 minutes between the air-pressure wave and the tsunami arrival in the near field, suggesting the potential for short-term early warning.

The findings provide a novel explanation for volcanic tsunami generation, significantly different from previously proposed mechanisms. The coupled atmosphere-ocean mechanism offers a unified framework for understanding the simultaneous generation of air-pressure waves and tsunamis. The crucial observation is the direct link between air pressure changes and tsunami generation, mediated by the mass balance between the atmosphere and ocean. This link allows for the prediction of tsunami characteristics from readily available air-pressure data. This discovery challenges previously held assumptions about historical volcanic tsunamis and offers the potential to significantly improve early warning systems. The relatively small seafloor deformation observed following the eruption suggests that other mechanisms like landslides played a less significant role in generating the destructive tsunami.

This study unveils a coupled atmosphere-ocean source mechanism for volcanic tsunami generation, where the leading air-pressure wave acts as a reliable predictor of tsunami behavior. The findings support the development of early warning systems using readily available air-pressure sensors. Future research should focus on refining the predictive model by incorporating additional factors and validating it with data from other volcanic events. Further investigation into the specific characteristics of atmospheric waves generated by different types of volcanic eruptions is essential.

The model used simplified assumptions, such as constant water density, and may not fully capture all the complexities of the real-world scenario. The empirical runup formula used for tsunami prediction may also have inherent limitations. The study primarily focuses on the 2022 Tonga event; further testing with data from other volcanic eruptions is necessary to validate the generality of the findings. Additionally, the accuracy of the predictions might be affected by local bathymetry and coastal features, which are not fully considered in the current model.