4D seismic tomography unveiling velocity increase prior to seismic swarms at the Tatun volcano group of Taiwan

4D seismic tomography unveiling velocity increase prior to seismic swarms at the Tatun volcano group of Taiwan | 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 Article 4D seismic tomography unveiling velocity increase prior to seismic swarms at the Tatun volcano group of Taiwan Hsin-Chieh Pu, Cheng-Horng Lin, Ya-Chuan Lai, Min-Hung Shih, Yi-Heng Li This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5675134/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 25 May, 2025 Read the published version in Scientific Reports → Version 1 posted 6 You are reading this latest preprint version Abstract The Tatun Volcano Group (TVG) is in close proximity to the Taipei metropolitan area, which is home to over 7 million residents. Although there has been no recorded volcanic eruption in the TVG throughout human history, recent seismic observations suggest that it may still be active. To investigate possible volcanic seismicity and activity in the TVG, we conducted a 4-dimensional seismic tomography study using abundant seismic data collected from 2014 to 2021. We obtained 3D seismic velocity structures to examine both the temporal and spatial variations in seismicity each year. Our results show that the dramatic increase in seismicity in 2019 followed an increase in P-wave seismic velocity in the Dayoukeng fumarole area and Mt. Chihsin in 2018. This increase in seismic velocity may have been caused by rocks or sediments subjected to higher pressure beneath the TVG, resembling a pressure cooker on a flame. Thus, the sequential rise in both seismic velocity and seismicity strongly suggests that careful monitoring of temporal velocity variations in the volcanic area might provide an early warning of potential seismic swarms or volcanic activity in the future. Earth and environmental sciences/Natural hazards Earth and environmental sciences/Solid earth sciences Earth and environmental sciences/Solid earth sciences/Seismology Earth and environmental sciences/Solid earth sciences/Volcanology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Introduction The Tatun Volcano Group (TVG) is located at the northern tip of Taiwan (Fig. 1 ) and is directly linked to the Taipei metropolitan area, home to over 7 million residents across the cities of Taipei and New Taipei. The horizontal distance from Mt. Chihsin, the highest peak in the TVG and likely the site of its most recent eruption 1 , 2 , to the Taipei 101 skyscraper in downtown Taipei is less than 15 km. In fact, the northernmost parts of Taipei City, including the Beitou and Shilin districts, partially overlap with the TVG. As a result, Taipei City, similar to Naples in Italy, Auckland in New Zealand, and Cartago in Costa Rica, can be considered one of the major cities located on an active volcano, where a volcanic eruption could potentially cause widespread loss of life and regional disruption 3 . One of the most well-known cities affected by a volcanic eruption is ancient Pompeii in Italy, which was completely buried by the eruption of Mt. Vesuvius in 79 A.D. 4 . In fact, the potential global economic impact of a future eruption of the TVG has been seriously considered 5 . Although there is no recorded volcanic eruption in the TVG throughout human history, recent geological and seismic observations suggest that it may still be active. The most significant evidence from these recent observations includes the dating of volcanic ash layers 1 , 2 and seismic detections of a magma reservoir 6 – 8 . First, radiocarbon dating of various volcaniclastic deposits in the TVG suggests that the last eruption occurred approximately 6,000 years ago at Mt. Chihsin 1 . Besides, a later study by Zellmer et al. 2 suggests that the last eruption may have occurred as recently as 1,350 years ago. Both studies consistently indicate that the last eruption occurred less than 10,000 years ago. Second, a magma reservoir at mid-crust depth was first detected through observations of both S-wave shadowing and P-wave delays recorded by a dense seismic array in the TVG 6 . This magma reservoir was further confirmed by tomographic images showing low-velocity zones 7 and strong seismic attenuation 8 following the successful deployment of a dense seismic array in northern Taiwan in 2018 9–11 . Taken together, the dating results and seismic observations strongly suggest that the TVG is an active volcano according to established criteria 12 . Therefore, disaster prevention and mitigation measures for potential volcanic hazards are critical issues for the Taipei metropolitan area. Seismicity at the TVG Since 2003, a dense seismic array of 40 broadband stations has been gradually deployed to monitor volcanic activity in the TVG 13 – 16 . Many micro-earthquakes have been detected by the Taiwan Volcano Observatory at Tatun. Volcanic seismicity is generally concentrated around the