Recurrent earthquake swarms and transient crustal deformation off the Tokara Islands, southern Japan: Insights from the 2025 swarm sequence

Recurrent earthquake swarms and transient crustal deformation off the Tokara Islands, southern Japan: Insights from the 2025 swarm sequence | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Recurrent earthquake swarms and transient crustal deformation off the Tokara Islands, southern Japan: Insights from the 2025 swarm sequence Yutaro Okada, Yusaku Ohta, Miku Ohtate, Yoshiaki Ito, Mako Ohzono, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8399343/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 5 You are reading this latest preprint version Abstract Transient crustal deformation is often observed in conjunction with earthquake swarms and is attributed to aseismic processes that drive swarm activity, such as magma intrusion, fluid migration, or slow slip events. Off the Tokara Islands, along the volcanic front of the Ryukyu subduction zone in the southern Japanese Archipelago, earthquake swarms have repeatedly occurred; however, the driving mechanism of this anomalous seismicity remains unclear. Here, we investigate this mechanism through integrative analyses of Global Navigation Satellite System (GNSS) data from the Geospatial Information Authority of Japan and SoftBank Corp. The daily GNSS position time series from multiple stations exhibits clear transient signals that coincide with recurrent earthquake swarms. Notably, during the June–July 2025 episode, both the temporal trend and spatial pattern of the displacement changed abruptly at the onset of the secondary swarm. Motivated by these distinctive characteristics, we estimated the source models for the 2025 episode by dividing it into three phases. For all phases, the displacements calculated from rectangular faults with shear slip reproduce the observed displacements well. Vertical open cracks provide an alternative source model, but their data fit is systematically worse than that of the shear-slip models. A stacked time series of 30-second kinematic GNSS positions further suggests that the temporal relationship between aseismic deformation and seismicity can vary even within a single episode. Based on the shear-slip and open crack models for the 2025 episode, we propose two scenarios for the driving mechanisms of aseismic displacements and earthquake swarms. In both scenarios, we speculate that the supply of material from a region deeper than the aseismic source and swarm hypocenters plays a key role in generating these episodes. Earthquake swarm Transient displacement GNSS GEONET Private-sector GNSS Aseismic fault slip Open crack Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1 Introduction Earthquake swarms are sequences of earthquakes without a clear mainshock and are often synchronized with significant transient crustal deformation. For example, synchronizations between earthquake swarms and transient deformation have been observed around volcanoes and are interpreted to be related to magma intrusion (e.g., Aoki et al. 1999 ; Ellis et al. 2024 ; Isken et al. 2025 ) or hydrothermal fluid migration (Kawai et al. 2024 ). Such synchronizations have also been reported in regions without major volcanoes and are interpreted to be caused by fluid migration (Cappa et al. 2009 ; Nishimura et al. 2023 ). Slow slip events (SSEs), which represent aseismic transient fault-slip phenomena that occur without significant external stress loading, are sometimes accompanied by active seismicity (e.g., Ozawa et al. 2003 ; Delahaye et al. 2009 ; Vallée et al. 2013 ). Thus, analyses of geodetic data that can capture seismic and aseismic processes, are crucial for revealing the driving mechanisms of earthquake swarms. The Global Navigation Satellite System (GNSS) is a game changer in geodesy. For example, Japan’s nationwide GNSS network, the GNSS Earth Observation NETwork System (GEONET) operated by the Geospatial Information Authority of Japan (GSI), is composed of ~ 1,300 stations and has contributed to revealing interseismic, coseismic, postseismic, and volcanic deformation since the 1990s (e.g., Heki et al. 1997 ; Aoki et al. 1999 ; Hirose et al. 1999 ; Sagiya et al. 2000 ; Nishimura et al. 2001 , 2011 ; Sun et al. 2014 ). More than 3,300 GNSS stations operated by SoftBank Corp. (hereafter “SoftBank”) have recently begun to be used for scientific purposes (Ohta and Ohzono 2022 ). Analyses that integrate GEONET and SoftBank stations have contributed to a significantly improved understanding of aseismic transient deformation, coseismic slip distribution, postseismic deformation, and volcanic activity (Nishimura et al. 2023 ; Yamada et al. 2025 ; Ohzono et al. 2025 ; Ohtate et al. 2025 ). However, the SoftBank GNSS stations located in the southern Japanese Archipelago have not been used in published studies, and their ability to resolve geodetic phenomena remains uncertain. The Ryukyu subduction zone extends from southern Kyushu Island to Taiwan. Its crustal deformation is governed by two major tectonic processes: subduction of the Philippine Sea plate beneath the Okinawa plate along the Ryukyu Trench, and back-arc spreading between the Okinawa and Yangtze plate along the Okinawa Trough (e.g., Nishimura et al. 2004 ; Kano et al. 2021 ) (Fig. 1 a). In the southern Okinawa Trough, seismic and geodetic observations have indicated a dike intrusion coincident with an earthquake swarm, which is accompanied by long-term temporal changes in seismic and SSE activity attributed to this intrusion (Ando et al. 2015 ; Tu and Heki 2017 ; Nakamura and Kinjo 2018 ). A volcanic front has developed in this subduction zone, and the Tokara Islands are part of this zone (Fig. 1 a). Earthquake swarms, including events of M ≧ 5, have repeatedly occurred off the Tokara Islands, particularly after 2021, with shorter recurrence intervals (Figure S1 ). The most recent episode in 2025 occurred between Takara and Akuseki Islands and persisted for approximately one month (Fig. 1 ), resulting in the evacuation of several dozen residents. Despite these frequent and societally impactful episodes, the driving mechanisms of earthquake swarms off the Tokara Islands remain unclear. Given this background, we investigate crustal deformation accompanying earthquake swarms off the Tokara Islands by jointly analyzing GNSS datasets from GEONET and SoftBank stations (Fig. 1 b) to elucidate their driving mechanisms. We also used earthquake and centroid moment tensor (CMT) catalogs provided by the Japan Meteorological Agency (JMA) and the National Research Institute for Earth Science and Disaster Resilience (NIED), respectively, to compare crustal deformation with seismicity. In this paper, we demonstrate a clear synchronization between earthquake swarms and transient crustal deformation, estimate source models for aseismic deformation, and propose multiple scenarios for the driving mechanisms. 2 Data analysis 2.1 GNSS data and preprocessing of daily positions We analyzed GNSS data from 30 GEONET and 59 SoftBank stations located in and around the Tokara Islands from November 1, 2019, to October 29, 2025 (Figure S2 ). Daily positions were estimated from GPS observations using GipsyX (version 2.3), employing a Precise Point Positioning with Ambiguity Resolution (PPP-AR) strategy (Bertiger et al. 2020 ). The resulting daily positions were aligned into the International Terrestrial Reference Frame 2020 (Altamimi et al. 2023 ). In addition, high-rate GNSS (HR-GNSS) positions, sampled every 30 s, at BQ1I, 1243, BQ1J, and BQ1K (Figs. 1 b and S2) were used to examine detailed temporal correlations with seismicity (Section 4.2 ). These high-rate positions were estimated by applying the kinematic PPP-AR approach implemented in PRIDE-PPPAR (version 3.1.0) (Geng et al. 2019 ) to multi-GNSS (GPS, GLONASS, and Galileo) observations. We removed common-mode errors (Okada and Nishimura 2025 ), maintenance offsets at GEONET stations, and seasonal oscillations (Nishimura et al. 2013 ; Okada and Ueda 2025 ) from the daily positions as preprocessing (Text S1, Figures S3 and S4, and Table S1 ). The common-mode errors were calculated using data from stations away from the swarms (e.g., Fasola et al. 2023 ), which were denoised based on step master files (GSI, 2025; Ohta and Ohtate, in-press) and statistical and empirical techniques (Cleveland et al. 1990 ; Okada 1995 ). For station 1243, we did not estimate the seasonal component because of the short observation period. 2.2 Comparison of time series and displacement fields with seismicity The daily position time series which is detrended using linear velocities estimated from December 2024 to May 2025, show clear synchronicity between earthquake swarms and transient crustal deformation in multiple episodes (Figure S5). Focusing on the north–south (NS) components at stations on Takara Island (Figures S5 and S6), the 2025 episode exhibits particularly pronounced signals (~ 40 mm), which are unique among episodes since November 2019. In addition, the inter-episode velocities at multiple stations exhibit slight temporal variations. Transient motions without any clear accompanying increase in seismicity also appear in the time series, most notably at BQ1J on Kodakara Island. As our primary objective is to investigate crustal deformation coincident with earthquake swarms, we do not examine these signals in detail. Next, we examine the temporal changes during the 2025 episode by dividing it into three phases (Fig. 1 c-e). In Phase 1 (June 20, 2025–June 30, 2025), a swarm episode was initiated between Takara and Akuseki Islands. Coincident with the increase in seismicity, the GNSS stations on Kodakara and Akuseki Islands show a gradual but large displacement. In contrast, the NS component of BQ1I on Takara Island does not exhibit any noticeable signal during this phase. Subsequently, in Phase 2 (July 1, 2025–July 3, 2025), the seismic activity between Takara and Akuseki Islands migrated ~ 10 km eastward, and a secondary swarm began west of Takara Island (Fig. 1 d). Simultaneously, the GNSS time series on Takara and Kodakara Islands show rapid trend changes that are markedly different from the deformation in Phase 1, whereas BQ1K on Akuseki Island continues to show a persistent signal with a slow rate. Taken together, the GNSS data suggest that at least two sources drive deformation at different rates. In Phase 3 (July 4, 2025–July 20, 2025), earthquakes west of Takara Island ceased, whereas activity between Kodakara and Akuseki Islands persisted. The GNSS time series in Phase 3 resemble those in Phase 1: slow deformation at stations on Kodakara and Akuseki Islands and no marked signal at BQ1I on Takara Island. After Phase 3, the transient crustal deformation ceased, and no reactivation was observed until October 29, 2025. However, M ~ 3 earthquakes continued to occur around the Tokara Islands for at least three months after their onset (Figure S5). Finally, we examine the displacement fields for each phase of the 2025 episode and for the April 2021, December 2021, and September 2023 episodes (Fig. 2 ). Using the detrended dataset, we calculated the difference between the average daily positions over the two time windows defined for each episode (Table S2 ) and treated these differences as displacements. In previous episodes, the GNSS stations on Kodakara and Akuseki Islands are displaced northeastward and southwestward, respectively, although the relative magnitudes of the horizontal displacement vary among episodes. Furthermore, the horizontal displacement amplitudes on Takara Island are smaller than those on Kodakara Island, whereas the displacement azimuths are similar. No systematic patterns were observed in the vertical displacement field. The displacement fields in Phases 1 and 3 of the 2025 episode share these characteristics, whereas those in Phase 2 shows a previously unseen pattern: stations on Takara and Kodakara Islands are displaced southward and northward, respectively. Additionally, the displacement of BQ1K on Akuseki Island continues to trend southwesterly, similar to the Phases 1 and 3. 3 Source model of the 2025 transient deformation episode As described above, crustal deformation during the 2025 earthquake swarm exhibits distinctive characteristics, particularly in Phase 2. Therefore, focusing on the 2025 episode, we estimate the source models of aseismic deformation in each phase and discuss the results in the following sections. 3.1 Inversion strategy To isolate the aseismic processes, we first calculated the coseismic displacements predicted from F-net CMT solutions (Fukuyama et al. 1998 ) of earthquakes off the Tokara Islands using a point source dislocation model in a homogeneous elastic half-space (Okada 1992 ). We then stacked the displacements due to all listed earthquakes within the time window for each phase (Table S3 ) and subtracted these from the observed displacements (Figure S7). The observational errors were computed from the standard deviations of the daily positions over the two time windows used to define each displacement (Table S2 ). We employed a nonlinear least-squares inversion (Matsu’ura and Hasegawa 1987 ) to estimate rectangular dislocation source models (Okada 1992 ), while simultaneously adopting a grid-search strategy because the inversion is highly sensitive to prior constraints. To reduce bias due to the assumed source location, we tested multiple combinations of longitude and latitude (cf. Nishimura 2014 ; Takagi et al. 2019 ) (Tables S4 and S5). In addition, we determined the optimal fault-top depth below sea level using a grid search, rather than treating it as an inversion parameter, to avoid unrealistic models. We further imposed a strict prior constraint on the strike so that the resulting models remained consistent with regional tectonic structures (Table S6; Minami et al. 2021 ; Koge et al. 2025 ). Moreover, to limit the number of free parameters, we assumed that the dislocation mechanism belongs to one of four types: normal, dextral, or sinistral faulting, or a vertical open crack. Finally, we estimated three-dimensional translation together with the rectangular faults to account for possible remaining common-mode errors (cf. Nishimura et al. 2013 ). For each assumed dislocation mechanism, we systematically selected a model that minimized the chi-square statistic among the solutions obtained for different prior source locations. The preferred dislocation mechanism was selected based on the Akaike Information Criterion (AIC) and its consistency with regional tectonics (Section 3.2 ). Using this strategy, we estimated a source model and three-dimensional translations for Phases 1 and 3 (Figure S8 and Tables S4–S6). In contrast, the GNSS and seismicity patterns in Phase 2 suggest the presence of multiple sources. Therefore, we fixed the location, geometry, and dislocation mechanism of one source to be identical to those of the Phase 3 model because the earthquake distribution between Kodakara and Akuseki Islands in Phase 3 closely resembles that in Phase 2 (Figs. 2 e and 2 f), and estimated only the dislocation magnitude under the same prior constraint as in Phases 1 and 3 (Table S6). We then introduced an additional source and three-dimensional translations and estimated their parameters simultaneously using the same inversion strategy as in Phases 1 and 3. 