[M-GI25] Holocene paleoenvironment, paleoclimate, and paleohazards in the Pacific Islands

[MGI25-P02] Numerical Simulation of Large-Scale Tsunami Propagation Caused by the 15^th-Century Kuwae Eruption

*Yuri Kijima ¹, Hideaki Yanagisawa², Koki Nakata¹, Kazuhisa Goto¹ (1.The University of Tokyo, 2.Tohoku Gakuin University)

Keywords:tsunami, numerical simulation, south pacific

Tsunami deposits from a massive 15^th-century tsunami have been recorded in various locations across the South Pacific (Goff et al., 2022). Potential sources of this tsunami include earthquakes along the Tonga Trench and eruptions of the Kuwae submarine volcano. However, the possibility of a tsunami generated by the latter has not been thoroughly examined to date. Kuwae volcano is a submarine caldera situated between Epi and Tongoa Islands in the Republic of Vanuatu, northeast of the Australian continent. The caldera's estimated area, volume, and depth are approximately 45 km², 32-39 km³, and 650-950 m, respectively (Monzier et al., 1994). It is believed that the caldera collapsed during a 15^th-century eruption, and prior to this event, an island with an elevation of 500-600 m, comparable to Epi and Tongoa Islands, likely existed in its place. There are historical, archaeological, and geological records of the 15^th century eruption. Geological evidence suggested it was a large-scale event with a Volcanic Explosivity Index (VEI) of 6-7 (Ballard et al., 2023). Similar large eruptions, such as the collapse of the Kikai Caldera, are known to have generated large tsunamis (Maeno et al., 2006), suggesting that the 15^th-century Kuwae eruption likely caused a large tsunami as well. However, no studies have quantitatively evaluated the specific scale of the Kuwae tsunami. This study aims to quantitatively assess the impact of the Kuwae caldera collapse on surrounding islands through numerical simulations of large-scale tsunami propagation.

Several mechanisms have been proposed for tsunami generation during caldera formation (Schindele et al., 2024), with the most energy-intensive scenario being the collapse and subsequent subsidence of the caldera (Maeno et al., 2006). Therefore, this study assumes that the primary cause of the tsunami generated during the Kuwae eruption was caldera collapse. A pre-collapse topographic model was developed using 15-arcsecond resolution bathymetric data from GEBCO2024, which was converted to a 5-arcsecond resolution grid using GMT. A Gaussian-shaped elevation approximately 600 m high was modeled between Epi and Tongoa Islands to represent the pre-collapse topography. Since the time scale of the caldera is unknown (Maeno et al., 2006), we conducted numerical simulations under two primary assumptions: free-fall and uniform velocity subsidence. The uniform velocity model was represented by the equation: h(t)=h_before-{(h_before-h_after)/t_s}, where h_before is the pre-collapse elevation, h_after is the post-collapse elevation, and t_s is the total subsidence duration (in seconds), tested under conditions of 300, 600, 1200, and 3600 seconds. For the free-fall model, the equation used was: h(t)=h_before-(1/2)gt² , where g is the gravitational acceleration. In both models, subsidence was halted once caldera reached its current depth. These topographic changes were input into a nonlinear long-wave tsunami model for simulation.

The results demonstrated that faster subsidence rates produced higher maximum wave heights on surrounding islands. On the coastal areas of Epi and Tongoa Islands, which are both in close proximity to Kuwae, tsunami heights ranging from several meters to tens of meters were observed under all conditions. These findings suggest that the tsunami deposits identified in previous studies, such as those in Southern Epi Island (Goff et al., 2008), could potentially be explained by the tsunami generated from the 15^th-century Kuwae eruption.

Acknowledgments: This research was supported by SATREPS, JST/JICA and JSPS.