administrative boundary between Taipei and New Taipei cities, as shown in Fig. 1 . Several key studies have examined the volcanic activity in the TVG, including reports of significant seismic events 13 – 14 , 16 – 21 . Notably, between 2014 and 2018, more than 1,000 earthquakes occurred at depths of 0 to 2 km beneath the Dayoukeng fumarole, forming a distinct seismic conduit 16 . Additionally, several "heartbeat-like" earthquake sequences have been observed in the same region 20 . These seismic phenomena strongly suggest ongoing degassing processes beneath the Dayoukeng fumarole 21 . The presence of both the seismic conduit and the heartbeat-like sequences indicates that the area is a likely pathway for future volcanic eruptions. An analysis of the temporal variation in background seismicity from 2014 to 2021 reveals that the number of earthquakes typically ranged from 100 to 300 per month. However, seismic swarms occasionally occurred, with several hundred to over a thousand earthquakes occurring over just a few days or within a single month (Fig. 2 ). Based on the slope of the accumulated earthquake count, the background seismicity can be roughly divided into three periods: (1) 2014 to early 2015, (2) early 2015 to 2018, and (3) 2019 to 2021. On average, the number of earthquakes per month began at a relatively low level of ~ 115 events during the first period, increased to ~ 198 events during the second period, and surged to ~ 300 events in the final period. Notably, there was a significant spike in the total accumulated events in 2019, primarily driven by a major seismic event. This jump resulted from more than 1,400 earthquakes triggered by a magnitude 4.2 earthquake on January 28, 2019. In addition to the significant increase in the total number of volcanic earthquakes observed from 2019 onward, larger earthquakes (M > 3) occurred more frequently during the third period compared to previous years (Fig. 3 ). In 2014, only one such earthquake was recorded, and there were none from 2015 to 2018. However, it is noteworthy that more than 20 larger earthquakes (M > 3) were detected between 2019 and 2021 (Fig. 3 ). Interestingly, most of these larger earthquakes (M > 3) were concentrated around the Dayoukeng and Mt. Chihsin areas (Fig. 3 ), which are two major degassing sites in the Tatun Volcanic Group (TVG) 16 . Among these events, one of the largest earthquakes, with a magnitude of 4.2, occurred near Dayoukeng on January 28, 2019. Additionally, two significant earthquakes with magnitudes of 3.6 and 3.7 took place near Mt. Chihsin on February 9 and August 6, 2019, respectively. These dramatic increases in volcanic seismicity around the Dayoukeng fumarole and Mt. Chihsin strongly suggest that temporal changes in physical or chemical conditions within the shallow crust beneath the TVG may have occurred. To investigate potential temporal changes in the shallow crustal velocity structures beneath the TVG (Fig. 2 ), we conducted a 4D seismic tomography analysis 22 using the extensive seismic data recorded at the TVG between 2014 and 2021. We carefully selected non-redundant earthquake data to ensure reliable tomographic images for each year. We then calculated the differences between the velocity structures obtained in 2018 and those from other years to assess the dramatic increase in volcanic earthquakes observed in the TVG. By examining the strong correlations between velocity changes and seismicity variations, both temporally and spatially, we propose that temporal velocity variations in the TVG could serve as a potential warning signal for seismic swarms or volcanic activity. Results Velocity structures at a depth of 3 km The results from the 4D seismic tomography reveal significant perturbations in P-wave seismic velocity (Vp) in 2018. For discussing the reliable inverted results in the Figs. 4 – 8 , we only plotted the derivative weight sum (DWS 22 – 25 ) of the velocity node is higher than 5, which means the ray density at the velocity node is five times higher than the average ray density during the inversion. One of the most notable layers of perturbations occurs at a depth of 3 km, with data from 2014 to 2021 (Fig. 4 ). In general, lateral velocity perturbations during the first five years (2014–2018) are more pronounced than those in the last three years (2019–2021). Among these, the strongest lateral perturbations were observed in 2018, particularly around the Dayoukeng fumarole (marked by a pink triangle) and Mt. Chihsin (marked by a yellow triangle). Beneath the Dayoukeng fumarole, a significant positive velocity perturbation (H1, marked in blue) is clearly identifiable. This major positive anomaly is surrounded by a zone of strong negative velocity perturbations (marked by a dashed circle). Similarly, around Mt. Chihsin, another prominent positive velocity perturbation (H2) is encircled by negative perturbations (marked by another dashed circle). To further highlight the intensity of