3.2 Estimated source models The source models for all trial combinations in all phases are shown in Figures A1–A42, and the selected models are summarized in Fig. 3 . In Phases 1 and 3, the northwest-dipping faults with dextral slip yield the minimum AIC values (Figure S9), although the pattern of horizontal displacements differs between the phases (Figs. 3 a and 3 c). We also consider northwest-dipping faults with normal slip, which give the fourth- and second-smallest AIC values in Phases 1 and 3, respectively (Figs. 3 d and 3 f). We excluded the sinistral-slip models in Phase 1 from further discussion because its negative slip (Figure A2) and lateral faulting with an extremely low dip angle (Figure A3) is physically unreasonable in terms of the direction of the P- and T-axes, even though its AIC values is smaller than that in the normal-slip case. The calculated horizontal displacements from the dextral- and normal-fault models reproduce the observed horizontal displacements with comparable residuals (Figs. 3 a and 3 d). In contrast, the vertical displacement at BQ1J on Kodakara Island calculated from the normal-faulting model is significantly larger than the observed displacements (Figure A4b), whereas the vertical displacements from the dextral-faulting model are consistent with the observation (Figure A6b). In Phase 3, however, the fits to the observations obtained from the dextral- and normal-fault models are similar in terms of both visual inspection and AIC values (Figs. 3 c, 3 f, and S9). In Phase 2, the AIC values remain small when sinistral slip is selected for the western fault, irrespective of the assumed mechanism for the eastern fault between Kodakara and Akuseki Islands (Figure S10). Note that a dextral fault striking 315° in combination with an eastern normal fault also yields a small AIC; however, we interpret this case as sinistral faulting because the estimated slip magnitude is negative (Figure A28). Therefore, based on the data misfit, we adopted the sinistral-slip model striking 315° as the western source model (Figs. 3 b and 3 e). Hereafter, we refer to the source models with dextral faulting on the eastern fault (Figs. 3 a–c) as the minimum-misfit (MM) model. The MM model is acceptable in light of the shallow seismic velocity structure, because lateral faulting on the back-arc side of the volcanic front has been suggested by seismic reflection images, in addition to normal faulting (Arai et al. 2018 ). Meanwhile, the normal-faulting models (Figs. 3 d–f) are also acceptable owing to their small AIC values. Moreover, these models are more consistent with the regional tectonics around the Tokara Islands (Nishimura et al. 2004 ; Minami et al. 2021 ), the stress field inferred from earthquake focal mechanisms (Koge et al. 2025 ), and previous geodetic observations of other volcanic regions (Rubin and Pollard 1988 ; Himematsu et al. 2019 ). For these reasons, we refer to the normal-faulting models as the preferred model. The magnitude of the aseismic slip ranges from M w 6.0 to 6.4, depending on the phase and mechanism. On the other hand, the cumulative moment of earthquakes in Phases 1, 2, and 3 is 1.17×10 18 , 1.81×10 18 , and 1.09×10 18 Nm, respectively. The ratio between the aseismic and seismic moments varies from 1.1 to 3.9 (Tables S7 and S8), which is larger than typical afterslips regardless of phase and mechanism (e.g., Churchill et al. 2022 ). In addition, the estimated fault planes are inconsistent with the hypocenter distribution of earthquakes in the JMA catalog. However, this comparison has uncertainty associated with hypocenter locations because the catalog is not relocated (Figures A4, A6, A16, A17, A39, and A41). We evaluate uncertainties in the fault plane extent of the MM and preferred models using a remove-one-station jackknife test (Figure S11). The results show that the fault lengths in Phases 1 and 3 and the location of the western fault in Phase 2 vary depending on which station is removed, whereas the fault widths in Phases 1 and 3 and the length of the western fault in Phase 2 remain almost unchanged. The former indicates that these parameters strongly depend on displacements observed at a limited number of stations particularly BQ1J, on Kodakara Island. In contrast, the latter suggests that these parameters are largely controlled by prior constraints (Table S6) rather than by observational data. 4 Discussion 4.1 Source models for the 2025 episode assuming vertical open cracks In addition to shear-slip models, we examine vertical open crack sources in each phase (Figs. 3 g–i), anticipating the possibility of dike intrusion because the Tokara Islands are part of the volcanic front in the Ryukyu subduction zone (Fig. 1 a) and earthquake swarms sometimes coincide with magma intrusions (e.g., Aoki et al. 1999 ; Ellis et al. 2024 ; Isken et al. 2025 ). The calculated displacements from open cracks with a top depth of 1 km below sea level (Table S9) partially reproduce the observed displacements in each phase; however, their AIC values are significantly larger than those of the MM and preferred models (Figures S9 and S10). No evidence indicative of magma discharge into the sea, such as newly formed submarine topography, discolored seawater (e.g., Urai and Machida 2005 ), or pumice rafts (e.g., Nishikawa et al. 2023 ), had not been reported as of December 19, 2025 (Japan Coast Guard, 2025 ). To consider this information, we evaluate the data fit of the crack models as a function of their top depth using the chi-square ratio \(\:{R}_{d}\) , defined as \(\:\begin{array}{c}{R}_{d}=\frac{{\chi\:}_{d}^{2}}{{\chi\:}_{min}^{2}},\:\#\left(1\right)\end{array}\) where \(\:{\chi\:}_{d}^{2}\) and \(\:{\chi\:}_{min}^{2}\) are the chi-square statistic of the best-fitting model with a top depth of \(\:d\) km in a given phase and of the overall minimum chi-squared model in that phase, respectively. For the Phase 2 source model, we varied the top depth of the western fault while fixing that of the eastern fault at 1 km. The comparison (Figure S1 2) shows that the top depth must be close to 1 km, which is comparable to the water depth around the islands (Figure S1 a), to reproduce the amplitude of the observed displacements (Figures S13 and S14). In conclusion, our results do not support a scenario in which magma is discharged into the sea at a high rate, given that the GNSS time series does not show any reversal of the transient trend after Phase 3 (Figs. 1 c and S5) and considering other available observations. Nevertheless, magma intrusion or a fluid supply reaching shallow depths cannot be ruled out based solely on GNSS data. However, the geodetic data are more consistently explained by the shear-slip models than by the open crack models (Figures S9 and S10). Therefore, we regard the vertical open crack model as an alternative model and discuss its implications further in Section 4.4 . 4.2 Temporal relationship between transient deformation and seismicity clarified by HR-GNSS data Investigating the temporal relationship between transient deformation and seismicity is essential for discussing causal linkages. Therefore, we examine this relationship using stacked HR-GNSS data (cf. Okada et al. 2022 ) (Text S2, Figure S15, and Table S10). We calculated a moving median with a 3-hour window for the stacked time series (Itoh et al. 2025 ) to suppress short-term fluctuations, which are likely caused by multipath effects and other noise. For comparison with seismic processes, we calculated the cumulative moment of earthquakes with M j ≧ 2.5 within predefined subregions (Figure S16). We converted M j to M w using the empirical relationship of Uchide and Imanishi ( 2018 ). The stacked HR-GNSS time series exhibit distinct behaviors in each phase (Fig. 4 ). At the macroscopic level, both the stacked HR-GNSS time series and the cumulative seismic moment begin to increase around the onset of Phase 1 (Fig. 4 a). Although a precise comparison of the onset times between the geodetic and seismic time series is difficult due to the high noise level, the observable trend change in the HR-GNSS time series appears to begin after the first M j ≧ 5 earthquake. In contrast, during Phase 2, a rapid trend change in the stacked HR-GNSS time series coincides with a rapid increase in the cumulative seismic moment in the western subregion (Figs. 4 b and S16). Subsequently, the transient motion in the HR-GNSS time series persists; however, its rate decreases from fast to slow, whereas the growth of the cumulative seismic moment quickly subsides, in contrast to Phase 1. 4.3. Crustal deformation during the interswarm periods The displacement rate field during interswarm periods is also important for discussing the driving mechanisms of earthquake swarms and the associated transient deformation. Therefore, we first fitted a linear function to the daily position data for each interswarm period (Table S2 ). We then computed the weighted average of the estimated slopes, using their uncertainties as weights. After that, we defined this as the long-term average interswarm displacement rate. This long-term average rate generally explains the observed time series well, except at BQ1J on Kodakara Island, where transient signals without accompanying seismicity (Section 2.2 ) or rapid subsidence are observed (Figure S17). The average horizontal rate at each station is directed southeastward with no notable spatial variation (Figure S18a). Assuming that the deformation caused by back-arc spreading in the Okinawa Trough (Nishimura et al. 2004 ) is approximately uniform across the analysis region, we calculated the weighted average velocity for all stations and all interswarm periods for each component. We then subtracted this average from the long-term average rate at each site to focus on spatial short-wavelength deformation. The resulting residual displacement rate field exhibits an incoherent pattern, which is difficult to explain using a simple dislocation source (Figure S18). 4.4 Possible driving mechanism of the transient deformation and earthquake swarms Based on the MM and the preferred models, we first discuss the possibility that transient crustal deformation is indicative of SSEs. We consider the occurrence of SSEs off the Tokara Islands unlikely because Tokara earthquake swarms do not occur at quasiperiodic intervals (Figure S1 b), and there is no clear evidence for systematic strain accumulation during the inter-swarm periods (Figure S1 8), in contrast to well-known SSEs on plate boundary faults. Therefore, to explain the MM and the preferred models, we propose a scenario in which volumetric changes drive aseismic slip and earthquake swarms (Fig. 5 a). We speculate that these volumetric changes are caused by magma or fluid supply given that the Tokara Islands are part of the volcanic front of the Ryukyu subduction zone (Fig. 1 a). However, such a supply is likely to occur at greater depths and/or with relatively small volumetric changes because the transient deformation is well explained by shear-slip fault models and the GNSS data have limited resolving power for deep sources (Figures S12–14). In our scenario, magma or fluid is supplied to a region between Kodakara and Akuseki Islands, at depths greater than those of the aseismic and seismic slip regions. Slow external stress loading and/or fluid migration from this “geodetically invisible” source induced aseismic slip in Phase 1. Simultaneously, the earthquake swarm was driven by a combination of stress loading from both aseismic slip and the deeper source. Subsequently, the geodetically invisible source appeared to a region between Takara and Kodakara Islands and triggered similar aseismic slip and an earthquake swarm, whereas continuous and modest volumetric changes in the eastern source generated aseismic faulting and earthquake swarms between Kodakara and Akuseki Islands. We speculate that the newly activated western source was located at a shallower depth and/or underwent larger volumetric changes than the eastern source because of the abrupt trend changes immediately after the onset of Phase 2 (Fig. 4 ). Given the significant trend change during Phase 2 (Fig. 4 b), we further infer that stress loading or fluid supply from the western source occurs over a timescale of a few hours and that the later part of Phase 2 represents a relaxation process similar to a postseismic afterslip, that is, aseismic slip following an earthquake. Moreover, combination of the geodetic source modeling and epicenter distribution suggests that a fault different from that involved in Phase 1 may have slipped aseismically in Phase 2. In the final stage Phase 3, the aseismic slip on the western fault terminated, whereas the volumetric change at the eastern source persisted and continued to drive aseismic slip and earthquake swarms through external stress loading and/or fluid migration. After Phase 3, the displacement rate in the GNSS time series (Figure S5) returned to the level observed before the earthquake swarm, although the number of earthquakes continues to increase gradually (Figure S4). Therefore, fluids are likely to play an important role in this scenario. In general, the frictional properties of faults vary with depth (or temperature) (Scholz 1998 ), and depth-dependent transitions in aseismic–seismic slip behavior have been widely documented on plate boundary faults (Obara and Kato 2016 ). At first glance, this seems inconsistent with our results because the inferred aseismic slip regions partially overlap earthquake hypocenters in depth (Figures A4, A6, A16, A17, A39, and A41). However, laboratory experiments show that wet granite at temperatures of 350°C–600°C exhibits rate-strengthening behavior (Blanpied et al. 1995 ), and high-temperature fluids are likely to exist beneath the Tokara Islands, as suggested by the widespread distribution of hot springs (Yokose et al. 2010 ). If the modeled faults correspond to fluid pathways, the frictional properties on the fault surface may be modified toward a rate-strengthening regime that can host aseismic slip indicative of a stress relaxation (Marone et al. 1991 ). Additionally, based on alternative crack models, we propose another possible mechanism in which magma or fluid supply directly drives aseismic deformation and coincident earthquake swarms (Fig. 5 b). In this scenario, aseismic displacements during Phases 1 to 3 are caused by volumetric changes in the vertical open cracks, whereas the earthquake swarms in these phases are generated by external stress loading from the open cracks and/or by fluid migration along associated pathways. 5 Conclusions and future perspectives We report for the first time that transient displacement episodes coincident with earthquake swarms have been repeatedly recorded at GEONET and SoftBank GNSS stations in the Tokara Islands, southern Japan. Focusing on the most recent 2025 episode, we estimate source models for the associated aseismic deformation. Based on the AIC values and regional tectonics, we identify shear-slip faulting, including both lateral and normal components, as the preferred mechanism for aseismic slips, rather than vertical open cracks. The stacked HR-GNSS time series further suggests that the temporal relationship between aseismic processes and seismicity can vary, even within a single episode. Finally, we propose two scenarios for the driving mechanisms of aseismic displacements and earthquake swarms, based on the shear-slip and vertical open crack models, respectively. In one scenario, the transient displacements are caused by aseismic slip on the faults; in the other scenario, they are caused by volumetric changes in open cracks. In both scenarios, volcanic activity likely plays an important role in generating these episodes. Although the displacements calculated from the MM and preferred models reproduce the observations more successfully than those from the alternative open crack model, it remains difficult to identify the “true” driving mechanism of the aseismic displacements and earthquake swarms using GNSS data alone. To better constrain these underlying processes, integrated studies of seismicity, crustal deformation, and crustal structure are required. Seafloor seismic and geodetic observations over long timescales are essential because of the limited land coverage in this region. Our study highlights the importance of “integration” between GEONET and private GNSS networks for hazard monitoring in the island region of the southern Japanese Archipelago. Industry–academia–government cooperation will be a key factor in achieving more effective hazard assessment and monitoring. Abbreviations AIC Akaike Information Criterion CMT Centroid Moment Tensor GEONET GNSS Earth Observation NETwork System GNSS Global Navigation Satellite System GSI Geospatial Information Authority of Japan HR-GNSS High-rate GNSS JMA Japan Meteorological Agency MM Minimum Misfit NS North–South NIED National Research Institute for Earth Science and Disaster Resilience PPP-AR Precise Point Positioning with Ambiguity Resolution RINEX Receiver Independent Exchange format SSE Slow