Shunsuke Takemura

We investigated the energy rate functions of shallow tremors southeast of the Kii Peninsula in the Nankai subduction zone. Unimodal functions (triangle or bell‐shaped) explain the characteristics of small (4 J) shallow tremors. As the seismic radiated energy increases, the energy rate functions become more complex and are better characterized by a unimodal function with a sharper peak than a triangle function or with small sub‐events. These tremor characteristics differ from those of ordinary earthquakes. We compared the moment‐duration relationships between shallow and deep tremors and found that shallow tremors have larger seismic moments. This difference can be explained by the thermal conditions along the plate boundary fault.

We investigated the energy rate functions of shallow tremors southeast of the Kii Peninsula in the Nankai subduction zone. Unimodal functions (triangle or bell-shaped) explain the characteristics of small (< 10 ⁴ J) shallow tremors. As the seismic radiated energy increases, the energy rate functions become more complex and are better characterized by a unimodal function with a sharper peak than a triangle function or with small sub-events. These tremor characteristics differ from those of ordinary earthquakes. We compared the moment-duration relationships between shallow and deep tremors and found that shallow tremors have larger seismic moments. This difference can be explained by the thermal conditions along the plate boundary fault.

On 8 October 2023, enigmatic tsunamis with maximum wave heights of ~60 cm were observed in the Izu Islands and southwestern Japan, although only mb 4.3–5.4 seismic events were reported near Sofu Seamount in the Izu–Bonin arc in the USGS catalog. To investigate the source process, we analyze tsunami waveforms recorded by ocean-bottom pressure gauges of the DONET array off southwestern Japan. Stacked waveforms reveal recurrent arrivals of multiple wave trains. Deconvolution using a waveform segment of the tsunami from an early isolated event identifies at least 14 successive events that intermittently generated tsunamis over ~1.5 h. Their timings closely coincide with individual events of the seismic swarm and strong seawater acoustic waves (T waves) recorded by ocean-bottom seismometers, indicating a common source. Larger events later in the sequence occurred at intervals comparable to the tsunami period, amplifying later wave phases. Our tsunami waveform analyses summarized above, reported in Sandanbata et al. (2024, GRL), indicate a shallow, repetitive, and atypical non-tectonic tsunami source processes, consistent with volcanic activity. Subsequent independent studies have provided additional constraints supporting a volcanic origin. Recent bathymetric surveys revealed evidence of a submarine eruption near Sofu Seamount (Minami and Tani, 2024, Mar. Geol.). In addition, Kubota et al. (2024, GRL) constrained tsunami source locations to the vicinity of the seamount based on independent tsunami waveform analyses, while Takemura et al. (2024, JGR: Solid Earth) inferred shallow source depths using high-frequency seismic body and T waves. We propose that repetitive volcanic unrest, potentially involving submarine eruptions, caldera deformation, and/or flank collapses, generated the enigmatic tsunamis, although the exact mechanisms remain unresolved.

Recent seismological and geodetic observations have revealed shallow slow earthquakes on subduction megathrusts, offering an opportunity to explore the factors controlling megathrust slip. Identifying these controlling factors helps constrain the potential updip extent of future megathrust ruptures. Here, we investigate the factors controlling the spatial extent of shallow slow earthquakes by integrating tremor locations with uncertainty estimates, detailed bathymetry data, and multichannel seismic data for the 2020–2021 slow earthquake sequence in the Kumano-nada region of the Nankai Trough. Tremor probability maps reveal that the updip, downdip, and lateral limits of tremor activity coincide with distinct structural boundaries. These include the kink or branching of the décollement, a deep-seated strike-slip fault, and the inner–outer wedge boundary. These structures likely act as geometrical or mechanical barriers, further influenced by variations in pore-fluid pressure and by the subducted Paleo–Zenisu ridge. The results indicate that megathrust geometry, material properties of the overriding prism, fluid distribution, and ridge subduction jointly govern the spatial extent of slow earthquakes, emphasizing the need to account for these factors in assessing the potential extent of future megathrust ruptures.