the velocity perturbations in the Dayoukeng and Mt. Chihsin areas during 2018, we present velocity differences between 2018 and the other years in four depth profiles below. Velocity anomaly beneath Dayoukeng To examine the strong perturbations beneath the Dayoukeng fumarole, we selected reliable P-wave velocity structures, validated by the checkerboard test, along two vertical profiles (Profiles A-A’ and B-B’) that intersect near Dayoukeng (Fig. 5 ). These two profiles are nearly perpendicular to each other. First, we plotted the P-wave velocity differences between 2018 and each of the other years along Profile A-A’. Notably, the most prominent and consistent anomaly across most profiles is a nearly vertical channel characterized by negative velocity differences extending from sea level down to a depth of approximately 3 to 4 km beneath the Dayoukeng fumarole. One of the most representative profiles displaying this nearly vertical channel (R1) is shown in Profile A-A’ for the velocity difference between 2017 and 2018. Interestingly, similar negative anomalies (R1) beneath the Dayoukeng fumarole are also observed along Profile BB’ in the other years (Fig. A1). To further emphasize the significant change in P-wave velocities within the nearly vertical channel beneath the Dayoukeng fumarole (R1), we plotted the velocity differences between 2017 and 2018 along both Profiles A-A’ and B-B’ (Fig. 6 ). Overall, both profiles indicate that the P-wave velocities within the nearly vertical channel (R1) in 2018 are consistently higher compared to the other years. Velocity anomaly beneath Mt. Chihsin To examine the strong perturbations beneath Mt. Chihsin, we analyzed two additional vertical profiles (Profiles C-C’ and D-D’) that intersect at Mt. Chihsin and are nearly perpendicular to each other. For Profile C-C’ (Fig. 7 ), a major negative anomaly (R2) is consistently observed at depths of 3–4 km beneath Mt. Chihsin in all years, despite the presence of other individual anomalies in different years. The most representative anomaly (R2) is clearly visible in the velocity difference between 2017 and 2018. These negative anomalies are not only observed along Profile C-C’ (Fig. 7 ), but also along Profile D-D’ (Fig. A2). For example, Fig. 8 shows that the representative anomaly beneath Mt. Chihsin is clearly identified along both Profiles C-C’ and D-D’. Overall, the negative anomalies consistently observed along both profiles indicate a significant change in P-wave velocities at depths of 3–4 km beneath Mt. Chihsin between 2018 and the other years. Specifically, the velocities at R2 in 2018 are significantly higher than those in the other years. Discussion It is particularly intriguing to note that the dramatic increase in seismicity observed in the TVG in 2019 (Fig. 2 ) occurred shortly after a significant increase in P-wave seismic velocity beneath both the Dayoukeng fumarole and Mt. Chihsin areas in 2018 (Figs. 4 – 8 ). On one hand, the sharp rise in seismic activity in 2019 is evident in both the total number of volcanic earthquakes (Fig. 2 ) and the occurrence of larger earthquakes (M > 3.0) (Fig. 3 ). This marked increase in seismicity was the strongest observed in the TVG since the deployment of a dense seismic array in 2003 13 . On the other hand, the notable increase in P-wave seismic velocity beneath both the Dayoukeng and Mt. Chihsin areas in 2018 occurred just prior to the surge in seismic activity in 2019 (Figs. 4 – 8 ). This temporal sequence of changes in velocity structures and seismicity suggests a potentially strong causal relationship between the two phenomena. The significant alterations in Vp structures, along with the clustering of larger earthquakes (M > 3), beneath both the Dayoukeng fumarole and Mt. Chihsin, are generally consistent with findings from previous studies 15 , 21 . Notably, background seismicity in the TVG often clusters around these two areas 16 . Beneath Mt. Chihsin, earthquakes primarily occur within a sphere-like zone at depths ranging from 0.5 km to 2.0 km. Interestingly, earthquake clustering beneath the Dayoukeng fumarole appears to be confined to a nearly vertical conduit, approximately 500 meters in diameter, extending from sea level down to 2 km depth 16 . Detailed stress analyses of focal mechanisms from numerous micro-earthquakes, coupled with surface observations of the sulfur-to-carbon dioxide ratio, suggest that the seismic conduit beneath Dayoukeng may gradually seal with smectite-rich rocks at its upper levels, leading to the accumulation of inflationary pressure over time 21 . In other words, both the sphere-like and conduit seismic zones are likely influenced by the ascent of volcanic fluids from a deep magma reservoir. Periodically, these zones may seal at the top, causing inflationary pressure to build within the seismic zones 21 . The sequential changes in seismic velocities in 2018 and seismicity in 2019 may suggest a plausible link, wherein the increase