Slip Event Declarations Ethics approval and consent to participate Not applicable Consent for publication Not applicable Availability of data and materials We used GNSS data in Receiver Independent Exchange format (RINEX) at GEONET stations (Geodetic Observation Center, GSI 2025). RINEX files of SoftBank stations were provided by SoftBank Corp. and ALES Corp. under the framework of the “Consortium to Utilize the SoftBank Original Reference Sites for Earth and Space Science.” We also utilized the earthquake and CMT catalogs provided by JMA (https://www.data.jma.go.jp/eqev/data/bulletin/hypo.html) and the NIED (https://www.fnet.bosai.go.jp/event/dreger.php?LANG=en), respectively. We carried out STL and STEL decompositions using the PySTEL module (Okada and Ueda 2024). Figures are drawn by Python with the Generic Mapping Tools (Wessel et al. 2019) and Matplotlib. Competing interests The authors declare that they have no competing interests. Funding This work was supported by the Uehiro Foundation on Ethics and Education, JSPS (Japan Society for the Promotion of Science) KAKENHI Grant 25K24462, and the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, under its The Third Earthquake and Volcano Hazards Observation and Research Program. This study was also supported by the JST FOREST Program (Grant Number: JPMJFR202P, Japan). Authors' contributions Here, we use the family name of each author rather than the common way of indicating by initials to identify them uniquely. Okada, Ohta, Ohtate, and Ito conducted preliminary analyses as a basis for this study. Ohta and Ito estimated daily and HR-GNSS positions, respectively. Okada and Ohzono analyzed GNSS position data. All authors discussed scenarios of the 2025 episode. Okada drafted the first version of the manuscript, and all authors revised it. Acknowledgements Discussions with Drs. Yuji Itoh, Yo Fukushima, Takuya Nishimura, and Hanaya Okuda were fruitful. Authors' information Affiliations International Research Institute of Disaster Science, Tohoku University, 468-1 Aoba, Aramaki, Aoba-ku, Sendai 980-8572, Japan. Yutaro Okada Research Center for Prediction of Earthquakes and Volcanic Eruptions, Graduate School of Science, Tohoku University, 6-6 Aza-Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan. Advanced Institute for Marine Ecosystem Change (WPI-AIMEC), Tohoku University, 6-3 Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan. Division for the Establishment of Frontier Sciences of Organization for Advanced Studies, Tohoku University, 2-1-1 Katahira, Aoba-ku, Sendai 980-8577, Japan. International Research Institute of Disaster Science, Tohoku University, 468-1 Aoba, Aramaki, Aoba-ku, Sendai 980-8572, Japan. Core Research Cluster of Disaster Science, Tohoku University, 468-1 Aoba, Aramaki, Aoba-ku, Sendai, 980-8572, Japan Yusaku Ohta Research Center for Prediction of Earthquakes and Volcanic Eruptions, Graduate School of Science, Tohoku University, 6-6 Aza-Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan. Miku Ohtate Research Center for Prediction of Earthquakes and Volcanic Eruptions, Graduate School of Science, Tohoku University, 6-6 Aza-Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan. Yoshiaki Ito Institute of Seismology and Volcanology, Faulty of Science, Hokkaido Univer- sity, N10W8 Kita-ku, Sapporo 060-0810, Japan. Mako Ohzono Graduate School of Science and Engineering, Kagoshima University, 1-21-40 Korimoto, Kagoshima, 890-0065, Japan Hiroshi Yakiwara Graduate School of Science and Engineering, Kagoshima University, 1-21-40 Korimoto, Kagoshima, 890-0065, Japan Shigeru Nakao Endnotes We do not have any endnotes. References Altamimi Z, Rebischung P, Collilieux X et al (2023) ITRF2020: an augmented reference frame refining the modeling of nonlinear station motions. J Geod 97:47. https://doi.org/10.1007/s00190-023-01738-w Ando M, Ikuta R, Tu Y et al (2015) The Apr 2013 earthquake swarm and dyke intrusion in the Okinawa trough. Paper presented at the Japan Geoscience Union Meeting 2015, Makuhari Messe, Chiba, 24–28 May 2015 Aoki Y, Segall P, Kato T et al (1999) Imaging magma transport during the 1997 seismic swarm off the izu peninsula, japan. Science 286:927–930. https://doi.org/10.1126/science.286.5441.927 Arai R, Kodaira S, Takahashi T et al (2018) Seismic evidence for arc segmentation, active magmatic intrusions and syn-rift fault system in the northern Ryukyu volcanic arc. Earth Planet Space 70:61. https://doi.org/10.1186/s40623-018-0830-8 Bertiger W, Bar-Sever Y, Dorsey A et al (2020) GipsyX/RTGx, a new tool set for space geodetic operations and research. Adv Space Res 66:469–489. https://doi.org/10.1016/j.asr.2020.04.015 Blanpied MLB, Lockner DA, Byerlee JD (1995) Frictional slip of granite at hydrothermal conditions. J Geophys Res 100 :13,045 – 13,064. https://doi.org/10.1029/95JB00862 Cappa F, Rutqvist J, Yamamoto K (2009) Modeling crustal deformation and rupture processes related to upwelling of deep CO2-rich fluids during the 1965–1967 Matsushiro earthquake swarm in Japan. J Geophys Res 114:B10304. https://doi.org/10.1029/2009jb006398 Churchill RM, Werner MJ, Biggs J, Fagereng Å (2022) Afterslip Moment Scaling and Variability From a Global Compilation of Estimates. J Geophys Res Solid Earth 127. https://doi.org/10.1029/2021JB023897 . e2021JB023897 Cleveland RB, Cleveland WS, McRae JE, Terpenning I (1990) STL: A Seasonal-Trend Decomposition Procedure Based on Loess. J Off Stat 6:3–33 Delahaye EJ, Townend J, Reyners ME, Rogers G (2009) Microseismicity but no tremor accompanying slow slip in the Hikurangi subduction zone, New Zealand. Earth Planet Sci Lett 277:21–28. https://doi.org/10.1016/j.epsl.2008.09.038 Ellis AP, Johanson IA, Poland MP (2024) Deformation of Mauna Loa before, during, and after its 2022 eruption. Bull Volcanol 87:8. https://doi.org/10.1007/s00445-024-01788-8 Fasola SL, Jackson NM, Williams CA (2023) Deep short-term slow slip and tremor in the Manawatu region, New Zealand. Geophys Res Lett 50. https://doi.org/10.1029/2023gl105428 . e2023GL105428 Fukuyama E, Ishida M, Dreger DS, Kawai H (1998) Automated seismic moment tensor determination by using on-line broadband seismic waveforms. Jishin 51:149–156. https://doi.org/10.4294/zisin1948.51.1_149 Geng J, Chen X, Pan Y et al (2019) PRIDE PPP-AR: an open-source software for GPS PPP ambiguity resolution. GPS Solut 23:91. https://doi.org/10.1007/s10291-019-0888-1 Geospatial Information Authority of Japan (2025) Maintenance list at GEONET stations. http://terras.gsi.go.jp/denshi_hosyu.php . Accessed 18 August 2025 Geodetic Observation Center, Geospatial Information Authority of Japan (2025) GNSS data. https://doi.org/10.57499/GSI_GNSS_2025_001 . Accessed 3 December 2025 Heki K, Miyazaki S, Tsuji H (1997) Silent fault slip following an interplate thrust earthquake at the Japan Trench. Nature 386:595–598. https://doi.org/10.1038/386595a0 Himematsu Y, Sigmundsson F, Furuya M (2019) Icecap and subglacial crustal deformation inferred from SAR pixel tracking: The 2014 dike intrusion episode in the bárðarbunga volcanic system, Iceland. J Geophys Res [Solid Earth] 124:9940–9955. https://doi.org/10.1029/2019jb017652 Hirose H, Hirahara K, Kimata F et al (1999) A slow thrust slip event following the two 1996 Hyuganada Earthquakes beneath the Bungo Channel, southwest Japan. Geophys Res Lett 26:3237–3240. https://doi.org/10.1029/1999gl010999 Isken MP, Karstens J, Nomikou P et al (2025) Volcanic crisis reveals coupled magma system at Santorini and Kolumbo. Nature 645:939–945. https://doi.org/10.1038/s41586-025-09525-7 Itoh Y, Socquet A, Radiguet M (2025) Slip-tremor interaction at the very beginning of Episodic Tremor and slip in Cascadia. AGU Adv 6. https://doi.org/10.1029/2024av001425 . e2024AV001425 Japan Coast Guard (2025) The first postseismic survey of submarine topography has been conducted in off the Tokara Islands: the findings expected to serve as fundamental data for elucidating the causes of seismic activity. https://www.kaiho.mlit.go.jp/info/kouhou/post-1240.html . Accessed 3 December 2025 Kano M, Ikeuchi A, Nishimura T et al (2021) Potential of megathrust earthquakes along the southern Ryukyu Trench inferred from GNSS data. Earth Planet Space 73:1–11. https://doi.org/10.1186/s40623-021-01531-z Kawai T, Yukutake Y, Doke R, Honda R (2024) Contribution of aseismic slips to earthquake swarms at the Hakone volcano. Earth Planet Space 76:152. https://doi.org/10.1186/s40623-024-02098-1 Koge H, Sato T, Arimoto J et al (2025) Inception of ridge-ridge-ridge triple junctions: Morphostructural analysis and dynamics in the early back-arc extension of the northern Okinawa Trough. Geology 53:269–273. https://doi.org/10.1130/g52640.1 Marone CJ, Scholtz CH, Bilham R (1991) On the mechanics of earthquake afterslip. J Geophys Res 96:8441–8452. https://doi.org/10.1029/91jb00275 Matsu’ura M, Hasegawa Y (1987) A maximum likelihood approach to nonlinear inversion under constraints. Phys Earth Planet Inter 47:179–187. https://doi.org/10.1016/0031-9201(87)90076-8 Minami H, Ohara Y, Tendo H (2021) Volcanic and tectonic features of Shirahama Bank in the northern Ryukyu Arc: Implications for cross-arc volcanism controlled by arc-parallel extension. Mar Geol 441:106623. https://doi.org/10.1016/j.margeo.2021.106623 Nakamura M, Kinjo A (2018) Activated seismicity by strain rate change in the Yaeyama region, south Ryukyu. Earth Planet Space 70:154. https://doi.org/10.1186/s40623-018-0929-y Nishikawa H, Kuwatani T, Tada N, Kayama Watanabe H (2023) Simulated distributions of pumice rafts in Japan following eruptions at volcanic islands and submarine volcanoes. Prog Earth Planet Sci 10:21. https://doi.org/10.1186/s40645-023-00552-4 Nishimura S, Hashimoto M, Ando M (2004) A rigid block rotation model for the GPS derived velocity field along the Ryukyu arc. Phys Earth Planet Inter 142:185–203. https://doi.org/10.1016/j.pepi.2003.12.014 Nishimura T (2014) Short-term slow slip events along the Ryukyu Trench, southwestern Japan, observed by continuous GNSS. Progress Earth Planet Sci 1:1–13. https://doi.org/10.1186/s40645-014-0022-5 Nishimura T, Hiramatsu Y, Ohta Y (2023) Episodic transient deformation revealed by the analysis of multiple GNSS networks in the Noto Peninsula, central Japan. Sci Rep 13:8381. https://doi.org/10.1038/s41598-023-35459-z Nishimura T, Matsuzawa T, Obara K (2013) Detection of short-term slow slip events along the Nankai Trough, southwest Japan, using GNSS data. J Geophys Res Solid Earth 118:3112–3125. https://doi.org/10.1002/jgrb.50222 Nishimura T, Munekane H, Yarai H (2011) The 2011 off the Pacific coast of Tohoku Earthquake and its aftershocks observed by GEONET. Earth Planet Space 63:22. https://doi.org/10.5047/eps.2011.06.025 Nishimura T, Ozawa S, Murakami M et al (2001) Crustal Deformation caused by magma migration in the. Geophys Res Lett 28:3745–3748. https://doi.org/10.1029/2001GL013051 Obara K, Kato A (2016) Connecting slow earthquakes to huge earthquakes. Science 353:253–257. https://doi.org/10.1126/science.aaf1512 Ohta Y, Ohtate M Quality assessment of daily GNSS coordinate time series from the SoftBank proprietary reference network. Earth Planets Space in- Ohta Y, Ohzono M (2022) Potential for crustal deformation monitoring using a dense cell phone carrier Global Navigation Satellite System network. Earth Planet Space 74:25. https://doi.org/10.1186/s40623-022-01585-7 Ohtate M, Ohta Y, Mitsui Y (2025) Significant afterslip contribution to postseismic deformation in Sado Island following the 2024 Noto Peninsula earthquake: insights from two dense GNSS observation networks. Earth Planet Space 77:74. https://doi.org/10.1186/s40623-025-02203-y Ohzono M, Takahashi H, Ohta Y, Ohno K (2025) Transient inflation event at Kussharo caldera, eastern Hokkaido, Japan in 2021–2023 captured by multiple GNSS observation networks. Earth Planet Space 77:41. https://doi.org/10.1186/s40623-025-02173-1 Okada Y (1995) Simulated empirical law of coseismic crustal deformation. J Phys Earth 43:697–713. https://doi.org/10.4294/jpe1952.43.697 Okada Y (1992) Internal deformation due to shear and tensile faults in a half-space. Bull Seismol Soc Am 82:1018–1040. https://doi.org/10.1785/BSSA0820021018 Okada Y, Nishimura T (2025) Investigation on short-term slow slip events in the northeast Japan subduction zones using decadal GNSS data. Earth Planet Space 77:45. https://doi.org/10.1186/s40623-025-02175-z Okada Y, Nishimura T, Tabei T et al (2022) Development of a detection method for short-term slow slip events using GNSS data and its application to the Nankai subduction zone. Earth Planet Space 74:1–18. https://doi.org/10.1186/s40623-022-01576-8 Okada Y, Ueda T (2024) PySTEL: Python module for seasonal-trend-episodic decomposition technique based on LOESS (version 1.1.0). https://doi.org/10.5281/zenodo.13990750 . Zenodo Okada Y, Ueda T (2025) Development of a Seasonal Adjustment Technique STEL and its Application to Daily GNSS Data. J Geodetic Soc Japan 71:1–11. https://doi.org/10.11366/sokuchi.71.1 Ozawa S, Miyazaki S, Hatanaka Y et al (2003) Characteristic silent earthquakes in the eastern part of the Boso peninsula, Central Japan. Geophys Res Lett 30(6):1283. https://doi.org/10.1029/2002gl016665 Rubin AM, Pollard DD (1988) Dike-induced faulting in rift zones of Iceland and Afar. Geology 16:413–417. https://doi.org/10.1130/0091-7613(1988)016%3C0413:DIFIRZ%3E2.3.CO;2 Sagiya T, Miyazaki S, Tada T (2000) Continuous GPS Array and Present-day Crustal Deformation of Japan. Pure Appl Geophys 157:2303–2322. https://doi.org/10.1007/PL00022507 Scholz CH (1998) Earthquakes and friction laws. Nature 391:37–42. https://doi.org/10.1038/34097 Sun T, Wang K, Iinuma T et al (2014) Prevalence of viscoelastic relaxation after the 2011 Tohoku-oki earthquake. Nature 514:84–87. https://doi.org/10.1038/nature13778 Takagi R, Uchida N, Obara K (2019) Along-strike variation and migration of long‐term slow slip events in the western Nankai subduction zone, japan. J Geophys Res Solid Earth 124:3853–3880. https://doi.org/10.1029/2018jb016738 Tu Y, Heki K (2017) Decadal modulation of repeating slow slip event activity in the southwestern Ryukyu Arc possibly driven by rifting episodes at the Okinawa Trough. Geophys Res Lett 44:9308–9313. https://doi.org/10.1002/2017gl074455 Uchide T, Imanishi K (2018) Underestimation of microearthquake size by the magnitude scale of the Japan Meteorological Agency: Influence on earthquake statistics. J Geophys Res Solid Earth 123:606–620. https://doi.org/10.1002/2017jb014697 Urai M, Machida S (2005) Discolored seawater detection using ASTER reflectance products: A case study of Satsuma-Iwojima, Japan. Remote Sens Environ 99:95–104. https://doi.org/10.1016/j.rse.2005.04.028 Vallée M, Nocquet J-M, Battaglia J et al (2013) Intense interface seismicity triggered by a shallow slow slip event in the Central Ecuador subduction zone. J Geophys Res Solid Earth 118:2965–2981. https://doi.org/10.1002/jgrb.50216 Wessel P, Luis JF, Uieda L et al (2019) The Generic Mapping Tools Version 6. Geochem Geophys Geosys 20(11):5556–5564. https://doi.org/10.1029/2019GC008515 Yamada T, Ohta Y, Nishimura T et al (2025) Coseismic slip distribution of the 2024 Noto Peninsula earthquake deduced from dense global navigation satellite system network and interferometric synthetic aperture radar data: effect of assumed dip angle. Earth Planet Space 77:19. https://doi.org/10.1186/s40623-025-02154-4 Yokose H, Sato H, Fujimoto Y et al (2010) Mid-pleistocene submarine acidic volcanism of the Tokara Islands, Japan. J Geogr 119:46–68. https://doi.org/10.5026/jgeography.119.46 Supplementary Files additionalfile1epstokaratransientsokadaver8.pdf additionalfile2epstokaratransientsokadaver8.pdf gaver1.png Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Minor Revision 26 Feb, 2026 Reviewers agreed at journal 12 Jan, 2026 Reviewers invited by journal 12 Jan, 2026 Editor assigned by journal 24 Dec, 2025 First submitted to journal 18 Dec, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8399343","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":573283855,"identity":"81c677cf-f478-4961-882b-7607c2b82a39","order_by":0,"name":"Yutaro Okada","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0003-0725-4612","institution":"Tohoku University: Tohoku