Volcanic activity in submarine calderas recurrently generates characteristic earthquakes on ring‐fault systems driven by magma‐induced stress. At Sumisu Caldera, trapdoor faulting has recurred quasi‐regularly, roughly at ∼10‐year intervals, generating earthquakes of moment magnitude 5.4–5.7 and producing tsunamis due to meter‐scale co‐seismic uplift. To assess event similarity and infer controlling factors, we compare seismic and tsunami waveforms at broadband period range (0.2–500 s). The 1996, 2015, and 2024 events exhibit nearly identical waveforms, indicating rupture of a persistent major asperity with stable geometry and processes. Similar but non‐identical earthquakes suggest partial rupture or additional minor asperities. Comparable repeating sequences at other submarine calderas further support that trapdoor faulting is generally governed by major asperities, while variations in recurrence patterns may reflect differences in fault properties among calderas. These results demonstrate that trapdoor faulting recurrency is controlled by major asperities on ring faults, sustained by long‐term, stable magma supply.

Volcanic activity in submarine calderas recurrently generates characteristic earthquakes on ring-fault systems driven by magma-induced stress. At Sumisu Caldera, M _w 5.4–5.7 trapdoor fault earthquakes have recurred quasi-regularly, mostly at ~10-year intervals, producing tsunamis due to meter-scale caldera uplift. To assess event similarity and infer controlling factors, we compare seismic and tsunami waveforms at broadband period range (0.2–500 s). The 1996, 2015, and 2024 events exhibit nearly identical waveforms, indicating rupture of a persistent major asperity with stable geometry and processes. Similar but non-identical earthquakes suggest partial rupture or additional minor asperities. Comparable repeating sequences at other submarine calderas further support that trapdoor fault earthquakes are generally governed by major asperities, while variations in recurrence patterns may reflect differences in fault properties among calderas. These results demonstrate that trapdoor faulting recurrency is controlled by major asperities on ring faults, sustained by long-term, stable magma supply.

Unexpected major tsunamis have occurred at Sumisu Caldera with an interval of about 10 years. Recently, a similar unexpected major tsunami from Sofu Seamount was reported. Sumisu Caldera and Sofu Seamount are submarine volcanoes far south of Japan. Large T waves from the accompanying seismic events were observed during these tsunamigenic events. Characteristics of T waves could be useful as monitoring metrics for seismicity and tsunami excitation at submarine volcanoes. To understand T-wave generation around submarine volcanoes, I conducted numerical simulations of seismic-acoustic wave propagation from regions around Sumisu Caldera and Sofu Seamount and evaluated T-wave characteristics at various source locations. For sources just below the seafloor, efficient T waves are generated in wider areas of the simulation regions. As depth below the seafloor increases, epicenters with large-amplitude and short-duration T waves are concentrated within the submarine volcanoes of Sumisu Caldera, Sofu Seamount, and Sofugan Volcano. Large-amplitude and short-duration T waves at a certain seismic station are generated from the seismic source located near the steep (> 10º) downslopes oriented toward the seismic station. Large-amplitude and short-duration T waves from regions far from steep downslopes oriented toward the seismic station could not be observed at the station. Both slope angle and downslope direction are important for T-wave generation. The heterogeneous variation of downslopes at submarine volcanoes also causes anisotropic azimuthal patterns of T waves at far stations. The heterogeneities of bathymetric features around source regions cause variability of T-wave characteristics, and consequently, horizontal and vertical locations of seismic events could be constrained by not only body and T-wave travel times, but also T-wave characteristics. Graphical Abstract

Unexpected major tsunamis have often occurred at Sumisu Caldera and Sofu Seamount, submarine volcanoes far south of Japan. Large T waves from the accompanying seismic events were observed during these tsunamigenic events. Characteristics of T waves could be useful as monitoring metrics for seismicity and tsunami excitation at submarine volcanoes. To understand T-wave generation around submarine volcanoes, I conducted numerical simulations of seismic-acoustic wave propagation from regions around Sumisu Caldera and Sofu Seamount and evaluated T-wave characteristics at various source locations. For sources just below the seafloor, efficient T waves are generated in wider areas of the simulation regions. As depth below the seafloor increases, epicenters with efficient T waves are concentrated within the submarine volcanoes of Sumisu Caldera, Sofu Seamount, and Sofugan Volcano. Efficient T waves at a certain seismic station are generated in regions around the steep (> 10º) downslopes against the seismic station. Even in regions with steep downslopes, efficient T waves from regions without downslopes against the seismic station could not be observed. The heterogeneous variation of downslopes at submarine volcanoes also causes anisotropic azimuthal patterns of T waves at far stations. The heterogeneities of bathymetric features around source regions cause variability of T-wave characteristics, and consequently, horizontal and vertical locations of seismic events could be constrained by comparing the characteristics of both observed and synthetic T waves.