in P-wave seismic velocity (Vp) over the years could result from higher pressure in the rocks or sediments beneath the TVG. While Vp can be influenced by various factors, including temperature, pressure, chemical composition, and water content 26 – 32 , the most likely explanation for the Vp increase in this region is an increase in rock confining pressure, as other factors appear less likely to account for the change. Previous studies and laboratory experiments consistently show that Vp increases with confining pressure in rocks 27 – 31 . In fact, increased pressure has already been observed in the TVG, as indicated by spatial-temporal variations in the focal mechanisms of numerous micro-earthquakes 21 and precise leveling surveys 33 . In contrast, increases in temperature or water content typically lead to a reduction in Vp 26 , 31 – 32 . Additionally, changes in seismic velocity occurring over the course of months or even years are unlikely to result from chemical variations in the rocks, as such changes would generally require more time to manifest. The processes underlying the sequential changes in seismic velocity and earthquake swarms associated with confining pressure along major volcanic pathways in the TVG can be divided into three distinct periods from 2014 to 2021. The major pathway beneath the Dayoukeng fumarole is well delineated as a nearly vertical conduit 16 . A simplified conceptual model representing the changes in seismicity, inflation pressure, and seismic velocity over these three periods is shown in Fig. 9 . Period 1 (2014) During this period, seismicity is primarily driven by stable volcanic degassing beneath the fumarole site, following the smooth volcanic pathway. Earthquake activity is largely confined to clustering within the pathway itself 16 . Period 2 (2015–2018) Over this period, the top of the major volcanic degassing pathway begins to gradually seal, leading to an increase in confining pressure within the pathway. It may be like a pressure cooker on a flame. Consequently, a dramatic rise in seismic velocity is observed in 2018 (Figs. 5 and 6 ). In addition to the continued clustering of seismicity within the pathway, some earthquakes occur in the surrounding area, likely due to the inflationary stress exerted by the sealing of the pathway. Period 3 (2019 and beyond) Beginning in 2019, strong seismic swarms are unexpectedly observed in and around the pathway. The significant dynamic stress from larger earthquakes (M > 3) likely breaks the seal at the top of the pathway. It is just similar to a pressure cooker explosion. Following this, seismicity, seismic velocity, and confining pressure may return to a temporarily stable state, similar to the conditions in Period 1, until the sealing of the volcanic degassing pathway occurs again. The sequential changes in both seismicity and seismic velocity around the two major volcanic degassing pathways—Mt. Chihsin and Dayoukeng fumarole—suggest that monitoring temporal variations in seismic velocity could provide a valuable early warning system for seismic swarms or volcanic activity, as illustrated in the conceptual model in Fig. 9 . First, major degassing zones such as Mt. Chihsin and Dayoukeng fumarole can be clearly identified from clustering seismic events, which reflect the spatial distribution of seismicity 16 . Second, the migration of volcanic fluids or pressure variations along the primary pathways can cause transient changes in seismic clustering 21 . These variations in seismicity can be routinely monitored through local earthquake detection systems operated by the Taiwan Volcano Observatory at Tatun. Third, temporal variations in seismic velocity can be tracked using 4D seismic tomography, provided that sufficient seismic data is collected within a given time window. By comparing these temporal variations in seismicity and velocity structures, it may be possible to issue early warnings for strong seismic swarms or potential volcanic activity in the future, based on the detailed analysis proposed in the model in Fig. 9 . Method and Data Selection To obtain 3-D velocity structures that change over time beneath the TVG, we applied the double-difference tomography method 22 to seismic data collected by the Taiwan Volcano Observatory at Tatun from 2014 to 2021 (Fig. 1). During the tomographic inversion process, various parameter values were tested based on recommendations from Zhang and Thurber 22 . A detailed description of the inversion procedure can be found in one of our previous studies 16 . The final inversion parameters used in this study are as follows: the number of iterations was 18, and the damping factor for both velocity structure inversion and earthquake relocation was set to 125, which was optimized after trying different values ranging from 25 to 200 (Fig. A3). Such an optimal result is based on the trade-off between the damping factor and CND (Condition Number) of ~100, which is the ratio of the largest