Daigaku","correspondingAuthor":true,"prefix":"","firstName":"Yutaro","middleName":"","lastName":"Okada","suffix":""},{"id":573283856,"identity":"64e0192f-8009-4aaa-abc7-56da243a94b9","order_by":1,"name":"Yusaku Ohta","email":"","orcid":"","institution":"Tohoku University: Tohoku Daigaku","correspondingAuthor":false,"prefix":"","firstName":"Yusaku","middleName":"","lastName":"Ohta","suffix":""},{"id":573283857,"identity":"7b3f75ba-2936-46cb-a27f-69dc0e08efc9","order_by":2,"name":"Miku Ohtate","email":"","orcid":"","institution":"Tohoku University: Tohoku Daigaku","correspondingAuthor":false,"prefix":"","firstName":"Miku","middleName":"","lastName":"Ohtate","suffix":""},{"id":573283858,"identity":"ee81b086-90be-447c-8200-b4a2e6246391","order_by":3,"name":"Yoshiaki Ito","email":"","orcid":"","institution":"Tohoku University: Tohoku Daigaku","correspondingAuthor":false,"prefix":"","firstName":"Yoshiaki","middleName":"","lastName":"Ito","suffix":""},{"id":573283859,"identity":"1766d964-f3fb-47d9-9eac-dd6edde01ae7","order_by":4,"name":"Mako Ohzono","email":"","orcid":"","institution":"Hokkaido University: Hokkaido Daigaku","correspondingAuthor":false,"prefix":"","firstName":"Mako","middleName":"","lastName":"Ohzono","suffix":""},{"id":573283860,"identity":"ad4b6d7a-0f42-41e7-bb1f-31363fe27874","order_by":5,"name":"Hiroshi Yakiwara","email":"","orcid":"","institution":"Kagoshima University: Kagoshima Daigaku","correspondingAuthor":false,"prefix":"","firstName":"Hiroshi","middleName":"","lastName":"Yakiwara","suffix":""},{"id":573283861,"identity":"10760efc-83b5-4422-9736-ab1301d61b1f","order_by":6,"name":"Shigeru Nakao","email":"","orcid":"","institution":"Kagoshima University: Kagoshima Daigaku","correspondingAuthor":false,"prefix":"","firstName":"Shigeru","middleName":"","lastName":"Nakao","suffix":""}],"badges":[],"createdAt":"2025-12-19 00:00:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8399343/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8399343/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":100237167,"identity":"c954c93a-1314-4845-9257-b4c5c3f69503","added_by":"auto","created_at":"2026-01-14 12:40:52","extension":"xml","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13316,"visible":true,"origin":"","legend":"","description":"","filename":"epspEPSPD2500422.xml","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/9db3cdd10ffe8b7c96a4d0f0.xml"},{"id":100237144,"identity":"518f68a0-d1e6-4683-b125-b64367357785","added_by":"auto","created_at":"2026-01-14 12:40:47","extension":"xml","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1426,"visible":true,"origin":"","legend":"","description":"","filename":"EPSPD250042211763.go.xml","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/26a3734bdf64964d167cfacf.xml"},{"id":100237102,"identity":"085eddb8-b1bf-4a71-a607-814d8fdf447f","added_by":"auto","created_at":"2026-01-14 12:40:46","extension":"xml","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":859,"visible":true,"origin":"","legend":"","description":"","filename":"EPSPD2500422Import.xml","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/dd5d2d808a22c84de1ce7f7e.xml"},{"id":100237165,"identity":"b43b5c7b-7fdd-4cae-9a4b-740a18a46889","added_by":"auto","created_at":"2026-01-14 12:40:51","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":134416,"visible":true,"origin":"","legend":"","description":"","filename":"EPSPD25004220enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/c6bf28d55a05473fe1c04403.xml"},{"id":100237104,"identity":"698512e6-4a53-4d3f-a842-795e99160999","added_by":"auto","created_at":"2026-01-14 12:40:47","extension":"pdf","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3781338,"visible":true,"origin":"","legend":"","description":"","filename":"fig1tectonicsettingtsseismicity.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/9c6d9e7ab0876bf7883c1662.pdf"},{"id":100237101,"identity":"628bc921-e1bc-427d-a036-d09e329207e8","added_by":"auto","created_at":"2026-01-14 12:40:46","extension":"pdf","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1123694,"visible":true,"origin":"","legend":"","description":"","filename":"fig2displacementfieldmultievents.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/fb42f255a9ee2df5ee58b8e7.pdf"},{"id":100237169,"identity":"956134b3-32fd-4867-8e51-81f30c30fb49","added_by":"auto","created_at":"2026-01-14 12:40:52","extension":"pdf","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1526625,"visible":true,"origin":"","legend":"","description":"","filename":"fig3invertedsourcemodelselected.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/f8cc7b698d832ee23cb08180.pdf"},{"id":100237095,"identity":"ae318457-acac-4ceb-a1db-06399cfd9bfd","added_by":"auto","created_at":"2026-01-14 12:40:43","extension":"pdf","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":269710,"visible":true,"origin":"","legend":"","description":"","filename":"fig4stackedhrtimeseriesphases12.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/41a9439a81ab7989d358d073.pdf"},{"id":100237097,"identity":"2d3a24c1-e1ea-48cd-82e6-c7ab88f9d83d","added_by":"auto","created_at":"2026-01-14 12:40:44","extension":"pdf","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":592490,"visible":true,"origin":"","legend":"","description":"","filename":"fig5scenariosshematic.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/fd49a3e627c27a5e0fcda2c9.pdf"},{"id":100237096,"identity":"074e24a7-7a13-497f-b931-ce1da7d1d25f","added_by":"auto","created_at":"2026-01-14 12:40:43","extension":"xml","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":131986,"visible":true,"origin":"","legend":"","description":"","filename":"EPSPD25004220structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/11c52c0e609277a5cd6e2b15.xml"},{"id":100237157,"identity":"5ef4c7d1-c46b-4b64-a130-92b3c89d807d","added_by":"auto","created_at":"2026-01-14 12:40:48","extension":"html","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":146108,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/25586533a775a56af5d6a872.html"},{"id":100237103,"identity":"a10cc0e4-00dc-420a-aa3b-d34f2758a9eb","added_by":"auto","created_at":"2026-01-14 12:40:46","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":713947,"visible":true,"origin":"","legend":"\u003cp\u003eSpatiotemporal activity of earthquakes and GNSS observations off the Tokara Islands. \u003cstrong\u003e(a)\u003c/strong\u003eRegional map of southern Japan. Red triangles indicate active volcanoes. \u003cstrong\u003e(b)\u003c/strong\u003eDistribution of earthquakes and GNSS stations around the Tokara Islands. Pink circles denote earthquakes with M ≧ 2.5 between June 1, 2025, and July 31, 2025 to retain catalog completeness, and their size reflects earthquake magnitude. Red triangles are active volcanoes. Colored diamonds and squares show the locations of GEONET and SoftBank GNSS stations, respectively. Colored characters denote the names of GNSS stations, and underlined black characters indicate the names of the islands on which the stations are located. \u003cstrong\u003e(c)\u003c/strong\u003e Daily GNSS position time series at selected stations. \u003cstrong\u003e(d)\u003c/strong\u003e Longitude–time plot of earthquakes shown in (b). Pink circles indicate earthquakes, and their size reflects earthquake magnitude. Black dashed lines denote the longitudes of Takara, Kodakara, and Akuseki Islands. \u003cstrong\u003e(e)\u003c/strong\u003e Magnitude–time (MT) diagram of earthquakes shown in (b).\u003c/p\u003e","description":"","filename":"fig1tectonicsettingtsseismicity.png","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/b0ae73b295c5e293955e6920.png"},{"id":100371779,"identity":"2c626ba6-88da-46e9-a85f-6df957fa1156","added_by":"auto","created_at":"2026-01-16 08:10:53","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":124925,"visible":true,"origin":"","legend":"\u003cp\u003eDisplacement field associated with earthquake swarm episodes or their subphases. Pink and cyan arrows indicate horizontal and vertical displacements, respectively. Error ellipses denote 95% confidence intervals. Green circles show earthquakes that occurred between the start of the pre-episode window and the end date of the post-episode window (Table S2), except in panel (e). \u003cstrong\u003e(a)\u003c/strong\u003e April 2021 episode. \u003cstrong\u003e(b)\u003c/strong\u003e December 2021 episode. \u003cstrong\u003e(c)\u003c/strong\u003e September 2023 episode. Displacements at BQ1J are shown in lighter colors because their uncertainties are larger than those at other stations, owing to missing data during the earthquake swarm. \u003cstrong\u003e(d)\u003c/strong\u003ePhase 1 of the June–July 2025 episode. \u003cstrong\u003e(e)\u003c/strong\u003e Phase 2 of the June–July 2025 episode. Earthquakes between July 1, 2025, and July 3, 2025, are plotted. \u003cstrong\u003e(f)\u003c/strong\u003ePhase 3 of the June–July 2025 episode.\u003c/p\u003e","description":"","filename":"fig2displacementfieldmultievents.png","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/6ae6935c48e0cfd40657fdbf.png"},{"id":100237164,"identity":"d46143ed-9f96-4cfd-bf32-d5a07488ac72","added_by":"auto","created_at":"2026-01-14 12:40:50","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":97785,"visible":true,"origin":"","legend":"\u003cp\u003eEstimated source models for each phase of the 2025 episode. Blue lines or rectangles indicate fault planes, and purple arrows show dislocation mechanisms. Green circles denote earthquakes used in the modeling, and their size reflects earthquake magnitude. Black and red arrows indicate input and calculated horizontal displacements, respectively. Error ellipses denote 95% confidence intervals. (a)–(c) Dextral-faulting model of Phases 1–3. (d)–(f) Normal-faulting model of Phases 1–3. (g)–(i) Vertical open crack model of Phases 1–3\u003c/p\u003e","description":"","filename":"fig3invertedsourcemodelselected.png","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/ad7c23a0fb97175e9973b709.png"},{"id":100237098,"identity":"9dccaa7b-d371-408c-aac5-33bb4bea29f2","added_by":"auto","created_at":"2026-01-14 12:40:44","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1110756,"visible":true,"origin":"","legend":"\u003cp\u003eStacked HR-GNSS time series with a cumulative seismic moment in a subregion (Figure S16). Grey dots indicate stacked 30-s position data, whereas red lines show the moving median. Blue lines show temporal variations in the cumulative moment of earthquakes with M\u003csub\u003ej\u003c/sub\u003e ≧ 2.5. (a) Time series around the initiation of Phase 1, with cumulative moment in the eastern subregion. A black dashed line indicates the timing of the first M\u003csub\u003ej\u003c/sub\u003e ≧5 earthquake in the 2025 episode. (b) Time series in Phase 2, with cumulative moment in the western subregion.\u003c/p\u003e","description":"","filename":"fig4stackedhrtimeseriesphases12.png","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/078dd790623896207ae7f1ab.png"},{"id":100237158,"identity":"299decb3-f125-49ec-8368-c7b52ff6c39e","added_by":"auto","created_at":"2026-01-14 12:40:48","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":616699,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic illustrations of two possible scenarios for the driving mechanisms of the 2025 Tokara earthquake swarm and associated aseismic displacements, based on \u003cstrong\u003e(a) \u003c/strong\u003epreferred shear-slip models and \u003cstrong\u003e(b)\u003c/strong\u003e alternative open crack models.\u003c/p\u003e","description":"","filename":"fig5scenariosshematic.png","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/535b186de4bcc5e22da2bd7f.png"},{"id":100383654,"identity":"c81d65c4-63ee-49c2-97ba-b7f03f707db6","added_by":"auto","created_at":"2026-01-16 10:47:49","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3302336,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/e12d0313-ff9d-493d-a6df-e5e38cb44d41.pdf"},{"id":100237162,"identity":"b1a5cfb6-98d1-4b40-9df3-9885f4213a52","added_by":"auto","created_at":"2026-01-14 12:40:50","extension":"pdf","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":2871682,"visible":true,"origin":"","legend":"","description":"","filename":"additionalfile1epstokaratransientsokadaver8.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/e2155d06f0cd619ee966337c.pdf"},{"id":100237172,"identity":"457cb217-e540-47d6-bccc-4cea6b6f1e5d","added_by":"auto","created_at":"2026-01-14 12:40:53","extension":"pdf","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":5960090,"visible":true,"origin":"","legend":"","description":"","filename":"additionalfile2epstokaratransientsokadaver8.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/676867379bb9017ad0dfae9b.pdf"},{"id":100237161,"identity":"19e558bd-934f-4545-96cd-0f9370bea898","added_by":"auto","created_at":"2026-01-14 12:40:49","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":276443,"visible":true,"origin":"","legend":"","description":"","filename":"gaver1.png","url":"https://assets-eu.researchsquare.com/files/rs-8399343/v1/0a1320eb6f53accae8333f4d.png"}],"financialInterests":"","formattedTitle":"Recurrent earthquake swarms and transient crustal deformation off the Tokara Islands, southern Japan: Insights from the 2025 swarm sequence","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eEarthquake swarms are sequences of earthquakes without a clear mainshock and are often synchronized with significant transient crustal deformation. For example, synchronizations between earthquake swarms and transient deformation have been observed around volcanoes and are interpreted to be related to magma intrusion (e.g., Aoki et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Ellis et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Isken et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) or hydrothermal fluid migration (Kawai et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Such synchronizations have also been reported in regions without major volcanoes and are interpreted to be caused by fluid migration (Cappa et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Nishimura et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Slow slip events (SSEs), which represent aseismic transient fault-slip phenomena that occur without significant external stress loading, are sometimes accompanied by active seismicity (e.g., Ozawa et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Delahaye et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Vall\u0026eacute;e et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Thus, analyses of geodetic data that can capture seismic and aseismic processes, are crucial for revealing the driving mechanisms of earthquake swarms.\u003c/p\u003e \u003cp\u003eThe Global Navigation Satellite System (GNSS) is a game changer in geodesy. For example, Japan\u0026rsquo;s nationwide GNSS network, the GNSS Earth Observation NETwork System (GEONET) operated by the Geospatial Information Authority of Japan (GSI), is composed of ~\u0026thinsp;1,300 stations and has contributed to revealing interseismic, coseismic, postseismic, and volcanic deformation since the 1990s (e.g., Heki et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Aoki et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Hirose et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Sagiya et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Nishimura et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Sun et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). More than 3,300 GNSS stations operated by SoftBank Corp. (hereafter \u0026ldquo;SoftBank\u0026rdquo;) have recently begun to be used for scientific purposes (Ohta and Ohzono \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Analyses that integrate GEONET and SoftBank stations have contributed to a significantly improved understanding of aseismic transient deformation, coseismic slip distribution, postseismic deformation, and volcanic activity (Nishimura et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Yamada et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Ohzono et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Ohtate et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, the SoftBank GNSS stations located in the southern Japanese Archipelago have not been used in published studies, and their ability to resolve geodetic phenomena remains uncertain.\u003c/p\u003e \u003cp\u003eThe Ryukyu subduction zone extends from southern Kyushu Island to Taiwan. Its crustal deformation is governed by two major tectonic processes: subduction of the Philippine Sea plate beneath the Okinawa plate along the Ryukyu Trench, and back-arc spreading between the Okinawa and Yangtze plate along the Okinawa Trough (e.g., Nishimura et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Kano et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). In the southern Okinawa Trough, seismic and geodetic observations have indicated a dike intrusion coincident with an earthquake swarm, which is accompanied by long-term temporal changes in seismic and SSE activity attributed to this intrusion (Ando et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Tu and Heki \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Nakamura and Kinjo \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). A volcanic front has developed in this subduction zone, and the Tokara Islands are part of this zone (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). Earthquake swarms, including events of M\u0026thinsp;≧\u0026thinsp;5, have repeatedly occurred off the Tokara Islands, particularly after 2021, with shorter recurrence intervals (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The most recent episode in 2025 occurred between Takara and Akuseki Islands and persisted for approximately one month (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), resulting in the evacuation of several dozen residents. Despite these frequent and societally impactful episodes, the driving mechanisms of earthquake swarms off the Tokara Islands remain unclear.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eGiven this background, we investigate crustal deformation accompanying earthquake swarms off the Tokara Islands by jointly analyzing GNSS datasets from GEONET and SoftBank stations (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb) to elucidate their driving mechanisms. We also used earthquake and centroid moment tensor (CMT) catalogs provided by the Japan Meteorological Agency (JMA) and the National Research Institute for Earth Science and Disaster Resilience (NIED), respectively, to compare crustal deformation with seismicity. In this paper, we demonstrate a clear synchronization between earthquake swarms and transient crustal deformation, estimate source models for aseismic deformation, and propose multiple scenarios for the driving mechanisms.\u003c/p\u003e"},{"header":"2 Data analysis","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 GNSS data and preprocessing of daily positions\u003c/h2\u003e \u003cp\u003eWe analyzed GNSS data from 30 GEONET and 59 SoftBank stations located in and around the Tokara Islands from November 1, 2019, to October 29, 2025 (Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). Daily positions were estimated from GPS observations using GipsyX (version 2.3), employing a Precise Point Positioning with Ambiguity Resolution (PPP-AR) strategy (Bertiger et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The resulting daily positions were aligned into the International Terrestrial Reference Frame 2020 (Altamimi et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In addition, high-rate GNSS (HR-GNSS) positions, sampled every 30 s, at BQ1I, 1243, BQ1J, and BQ1K (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb and S2) were used to examine detailed temporal correlations with seismicity (Section \u003cspan refid=\"Sec10\" class=\"InternalRef\"\u003e4.2\u003c/span\u003e). These high-rate positions were estimated by applying the kinematic PPP-AR approach implemented in PRIDE-PPPAR (version 3.1.0) (Geng et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) to multi-GNSS (GPS, GLONASS, and Galileo) observations.\u003c/p\u003e \u003cp\u003eWe removed common-mode errors (Okada and Nishimura \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), maintenance offsets at GEONET stations, and seasonal oscillations (Nishimura et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Okada and Ueda \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) from the daily positions as preprocessing (Text S1, Figures \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e and S4, and Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The common-mode errors were calculated using data from stations away from the swarms (e.g., Fasola et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), which were denoised based on step master files (GSI, 2025; Ohta and Ohtate, in-press) and statistical and empirical techniques (Cleveland et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e1990\u003c/span\u003e; Okada \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). For station 1243, we did not estimate the seasonal component because of the short observation period.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Comparison of time series and displacement fields with seismicity\u003c/h2\u003e \u003cp\u003eThe daily position time series which is detrended using linear velocities estimated from December 2024 to May 2025, show clear synchronicity between earthquake swarms and transient crustal deformation in multiple episodes (Figure S5). Focusing on the north\u0026ndash;south (NS) components at stations on Takara Island (Figures S5 and S6), the 2025 episode exhibits particularly pronounced signals (~\u0026thinsp;40 mm), which are unique among episodes since November 2019. In addition, the inter-episode velocities at multiple stations exhibit slight temporal variations. Transient motions without any clear accompanying increase in seismicity also appear in the time series, most notably at BQ1J on Kodakara Island. As our primary objective is to investigate crustal deformation coincident with earthquake swarms, we do not examine these signals in detail.\u003c/p\u003e \u003cp\u003eNext, we examine the temporal changes during the 2025 episode by dividing it into three phases (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec-e). In Phase 1 (June 20, 2025\u0026ndash;June 30, 2025), a swarm episode was initiated between Takara and Akuseki Islands. Coincident with the increase in seismicity, the GNSS stations on Kodakara and Akuseki Islands show a gradual but large displacement. In contrast, the NS component of BQ1I on Takara Island does not exhibit any noticeable signal during this phase. Subsequently, in Phase 2 (July 1, 2025\u0026ndash;July 3, 2025), the seismic activity between Takara and Akuseki Islands migrated\u0026thinsp;~\u0026thinsp;10 km eastward, and a secondary swarm began west of Takara Island (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). Simultaneously, the GNSS time series on Takara and Kodakara Islands show rapid trend changes that are markedly different from the deformation in Phase 1, whereas BQ1K on Akuseki Island continues to show a persistent signal with a slow rate. Taken together, the GNSS data suggest that at least two sources drive deformation at different rates. In Phase 3 (July 4, 2025\u0026ndash;July 20, 2025), earthquakes west of Takara Island ceased, whereas activity between Kodakara and Akuseki Islands persisted. The GNSS time series in Phase 3 resemble those in Phase 1: slow deformation at stations on Kodakara and Akuseki Islands and no marked signal at BQ1I on Takara Island. After Phase 3, the transient crustal deformation ceased, and no reactivation was observed until October 29, 2025. However, M\u0026thinsp;~\u0026thinsp;3 earthquakes continued to occur around the Tokara Islands for at least three months after their onset (Figure S5).\u003c/p\u003e \u003cp\u003eFinally, we examine the displacement fields for each phase of the 2025 episode and for the April 2021, December 2021, and September 2023 episodes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Using the detrended dataset, we calculated the difference between the average daily positions over the two time windows defined for each episode (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e) and treated these differences as displacements. In previous episodes, the GNSS stations on Kodakara and Akuseki Islands are displaced northeastward and southwestward, respectively, although the relative magnitudes of the horizontal displacement vary among episodes. Furthermore, the horizontal displacement amplitudes on Takara Island are smaller than those on Kodakara Island, whereas the displacement azimuths are similar. No systematic patterns were observed in the vertical displacement field. The displacement fields in Phases 1 and 3 of the 2025 episode share these characteristics, whereas those in Phase 2 shows a previously unseen pattern: stations on Takara and Kodakara Islands are displaced southward and northward, respectively. Additionally, the displacement of BQ1K on Akuseki Island continues to trend southwesterly, similar to the Phases 1 and 3.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3 Source model of the 2025 transient deformation episode","content":"\u003cp\u003eAs described above, crustal deformation during the 2025 earthquake swarm exhibits distinctive characteristics, particularly in Phase 2. Therefore, focusing on the 2025 episode, we estimate the source models of aseismic deformation in each phase and discuss the results in the following sections.\u003c/p\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Inversion strategy\u003c/h2\u003e \u003cp\u003eTo isolate the aseismic processes, we first calculated the coseismic displacements predicted from F-net CMT solutions (Fukuyama et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1998\u003c/span\u003e) of earthquakes off the Tokara Islands using a point source dislocation model in a homogeneous elastic half-space (Okada \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1992\u003c/span\u003e). We then stacked the displacements due to all listed earthquakes within the time window for each phase (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e) and subtracted these from the observed displacements (Figure S7). The observational errors were computed from the standard deviations of the daily positions over the two time windows used to define each displacement (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe employed a nonlinear least-squares inversion (Matsu\u0026rsquo;ura and Hasegawa \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e1987\u003c/span\u003e) to estimate rectangular dislocation source models (Okada \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1992\u003c/span\u003e), while simultaneously adopting a grid-search strategy because the inversion is highly sensitive to prior constraints. To reduce bias due to the assumed source location, we tested multiple combinations of longitude and latitude (cf. Nishimura \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Takagi et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) (Tables S4 and S5). In addition, we determined the optimal fault-top depth below sea level using a grid search, rather than treating it as an inversion parameter, to avoid unrealistic models. We further imposed a strict prior constraint on the strike so that the resulting models remained consistent with regional tectonic structures (Table S6; Minami et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Koge et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Moreover, to limit the number of free parameters, we assumed that the dislocation mechanism belongs to one of four types: normal, dextral, or sinistral faulting, or a vertical open crack. Finally, we estimated three-dimensional translation together with the rectangular faults to account for possible remaining common-mode errors (cf. Nishimura et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). For each assumed dislocation mechanism, we systematically selected a model that minimized the chi-square statistic among the solutions obtained for different prior source locations. The preferred dislocation mechanism was selected based on the Akaike Information Criterion (AIC) and its consistency with regional tectonics (Section \u003cspan refid=\"Sec7\" class=\"InternalRef\"\u003e3.2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eUsing this strategy, we estimated a source model and three-dimensional translations for Phases 1 and 3 (Figure S8 and Tables S4\u0026ndash;S6). In contrast, the GNSS and seismicity patterns in Phase 2 suggest the presence of multiple sources. Therefore, we fixed the location, geometry, and dislocation mechanism of one source to be identical to those of the Phase 3 model because the earthquake distribution between Kodakara and Akuseki Islands in Phase 3 closely resembles that in Phase 2 (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef), and estimated only the dislocation magnitude under the same prior constraint as in Phases 1 and 3 (Table S6). We then introduced an additional source and three-dimensional translations and estimated their parameters simultaneously using the same inversion strategy as in Phases 1 and 3.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Estimated source models\u003c/h2\u003e \u003cp\u003eThe source models for all trial combinations in all phases are shown in Figures A1\u0026ndash;A42, and the selected models are summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. In Phases 1 and 3, the northwest-dipping faults with dextral slip yield the minimum AIC values (Figure S9), although the pattern of horizontal displacements differs between the phases (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). We also consider northwest-dipping faults with normal slip, which give the fourth- and second-smallest AIC values in Phases 1 and 3, respectively (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef). We excluded the sinistral-slip models in Phase 1 from further discussion because its negative slip (Figure A2) and lateral faulting with an extremely low dip angle (Figure A3) is physically unreasonable in terms of the direction of the P- and T-axes, even though its AIC values is smaller than that in the normal-slip case. The calculated horizontal displacements from the dextral- and normal-fault models reproduce the observed horizontal displacements with comparable residuals (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). In contrast, the vertical displacement at BQ1J on Kodakara Island calculated from the normal-faulting model is significantly larger than the observed displacements (Figure A4b), whereas the vertical displacements from the dextral-faulting model are consistent with the observation (Figure A6b). In Phase 3, however, the fits to the observations obtained from the dextral- and normal-fault models are similar in terms of both visual inspection and AIC values (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec, \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef, and S9).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn Phase 2, the AIC values remain small when sinistral slip is selected for the western fault, irrespective of the assumed mechanism for the eastern fault between Kodakara and Akuseki Islands (Figure S10). Note that a dextral fault striking 315\u0026deg; in combination with an eastern normal fault also yields a small AIC; however, we interpret this case as sinistral faulting because the estimated slip magnitude is negative (Figure A28). Therefore, based on the data misfit, we adopted the sinistral-slip model striking 315\u0026deg; as the western source model (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee).\u003c/p\u003e \u003cp\u003eHereafter, we refer to the source models with dextral faulting on the eastern fault (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea\u0026ndash;c) as the minimum-misfit (MM) model. The MM model is acceptable in light of the shallow seismic velocity structure, because lateral faulting on the back-arc side of the volcanic front has been suggested by seismic reflection images, in addition to normal faulting (Arai et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Meanwhile, the normal-faulting models (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed\u0026ndash;f) are also acceptable owing to their small AIC values. Moreover, these models are more consistent with the regional tectonics around the Tokara Islands (Nishimura et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Minami et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), the stress field inferred from earthquake focal mechanisms (Koge et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), and previous geodetic observations of other volcanic regions (Rubin and Pollard \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Himematsu et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). For these reasons, we refer to the normal-faulting models as the preferred model.\u003c/p\u003e \u003cp\u003eThe magnitude of the aseismic slip ranges from M\u003csub\u003ew\u003c/sub\u003e 6.0 to 6.4, depending on the phase and mechanism. On the other hand, the cumulative moment of earthquakes in Phases 1, 2, and 3 is 1.17\u0026times;10\u003csup\u003e18\u003c/sup\u003e, 1.81\u0026times;10\u003csup\u003e18\u003c/sup\u003e, and 1.09\u0026times;10\u003csup\u003e18\u003c/sup\u003e Nm, respectively. The ratio between the aseismic and seismic moments varies from 1.1 to 3.9 (Tables S7 and S8), which is larger than typical afterslips regardless of phase and mechanism (e.g., Churchill et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). In addition, the estimated fault planes are inconsistent with the hypocenter distribution of earthquakes in the JMA catalog. However, this comparison has uncertainty associated with hypocenter locations because the catalog is not relocated (Figures A4, A6, A16, A17, A39, and A41).\u003c/p\u003e \u003cp\u003eWe evaluate uncertainties in the fault plane extent of the MM and preferred models using a remove-one-station jackknife test (Figure S11). The results show that the fault lengths in Phases 1 and 3 and the location of the western fault in Phase 2 vary depending on which station is removed, whereas the fault widths in Phases 1 and 3 and the length of the western fault in Phase 2 remain almost unchanged. The former indicates that these parameters strongly depend on displacements observed at a limited number of stations particularly BQ1J, on Kodakara Island. In contrast, the latter suggests that these parameters are largely controlled by prior constraints (Table S6) rather than by observational data.\u003c/p\u003e \u003c/div\u003e"},{"header":"4 Discussion","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Source models for the 2025 episode assuming vertical open cracks\u003c/h2\u003e \u003cp\u003eIn addition to shear-slip models, we examine vertical open crack sources in each phase (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg\u0026ndash;i), anticipating the possibility of dike intrusion because the Tokara Islands are part of the volcanic front in the Ryukyu subduction zone (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) and earthquake swarms sometimes coincide with magma intrusions (e.g., Aoki et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Ellis et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Isken et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). The calculated displacements from open cracks with a top depth of 1 km below sea level (Table S9) partially reproduce the observed displacements in each phase; however, their AIC values are significantly larger than those of the MM and preferred models (Figures S9 and S10).\u003c/p\u003e \u003cp\u003eNo evidence indicative of magma discharge into the sea, such as newly formed submarine topography, discolored seawater (e.g., Urai and Machida \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), or pumice rafts (e.g., Nishikawa et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), had not been reported as of December 19, 2025 (Japan Coast Guard, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). To consider this information, we evaluate the data fit of the crack models as a function of their top depth using the chi-square ratio \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}_{d}\\)\u003c/span\u003e\u003c/span\u003e, defined as\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:\\begin{array}{c}{R}_{d}=\\frac{{\\chi\\:}_{d}^{2}}{{\\chi\\:}_{min}^{2}},\\:\\#\\left(1\\right)\\end{array}\\)\u003c/span\u003e \u003c/span\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}_{d}^{2}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}_{min}^{2}\\)\u003c/span\u003e\u003c/span\u003e are the chi-square statistic of the best-fitting model with a top depth of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:d\\)\u003c/span\u003e\u003c/span\u003e km in a given phase and of the overall minimum chi-squared model in that phase, respectively. For the Phase 2 source model, we varied the top depth of the western fault while fixing that of the eastern fault at 1 km. The comparison (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e2) shows that the top depth must be close to 1 km, which is comparable to the water depth around the islands (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea), to reproduce the amplitude of the observed displacements (Figures S13 and S14).\u003c/p\u003e \u003cp\u003eIn conclusion, our results do not support a scenario in which magma is discharged into the sea at a high rate, given that the GNSS time series does not show any reversal of the transient trend after Phase 3 (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec and S5) and considering other available observations. Nevertheless, magma intrusion or a fluid supply reaching shallow depths cannot be ruled out based solely on GNSS data. However, the geodetic data are more consistently explained by the shear-slip models than by the open crack models (Figures S9 and S10). Therefore, we regard the vertical open crack model as an alternative model and discuss its implications further in Section \u003cspan refid=\"Sec12\" class=\"InternalRef\"\u003e4.4\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Temporal relationship between transient deformation and seismicity clarified by HR-GNSS data\u003c/h2\u003e \u003cp\u003eInvestigating the temporal relationship between transient deformation and seismicity is essential for discussing causal linkages. Therefore, we examine this relationship using stacked HR-GNSS data (cf. Okada et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) (Text S2, Figure S15, and Table S10). We calculated a moving median with a 3-hour window for the stacked time series (Itoh et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) to suppress short-term fluctuations, which are likely caused by multipath effects and other noise. For comparison with seismic processes, we calculated the cumulative moment of earthquakes with M\u003csub\u003ej\u003c/sub\u003e ≧ 2.5 within predefined subregions (Figure S16). We converted M\u003csub\u003ej\u003c/sub\u003e to M\u003csub\u003ew\u003c/sub\u003e using the empirical relationship of Uchide and Imanishi (\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe stacked HR-GNSS time series exhibit distinct behaviors in each phase (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). At the macroscopic level, both the stacked HR-GNSS time series and the cumulative seismic moment begin to increase around the onset of Phase 1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). Although a precise comparison of the onset times between the geodetic and seismic time series is difficult due to the high noise level, the observable trend change in the HR-GNSS time series appears to begin after the first M\u003csub\u003ej\u003c/sub\u003e ≧ 5 earthquake. In contrast, during Phase 2, a rapid trend change in the stacked HR-GNSS time series coincides with a rapid increase in the cumulative seismic moment in the western subregion (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb and S16). Subsequently, the transient motion in the HR-GNSS time series persists; however, its rate decreases from fast to slow, whereas the growth of the cumulative seismic moment quickly subsides, in contrast to Phase 1.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Crustal deformation during the interswarm periods\u003c/h2\u003e \u003cp\u003eThe displacement rate field during interswarm periods is also important for discussing the driving mechanisms of earthquake swarms and the associated transient deformation. Therefore, we first fitted a linear function to the daily position data for each interswarm period (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). We then computed the weighted average of the estimated slopes, using their uncertainties as weights. After that, we defined this as the long-term average interswarm displacement rate. This long-term average rate generally explains the observed time series well, except at BQ1J on Kodakara Island, where transient signals without accompanying seismicity (Section \u003cspan refid=\"Sec4\" class=\"InternalRef\"\u003e2.2\u003c/span\u003e) or rapid subsidence are observed (Figure S17). The average horizontal rate at each station is directed southeastward with no notable spatial variation (Figure S18a). Assuming that the deformation caused by back-arc spreading in the Okinawa Trough (Nishimura et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) is approximately uniform across the analysis region, we calculated the weighted average velocity for all stations and all interswarm periods for each component. We then subtracted this average from the long-term average rate at each site to focus on spatial short-wavelength deformation. The resulting residual displacement rate field exhibits an incoherent pattern, which is difficult to explain using a simple dislocation source (Figure S18).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Possible driving mechanism of the transient deformation and earthquake swarms\u003c/h2\u003e \u003cp\u003eBased on the MM and the preferred models, we first discuss the possibility that transient crustal deformation is indicative of SSEs. We consider the occurrence of SSEs off the Tokara Islands unlikely because Tokara earthquake swarms do not occur at quasiperiodic intervals (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb), and there is no clear evidence for systematic strain accumulation during the inter-swarm periods (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e8), in contrast to well-known SSEs on plate boundary faults.\u003c/p\u003e \u003cp\u003eTherefore, to explain the MM and the preferred models, we propose a scenario in which volumetric changes drive aseismic slip and earthquake swarms (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). We speculate that these volumetric changes are caused by magma or fluid supply given that the Tokara Islands are part of the volcanic front of the Ryukyu subduction zone (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). However, such a supply is likely to occur at greater depths and/or with relatively small volumetric changes because the transient deformation is well explained by shear-slip fault models and the GNSS data have limited resolving power for deep sources (Figures S12\u0026ndash;14). In our scenario, magma or fluid is supplied to a region between Kodakara and Akuseki Islands, at depths greater than those of the aseismic and seismic slip regions. Slow external stress loading and/or fluid migration from this \u0026ldquo;geodetically invisible\u0026rdquo; source induced aseismic slip in Phase 1. Simultaneously, the earthquake swarm was driven by a combination of stress loading from both aseismic slip and the deeper source. Subsequently, the geodetically invisible source appeared to a region between Takara and Kodakara Islands and triggered similar aseismic slip and an earthquake swarm, whereas continuous and modest volumetric changes in the eastern source generated aseismic faulting and earthquake swarms between Kodakara and Akuseki Islands. We speculate that the newly activated western source was located at a shallower depth and/or underwent larger volumetric changes than the eastern source because of the abrupt trend changes immediately after the onset of Phase 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Given the significant trend change during Phase 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb), we further infer that stress loading or fluid supply from the western source occurs over a timescale of a few hours and that the later part of Phase 2 represents a relaxation process similar to a postseismic afterslip, that is, aseismic slip following an earthquake. Moreover, combination of the geodetic source modeling and epicenter distribution suggests that a fault different from that involved in Phase 1 may have slipped aseismically in Phase 2. In the final stage Phase 3, the aseismic slip on the western fault terminated, whereas the volumetric change at the eastern source persisted and continued to drive aseismic slip and earthquake swarms through external stress loading and/or fluid migration. After Phase 3, the displacement rate in the GNSS time series (Figure S5) returned to the level observed before the earthquake swarm, although the number of earthquakes continues to increase gradually (Figure S4).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTherefore, fluids are likely to play an important role in this scenario. In general, the frictional properties of faults vary with depth (or temperature) (Scholz \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1998\u003c/span\u003e), and depth-dependent transitions in aseismic\u0026ndash;seismic slip behavior have been widely documented on plate boundary faults (Obara and Kato \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). At first glance, this seems inconsistent with our results because the inferred aseismic slip regions partially overlap earthquake hypocenters in depth (Figures A4, A6, A16, A17, A39, and A41). However, laboratory experiments show that wet granite at temperatures of 350\u0026deg;C\u0026ndash;600\u0026deg;C exhibits rate-strengthening behavior (Blanpied et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1995\u003c/span\u003e), and high-temperature fluids are likely to exist beneath the Tokara Islands, as suggested by the widespread distribution of hot springs (Yokose et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). If the modeled faults correspond to fluid pathways, the frictional properties on the fault surface may be modified toward a rate-strengthening regime that can host aseismic slip indicative of a stress relaxation (Marone et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1991\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAdditionally, based on alternative crack models, we propose another possible mechanism in which magma or fluid supply directly drives aseismic deformation and coincident earthquake swarms (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). In this scenario, aseismic displacements during Phases 1 to 3 are caused by volumetric changes in the vertical open cracks, whereas the earthquake swarms in these phases are generated by external stress loading from the open cracks and/or by fluid migration along associated pathways.\u003c/p\u003e \u003c/div\u003e"},{"header":"5 Conclusions and future perspectives","content":"\u003cp\u003eWe report for the first time that transient displacement episodes coincident with earthquake swarms have been repeatedly recorded at GEONET and SoftBank GNSS stations in the Tokara Islands, southern Japan. Focusing on the most recent 2025 episode, we estimate source models for the associated aseismic deformation. Based on the AIC values and regional tectonics, we identify shear-slip faulting, including both lateral and normal components, as the preferred mechanism for aseismic slips, rather than vertical open cracks. The stacked HR-GNSS time series further suggests that the temporal relationship between aseismic processes and seismicity can vary, even within a single episode. Finally, we propose two scenarios for the driving mechanisms of aseismic displacements and earthquake swarms, based on the shear-slip and vertical open crack models, respectively. In one scenario, the transient displacements are caused by aseismic slip on the faults; in the other scenario, they are caused by volumetric changes in open cracks. In both scenarios, volcanic activity likely plays an important role in generating these episodes.