to smallest eigenvalue 22 . For constraining the order smooth model, we calculated the recovery percentage of the checkerboard model under a series of smoothing constraints, ranging from 0.5 to 5.0 in increments of 0.5 (Fig. A4). We ultimately selected a smoothing constraint of 1.5, as the improvement in recovery percentage became marginal for smoothing constraints greater than 1.5. The weighting ratio between absolute and differential arrival times was reduced from 100 to 0.01. The large number of relative arrival times for both P- and S-waves, approximately four times the number of absolute arrivals, significantly minimized systematic errors, thus improving the TVG velocity model. The seismic data used in this study were recorded by a dense network of 40 seismic stations in the TVG 6 (Fig. 10). Each station is equipped with a three-component broadband seismometer (Guralp CMG-6TD or Maredian Compact). Seismic data are continuously recorded at a sampling rate of 100 Hz and transmitted in near-real-time to the Taiwan Volcano Observatory at Tatun and the Institute of Earth Sciences, Academia Sinica, Taiwan, via telephone lines or wireless radio. The arrival times of both the first P- and S-waves generated by local earthquakes are automatically picked by the software and later manually verified to determine the earthquake parameters, including hypocenter and magnitude. In total, 22,290 local earthquakes were detected between 2014 and 2021 (Fig. 1). Among the 22,290 earthquakes recorded at 40 seismic stations in the TVG between 2014 and 2021, we carefully selected 14,221 events to avoid redundancy in a model grid during the tomographic inversion. The dense gridding velocity model has been designed for an area of approximately 100 km² to cover the TVG (Fig. 10). The horizontal grid spacing is evenly set to 0.25 km, and the vertical grid is starting 0.5 km above sea level and increasing by 0.5 km increments down to a depth of 5.0 km. The denser grids at depths less than 5 km are designed to accommodate the high density of local events observed in the uppermost crust (Fig. 1). The velocity values at each grid node were slightly adjusted from the previous study by one of the previous studies 16 , who successfully inverted 3D seismic images based on earlier seismic data. For each grid cell, with a volume of 0.25 km × 0.25 km × 0.5 km (Fig. 10), only two earthquakes with the smallest residuals and location errors were selected for the inversion. Thus, the model distributions are expected to be same between the different time steps 34 . Since seismicity is not evenly distributed across the study area (Fig. 1), many redundant arrival times had to be adjusted for the tomographic inversion. In regions with dense clustering of earthquakes, such as around Mt. Chihsin and the Dayoukeng fumarole, we applied stricter criteria to select the most reliable seismic data. These criteria included not only smaller uncertainties in earthquake locations but also a greater number of seismic stations recording each event. In contrast, we aimed to retain all arrival times for ray paths passing through regions where seismic ray coverage was more limited. This approach ensures that the tomographic images are more evenly reliable across the entire study area. To examine potential temporal variations in the 3D velocity structures, the seismic data were divided into eight sets, one for each year from 2014 to 2021. To maintain consistency in the inversion process, the number of earthquakes selected for each year was roughly comparable (Table 1), ensuring that the ray-path coverage in each year provided similar resolution for the seismic images. This approach allows us to derive annual tomographic images and analyze the temporal variations in the 3D velocity structures. Table 1. Total number of earthquakes selected for tomographic inversion in each year. Year 2014 2015 2016 2017 2018 2019 2020 2021 No. 1428 1519 1913 1865 1230 2309 2326 1631 To assess the effectiveness of the inversion, a checkerboard test was conducted using the following procedure. First, a checkerboard velocity model with velocity perturbations of ±10% and a grid spacing of 1.0 km was applied to create a simulated dataset. Second, simulated arrivals of both P- and S-waves, generated from actual seismic events and stations in the TVG, were used for the double-difference tomographic inversion. Finally, a comparison between the checkerboard model and the inverted results showed that the P-wave velocity structures (Vp) were successfully recovered for most of the study area at depths shallower than 3.5 km (Fig. 11). This indicates that the inverted velocity structures in the shallow crust are highly reliable, although the area with reliable results gradually shrinks with depth due to the limited availability of deeper earthquakes. Besides, the results of the checkerboard tests from 2014 to 2021 for the top five layers are shown Fig. 12. The inverted images from the