\u003c/p\u003e \u003cp\u003eAlthough the displacements calculated from the MM and preferred models reproduce the observations more successfully than those from the alternative open crack model, it remains difficult to identify the \u0026ldquo;true\u0026rdquo; driving mechanism of the aseismic displacements and earthquake swarms using GNSS data alone. To better constrain these underlying processes, integrated studies of seismicity, crustal deformation, and crustal structure are required. Seafloor seismic and geodetic observations over long timescales are essential because of the limited land coverage in this region.\u003c/p\u003e \u003cp\u003eOur study highlights the importance of \u0026ldquo;integration\u0026rdquo; between GEONET and private GNSS networks for hazard monitoring in the island region of the southern Japanese Archipelago. Industry\u0026ndash;academia\u0026ndash;government cooperation will be a key factor in achieving more effective hazard assessment and monitoring.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAIC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAkaike Information Criterion\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCMT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eCentroid Moment Tensor\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGEONET\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGNSS Earth Observation NETwork System\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGNSS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGlobal Navigation Satellite System\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGSI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGeospatial Information Authority of Japan\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eHR-GNSS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eHigh-rate GNSS\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eJMA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eJapan Meteorological Agency\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMM\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMinimum Misfit\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eNorth\u0026ndash;South\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNIED\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eNational Research Institute for Earth Science and Disaster Resilience\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePPP-AR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePrecise Point Positioning with Ambiguity Resolution\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRINEX\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eReceiver Independent Exchange format\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSSE\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSlow Slip Event\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe used GNSS data in Receiver Independent Exchange format (RINEX) at GEONET stations (Geodetic Observation Center, GSI 2025). RINEX files of SoftBank stations were provided by SoftBank Corp. and ALES Corp. under the framework of the \u0026ldquo;Consortium to Utilize the SoftBank Original Reference Sites for Earth and Space Science.\u0026rdquo; We also utilized the earthquake and CMT catalogs provided by JMA (https://www.data.jma.go.jp/eqev/data/bulletin/hypo.html) and the NIED (https://www.fnet.bosai.go.jp/event/dreger.php?LANG=en), respectively. We carried out STL and STEL decompositions using the PySTEL module (Okada and Ueda 2024). Figures are drawn by Python with the Generic Mapping Tools (Wessel et al. 2019) and Matplotlib.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the Uehiro Foundation on Ethics and Education, JSPS (Japan Society for the Promotion of Science) KAKENHI Grant 25K24462, and the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, under its The Third Earthquake and Volcano Hazards Observation and Research Program.\u0026nbsp;This study was also supported by the JST FOREST Program (Grant Number: JPMJFR202P, Japan).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHere, we use the family name of each author rather than the common way of indicating by initials to identify them uniquely.\u003c/p\u003e\n\u003cp\u003eOkada, Ohta, Ohtate, and Ito conducted preliminary analyses as a basis for this study. Ohta and Ito estimated daily and HR-GNSS positions, respectively. Okada and Ohzono analyzed GNSS position data. All authors discussed scenarios of the 2025 episode. Okada drafted the first version of the manuscript, and all authors revised it.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDiscussions with Drs. Yuji Itoh, Yo Fukushima, Takuya Nishimura, and Hanaya Okuda were fruitful.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAffiliations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInternational Research Institute of Disaster Science, Tohoku University, 468-1 Aoba, Aramaki, Aoba-ku, Sendai 980-8572, Japan.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eYutaro Okada\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eResearch Center for Prediction of Earthquakes and Volcanic Eruptions, Graduate School of Science, Tohoku University, 6-6 Aza-Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan.\u003c/p\u003e\n\u003cp\u003eAdvanced Institute for Marine Ecosystem Change (WPI-AIMEC), Tohoku University, 6-3 Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan.\u003c/p\u003e\n\u003cp\u003eDivision for the Establishment of Frontier Sciences of Organization for Advanced Studies, Tohoku University, 2-1-1 Katahira, Aoba-ku, Sendai 980-8577, Japan.\u003c/p\u003e\n\u003cp\u003eInternational Research Institute of Disaster Science, Tohoku University, 468-1 Aoba, Aramaki, Aoba-ku, Sendai 980-8572, Japan.\u003c/p\u003e\n\u003cp\u003eCore Research Cluster of Disaster Science, Tohoku University, 468-1 Aoba, Aramaki, Aoba-ku, Sendai, 980-8572, Japan\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eYusaku Ohta\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eResearch Center for Prediction of Earthquakes and Volcanic Eruptions, Graduate School of Science, Tohoku University, 6-6 Aza-Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eMiku Ohtate\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eResearch Center for Prediction of Earthquakes and Volcanic Eruptions, Graduate School of Science, Tohoku University, 6-6 Aza-Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eYoshiaki Ito\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eInstitute of Seismology and Volcanology, Faulty of Science, Hokkaido Univer- sity, N10W8 Kita-ku, Sapporo 060-0810, Japan.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eMako Ohzono\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eGraduate School of Science and Engineering, Kagoshima University, 1-21-40 Korimoto, Kagoshima, 890-0065, Japan\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eHiroshi Yakiwara\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eGraduate School of Science and Engineering, Kagoshima University, 1-21-40 Korimoto, Kagoshima, 890-0065, Japan\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eShigeru Nakao\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEndnotes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe do not have any endnotes.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAltamimi Z, Rebischung P, Collilieux X et al (2023) ITRF2020: an augmented reference frame refining the modeling of nonlinear station motions. J Geod 97:47. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00190-023-01738-w\u003c/span\u003e\u003cspan address=\"10.1007/s00190-023-01738-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndo M, Ikuta R, Tu Y et al (2015) The Apr 2013 earthquake swarm and dyke intrusion in the Okinawa trough. Paper presented at the Japan Geoscience Union Meeting 2015, Makuhari Messe, Chiba, 24\u0026ndash;28 May 2015\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAoki Y, Segall P, Kato T et al (1999) Imaging magma transport during the 1997 seismic swarm off the izu peninsula, japan. Science 286:927\u0026ndash;930. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.286.5441.927\u003c/span\u003e\u003cspan address=\"10.1126/science.286.5441.927\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArai R, Kodaira S, Takahashi T et al (2018) Seismic evidence for arc segmentation, active magmatic intrusions and syn-rift fault system in the northern Ryukyu volcanic arc. Earth Planet Space 70:61. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-018-0830-8\u003c/span\u003e\u003cspan address=\"10.1186/s40623-018-0830-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBertiger W, Bar-Sever Y, Dorsey A et al (2020) GipsyX/RTGx, a new tool set for space geodetic operations and research. Adv Space Res 66:469\u0026ndash;489. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.asr.2020.04.015\u003c/span\u003e\u003cspan address=\"10.1016/j.asr.2020.04.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlanpied MLB, Lockner DA, Byerlee JD (1995) Frictional slip of granite at hydrothermal conditions. J Geophys Res 100 :13,045\u0026thinsp;\u0026ndash;\u0026thinsp;13,064. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/95JB00862\u003c/span\u003e\u003cspan address=\"10.1029/95JB00862\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCappa F, Rutqvist J, Yamamoto K (2009) Modeling crustal deformation and rupture processes related to upwelling of deep CO2-rich fluids during the 1965\u0026ndash;1967 Matsushiro earthquake swarm in Japan. J Geophys Res 114:B10304. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2009jb006398\u003c/span\u003e\u003cspan address=\"10.1029/2009jb006398\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChurchill RM, Werner MJ, Biggs J, Fagereng \u0026Aring; (2022) Afterslip Moment Scaling and Variability From a Global Compilation of Estimates. J Geophys Res Solid Earth 127. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2021JB023897\u003c/span\u003e\u003cspan address=\"10.1029/2021JB023897\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. e2021JB023897\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCleveland RB, Cleveland WS, McRae JE, Terpenning I (1990) STL: A Seasonal-Trend Decomposition Procedure Based on Loess. J Off Stat 6:3\u0026ndash;33\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDelahaye EJ, Townend J, Reyners ME, Rogers G (2009) Microseismicity but no tremor accompanying slow slip in the Hikurangi subduction zone, New Zealand. Earth Planet Sci Lett 277:21\u0026ndash;28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.epsl.2008.09.038\u003c/span\u003e\u003cspan address=\"10.1016/j.epsl.2008.09.038\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEllis AP, Johanson IA, Poland MP (2024) Deformation of Mauna Loa before, during, and after its 2022 eruption. Bull Volcanol 87:8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00445-024-01788-8\u003c/span\u003e\u003cspan address=\"10.1007/s00445-024-01788-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFasola SL, Jackson NM, Williams CA (2023) Deep short-term slow slip and tremor in the Manawatu region, New Zealand. Geophys Res Lett 50. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2023gl105428\u003c/span\u003e\u003cspan address=\"10.1029/2023gl105428\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. e2023GL105428\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFukuyama E, Ishida M, Dreger DS, Kawai H (1998) Automated seismic moment tensor determination by using on-line broadband seismic waveforms. Jishin 51:149\u0026ndash;156. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4294/zisin1948.51.1_149\u003c/span\u003e\u003cspan address=\"10.4294/zisin1948.51.1_149\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeng J, Chen X, Pan Y et al (2019) PRIDE PPP-AR: an open-source software for GPS PPP ambiguity resolution. GPS Solut 23:91. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10291-019-0888-1\u003c/span\u003e\u003cspan address=\"10.1007/s10291-019-0888-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeospatial Information Authority of Japan (2025) Maintenance list at GEONET stations. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://terras.gsi.go.jp/denshi_hosyu.php\u003c/span\u003e\u003cspan address=\"http://terras.gsi.go.jp/denshi_hosyu.php\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 18 August 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeodetic Observation Center, Geospatial Information Authority of Japan (2025) GNSS data. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.57499/GSI_GNSS_2025_001\u003c/span\u003e\u003cspan address=\"10.57499/GSI_GNSS_2025_001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 3 December 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeki K, Miyazaki S, Tsuji H (1997) Silent fault slip following an interplate thrust earthquake at the Japan Trench. Nature 386:595\u0026ndash;598. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/386595a0\u003c/span\u003e\u003cspan address=\"10.1038/386595a0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHimematsu Y, Sigmundsson F, Furuya M (2019) Icecap and subglacial crustal deformation inferred from SAR pixel tracking: The 2014 dike intrusion episode in the b\u0026aacute;r\u0026eth;arbunga volcanic system, Iceland. J Geophys Res [Solid Earth] 124:9940\u0026ndash;9955. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2019jb017652\u003c/span\u003e\u003cspan address=\"10.1029/2019jb017652\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHirose H, Hirahara K, Kimata F et al (1999) A slow thrust slip event following the two 1996 Hyuganada Earthquakes beneath the Bungo Channel, southwest Japan. Geophys Res Lett 26:3237\u0026ndash;3240. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/1999gl010999\u003c/span\u003e\u003cspan address=\"10.1029/1999gl010999\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIsken MP, Karstens J, Nomikou P et al (2025) Volcanic crisis reveals coupled magma system at Santorini and Kolumbo. Nature 645:939\u0026ndash;945. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41586-025-09525-7\u003c/span\u003e\u003cspan address=\"10.1038/s41586-025-09525-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eItoh Y, Socquet A, Radiguet M (2025) Slip-tremor interaction at the very beginning of Episodic Tremor and slip in Cascadia. AGU Adv 6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2024av001425\u003c/span\u003e\u003cspan address=\"10.1029/2024av001425\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. e2024AV001425\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJapan Coast Guard (2025) The first postseismic survey of submarine topography has been conducted in off the Tokara Islands: the findings expected to serve as fundamental data for elucidating the causes of seismic activity. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.kaiho.mlit.go.jp/info/kouhou/post-1240.html\u003c/span\u003e\u003cspan address=\"https://www.kaiho.mlit.go.jp/info/kouhou/post-1240.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 3 December 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKano M, Ikeuchi A, Nishimura T et al (2021) Potential of megathrust earthquakes along the southern Ryukyu Trench inferred from GNSS data. Earth Planet Space 73:1\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-021-01531-z\u003c/span\u003e\u003cspan address=\"10.1186/s40623-021-01531-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKawai T, Yukutake Y, Doke R, Honda R (2024) Contribution of aseismic slips to earthquake swarms at the Hakone volcano. Earth Planet Space 76:152. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-024-02098-1\u003c/span\u003e\u003cspan address=\"10.1186/s40623-024-02098-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoge H, Sato T, Arimoto J et al (2025) Inception of ridge-ridge-ridge triple junctions: Morphostructural analysis and dynamics in the early back-arc extension of the northern Okinawa Trough. Geology 53:269\u0026ndash;273. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1130/g52640.1\u003c/span\u003e\u003cspan address=\"10.1130/g52640.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarone CJ, Scholtz CH, Bilham R (1991) On the mechanics of earthquake afterslip. J Geophys Res 96:8441\u0026ndash;8452. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/91jb00275\u003c/span\u003e\u003cspan address=\"10.1029/91jb00275\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatsu\u0026rsquo;ura M, Hasegawa Y (1987) A maximum likelihood approach to nonlinear inversion under constraints. Phys Earth Planet Inter 47:179\u0026ndash;187. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0031-9201(87)90076-8\u003c/span\u003e\u003cspan address=\"10.1016/0031-9201(87)90076-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinami H, Ohara Y, Tendo H (2021) Volcanic and tectonic features of Shirahama Bank in the northern Ryukyu Arc: Implications for cross-arc volcanism controlled by arc-parallel extension. Mar Geol 441:106623. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.margeo.2021.106623\u003c/span\u003e\u003cspan address=\"10.1016/j.margeo.2021.106623\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNakamura M, Kinjo A (2018) Activated seismicity by strain rate change in the Yaeyama region, south Ryukyu. Earth Planet Space 70:154. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-018-0929-y\u003c/span\u003e\u003cspan address=\"10.1186/s40623-018-0929-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNishikawa H, Kuwatani T, Tada N, Kayama Watanabe H (2023) Simulated distributions of pumice rafts in Japan following eruptions at volcanic islands and submarine volcanoes. Prog Earth Planet Sci 10:21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40645-023-00552-4\u003c/span\u003e\u003cspan address=\"10.1186/s40645-023-00552-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNishimura S, Hashimoto M, Ando M (2004) A rigid block rotation model for the GPS derived velocity field along the Ryukyu arc. Phys Earth Planet Inter 142:185\u0026ndash;203. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.pepi.2003.12.014\u003c/span\u003e\u003cspan address=\"10.1016/j.pepi.2003.12.014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNishimura T (2014) Short-term slow slip events along the Ryukyu Trench, southwestern Japan, observed by continuous GNSS. Progress Earth Planet Sci 1:1\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40645-014-0022-5\u003c/span\u003e\u003cspan address=\"10.1186/s40645-014-0022-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNishimura T, Hiramatsu Y, Ohta Y (2023) Episodic transient deformation revealed by the analysis of multiple GNSS networks in the Noto Peninsula, central Japan. Sci Rep 13:8381. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-023-35459-z\u003c/span\u003e\u003cspan address=\"10.1038/s41598-023-35459-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNishimura T, Matsuzawa T, Obara K (2013) Detection of short-term slow slip events along the Nankai Trough, southwest Japan, using GNSS data. J Geophys Res Solid Earth 118:3112\u0026ndash;3125. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/jgrb.50222\u003c/span\u003e\u003cspan address=\"10.1002/jgrb.50222\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNishimura T, Munekane H, Yarai H (2011) The 2011 off the Pacific coast of Tohoku Earthquake and its aftershocks observed by GEONET. Earth Planet Space 63:22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5047/eps.2011.06.025\u003c/span\u003e\u003cspan address=\"10.5047/eps.2011.06.025\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNishimura T, Ozawa S, Murakami M et al (2001) Crustal Deformation caused by magma migration in the. Geophys Res Lett 28:3745\u0026ndash;3748. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2001GL013051\u003c/span\u003e\u003cspan address=\"10.1029/2001GL013051\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eObara K, Kato A (2016) Connecting slow earthquakes to huge earthquakes. Science 353:253\u0026ndash;257. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aaf1512\u003c/span\u003e\u003cspan address=\"10.1126/science.aaf1512\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhta Y, Ohtate M Quality assessment of daily GNSS coordinate time series from the SoftBank proprietary reference network. Earth Planets Space in-\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhta Y, Ohzono M (2022) Potential for crustal deformation monitoring using a dense cell phone carrier Global Navigation Satellite System network. Earth Planet Space 74:25. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-022-01585-7\u003c/span\u003e\u003cspan address=\"10.1186/s40623-022-01585-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhtate M, Ohta Y, Mitsui Y (2025) Significant afterslip contribution to postseismic deformation in Sado Island following the 2024 Noto Peninsula earthquake: insights from two dense GNSS observation networks. Earth Planet Space 77:74. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-025-02203-y\u003c/span\u003e\u003cspan address=\"10.1186/s40623-025-02203-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhzono M, Takahashi H, Ohta Y, Ohno K (2025) Transient inflation event at Kussharo caldera, eastern Hokkaido, Japan in 2021\u0026ndash;2023 captured by multiple GNSS observation networks. Earth Planet Space 77:41. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-025-02173-1\u003c/span\u003e\u003cspan address=\"10.1186/s40623-025-02173-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkada Y (1995) Simulated empirical law of coseismic crustal deformation. J Phys Earth 43:697\u0026ndash;713. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4294/jpe1952.43.697\u003c/span\u003e\u003cspan address=\"10.4294/jpe1952.43.697\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkada Y (1992) Internal deformation due to shear and tensile faults in a half-space. Bull Seismol Soc Am 82:1018\u0026ndash;1040. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1785/BSSA0820021018\u003c/span\u003e\u003cspan address=\"10.1785/BSSA0820021018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkada Y, Nishimura T (2025) Investigation on short-term slow slip events in the northeast Japan subduction zones using decadal GNSS data. Earth Planet Space 77:45. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-025-02175-z\u003c/span\u003e\u003cspan address=\"10.1186/s40623-025-02175-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkada Y, Nishimura T, Tabei T et al (2022) Development of a detection method for short-term slow slip events using GNSS data and its application to the Nankai subduction zone. Earth Planet Space 74:1\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-022-01576-8\u003c/span\u003e\u003cspan address=\"10.1186/s40623-022-01576-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkada Y, Ueda T (2024) PySTEL: Python module for seasonal-trend-episodic decomposition technique based on LOESS (version 1.1.0). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5281/zenodo.13990750\u003c/span\u003e\u003cspan address=\"10.5281/zenodo.13990750\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Zenodo\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkada Y, Ueda T (2025) Development of a Seasonal Adjustment Technique STEL and its Application to Daily GNSS Data. J Geodetic Soc Japan 71:1\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.11366/sokuchi.71.1\u003c/span\u003e\u003cspan address=\"10.11366/sokuchi.71.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOzawa S, Miyazaki S, Hatanaka Y et al (2003) Characteristic silent earthquakes in the eastern part of the Boso peninsula, Central Japan. Geophys Res Lett 30(6):1283. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2002gl016665\u003c/span\u003e\u003cspan address=\"10.1029/2002gl016665\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRubin AM, Pollard DD (1988) Dike-induced faulting in rift zones of Iceland and Afar. Geology 16:413\u0026ndash;417. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1130/0091-7613(1988)016%3C0413:DIFIRZ%3E2.3.CO;2\u003c/span\u003e\u003cspan address=\"10.1130/0091-7613(1988)016%3C0413:DIFIRZ%3E2.3.CO;2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSagiya T, Miyazaki S, Tada T (2000) Continuous GPS Array and Present-day Crustal Deformation of Japan. Pure Appl Geophys 157:2303\u0026ndash;2322. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/PL00022507\u003c/span\u003e\u003cspan address=\"10.1007/PL00022507\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScholz CH (1998) Earthquakes and friction laws. Nature 391:37\u0026ndash;42. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/34097\u003c/span\u003e\u003cspan address=\"10.1038/34097\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun T, Wang K, Iinuma T et al (2014) Prevalence of viscoelastic relaxation after the 2011 Tohoku-oki earthquake. Nature 514:84\u0026ndash;87. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nature13778\u003c/span\u003e\u003cspan address=\"10.1038/nature13778\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTakagi R, Uchida N, Obara K (2019) Along-strike variation and migration of long‐term slow slip events in the western Nankai subduction zone, japan. J Geophys Res Solid Earth 124:3853\u0026ndash;3880. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2018jb016738\u003c/span\u003e\u003cspan address=\"10.1029/2018jb016738\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTu Y, Heki K (2017) Decadal modulation of repeating slow slip event activity in the southwestern Ryukyu Arc possibly driven by rifting episodes at the Okinawa Trough. Geophys Res Lett 44:9308\u0026ndash;9313. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/2017gl074455\u003c/span\u003e\u003cspan address=\"10.1002/2017gl074455\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUchide T, Imanishi K (2018) Underestimation of microearthquake size by the magnitude scale of the Japan Meteorological Agency: Influence on earthquake statistics. J Geophys Res Solid Earth 123:606\u0026ndash;620. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/2017jb014697\u003c/span\u003e\u003cspan address=\"10.1002/2017jb014697\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUrai M, Machida S (2005) Discolored seawater detection using ASTER reflectance products: A case study of Satsuma-Iwojima, Japan. Remote Sens Environ 99:95\u0026ndash;104. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2005.04.028\u003c/span\u003e\u003cspan address=\"10.1016/j.rse.2005.04.028\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVall\u0026eacute;e M, Nocquet J-M, Battaglia J et al (2013) Intense interface seismicity triggered by a shallow slow slip event in the Central Ecuador subduction zone. J Geophys Res Solid Earth 118:2965\u0026ndash;2981. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/jgrb.50216\u003c/span\u003e\u003cspan address=\"10.1002/jgrb.50216\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWessel P, Luis JF, Uieda L et al (2019) The Generic Mapping Tools Version 6. Geochem Geophys Geosys 20(11):5556\u0026ndash;5564. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2019GC008515\u003c/span\u003e\u003cspan address=\"10.1029/2019GC008515\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYamada T, Ohta Y, Nishimura T et al (2025) Coseismic slip distribution of the 2024 Noto Peninsula earthquake deduced from dense global navigation satellite system network and interferometric synthetic aperture radar data: effect of assumed dip angle. Earth Planet Space 77:19. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40623-025-02154-4\u003c/span\u003e\u003cspan address=\"10.1186/s40623-025-02154-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYokose H, Sato H, Fujimoto Y et al (2010) Mid-pleistocene submarine acidic volcanism of the Tokara Islands, Japan. J Geogr 119:46\u0026ndash;68. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5026/jgeography.119.46\u003c/span\u003e\u003cspan address=\"10.5026/jgeography.119.46\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"earth-planets-and-space","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"epsp","sideBox":"Learn more about [Earth, Planets and Space](http://earth-planets-space.springeropen.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/epsp/default.aspx","title":"Earth, Planets and Space","twitterHandle":"@SpringerOpen","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Earthquake swarm, Transient displacement, GNSS, GEONET, Private-sector GNSS, Aseismic fault slip, Open crack","lastPublishedDoi":"10.21203/rs.3.rs-8399343/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8399343/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTransient crustal deformation is often observed in conjunction with earthquake swarms and is attributed to aseismic processes that drive swarm activity, such as magma intrusion, fluid migration, or slow slip events. Off the Tokara Islands, along the volcanic front of the Ryukyu subduction zone in the southern Japanese Archipelago, earthquake swarms have repeatedly occurred; however, the driving mechanism of this anomalous seismicity remains unclear. Here, we investigate this mechanism through integrative analyses of Global Navigation Satellite System (GNSS) data from the Geospatial Information Authority of Japan and SoftBank Corp. The daily GNSS position time series from multiple stations exhibits clear transient signals that coincide with recurrent earthquake swarms. Notably, during the June\u0026ndash;July 2025 episode, both the temporal trend and spatial pattern of the displacement changed abruptly at the onset of the secondary swarm. Motivated by these distinctive characteristics, we estimated the source models for the 2025 episode by dividing it into three phases. For all phases, the displacements calculated from rectangular faults with shear slip reproduce the observed displacements well. Vertical open cracks provide an alternative source model, but their data fit is systematically worse than that of the shear-slip models. A stacked time series of 30-second kinematic GNSS positions further suggests that the temporal relationship between aseismic deformation and seismicity can vary even within a single episode. Based on the shear-slip and open crack models for the 2025 episode, we propose two scenarios for the driving mechanisms of aseismic displacements and earthquake swarms. In both scenarios, we speculate that the supply of material from a region deeper than the aseismic source and swarm hypocenters plays a key role in generating these episodes.\u003c/p\u003e","manuscriptTitle":"Recurrent earthquake swarms and transient crustal deformation off the Tokara Islands, southern Japan: Insights from the 2025 swarm sequence","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-14 12:38:57","doi":"10.21203/rs.3.rs-8399343/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Minor Revision","date":"2026-02-26T21:43:10+00:00","index":"","fulltext":""},{"type":"reviewerAgreed","content":"","date":"2026-01-12T22:16:56+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-01-12T13:08:56+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-12-24T11:26:12+00:00","index":"","fulltext":""},{"type":"submitted","content":"Earth, Planets and Space","date":"2025-12-18T18:59:57+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"earth-planets-and-space","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"epsp","sideBox":"Learn more about [Earth, Planets and Space](http://earth-planets-space.springeropen.com)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/epsp/default.aspx","title":"Earth, Planets and Space","twitterHandle":"@SpringerOpen","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f4c5230a-1a28-4968-a698-25732c7662c5","owner":[],"postedDate":"January 14th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-18T04:35:21+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-14 12:38:57","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8399343","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8399343","identity":"rs-8399343","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}