checkerboard velocity model were clear and consistent across most of the study area each year. For instance, the differences in checkerboard tests between years were minimal in most areas shallower than 2 km in depth (Fig. 13). The velocity perturbations (> 1.0 km/s) at the areas where we have discussed at Figures 5-8 are significantly greater than the residuals around ± 0.25 to 0.5 km/s observed by differencing checkerboards (Figs. 14 and 15). It indicates that the major velocity changes between time steps from 2014 to 2021 can be well resolved. To further demonstrate the resolution similarity between the different time steps, the values of the derivative weight sum (DWS 22 ) for the model space in each year have been plotted for showing the seismic ray-path density is comparable (Fig. A5). It concludes that the inverted results are not only reliable, but also almost consistent year to year. Therefore, the temporal variations in the inverted results are meaningful and warrant further discussion. Declarations Author Contribution Y.C., M.H. and Y. H. collected seismic data and prepared some figures; H. C. analyzed the data and joined the writing as well as detailed discussion; C. H. analyzed the data and wrote the manuscript text.All authors reviewed the manuscript. Acknowledgments: We would like to express our gratitude to our colleagues at the Institute of Earth Sciences, Academia Sinica as well as the Taiwan Volcano Observatory at Tatun, National Centre for Research on Earthquake Engineering in Taipei, Taiwan for their efforts in deploying and maintaining the seismic arrays since 2013. This work was funded by the National Science and Technology Council of Taiwan. We are appreciated at the valuable comments and suggestions by two anonymous reviewers to improve this work. Data Availability Seismic data used in this study have been collected and carefully analyzed by the Taiwan Volcano Observatory at Tatun (TVO) in Taiwan. Due to their size and data regulation, completely seismic records can be made available upon reasonable request from the corresponding author (C.H. Lin). Some plots were made using the Generic Mapping Tools (version 4.5.2; URL:gmt.soest.hawaii.edu) and the SAC software (version 101.5; URL:ds.iris.edu). References Belousov, A., Belousova, M., Chen, C. H. & Zellmer, G. F. Deposits character and timing of recent eruptions and gravitational collapses in Tatun volcanic group, northern Taiwan: hazard-related issues. J. Volcanol. Geotherm. Res. 191 , 205-221 (2010). Zellmer, G. F., Rubin, K. H., Miller, C. A., Belousov, A. & Belousova, M. Resolving discordant U–Th–Ra ages: constraints on petrogenetic processes of recent effusive eruptions at Tatun Volcano Group, northern Taiwan. Geological Society 422 (2015). Chester, D. K., Degg, J. M., Duncan, A. M. & Guest, J. E. The increasing exposure of cities to the effects of volcanic eruptions: a global survey, Global Environmental Change Part B, Environmental Hazards 2 (3), 89-103 (2000). Giacomelli, L., Perrotta, A., Scandone, R. & Scarpati, C. The eruption of Vesuvius of 79 AD and its impact on human environment in Pompeii, Episodes 26 , 235-238 (2003). Mani, L., Tzachor, A. & Cole, P. Global catastrophic risk from lower magnitude volcanic eruptions. Nat. Commun. 12 , 4756 (2021). Lin, C. H.Evidence for a magma reservoir beneath the Taipei metropolis of Taiwan from both S-wave shadows and P-wave delays, Scientific Reports 6 , 39500 (2016). Huang, H. H. et al. Unveiling Tatun volcanic plumbing structure induced by post-collisional extension of Taiwan mountain belt, Scientific Reports 11 , 5286 (2021). Lin, Y. P., Ko, J. Y. T., Huang, B. S., Lin, C. H., & Shih, M. H. Unveiling attenuation structures in the northern Taiwan volcanic zone, Scientific Reports 14 , 4716 (2024). Lin, C. H., Shih, M. H., & Lai, Y. C. A strong seismic reflector in the mantle wedge above the Ryukyu subduction system beneath northern Taiwan, Seism. Res. Lett., 91 (1), 310-316 (2020). Lin, C. H., Shih, M. H., & Lai Y. C. Mantle wedge diapirs detected by a dense seismic array in northern Taiwan, Scientific Reports 11 , 1561 (2021). Liu, C. N. et al. High resolution 3-D shear velocity model of northern Taiwan via Bayesian joint inversion of Rayleigh wave ellipticity and phase velocity with Formosa Array, J. Geophy. Res: Solid Earth 126 (5), e2020JB21610 (2021). Szakacs, A. Redefining active volcanoes: A discussion, Bull. Vol- canol. 56 , 321–325 (1994). Lin, C. H. et al. Preliminary analysis of tectonic earthquakes and volcanoseismic signals recorded at the Tatun volcanic group, northern Taiwan, Geophysical Research Letters 32 (10), L10313. (2005a). Lin, C. H. et al. Preliminary results of seismic monitoring at Tatun volcanic area of northern Taiwan. Terr. Atm. Ocean 16 , 563-577 (2005b). Pu, H.C. et al. Volcano-hydrothermal inflation revealed through spatial variation in stress field in Tatun volcano group, northern Taiwan, J. Volcanology Geothermal Research 390 , 160712 (2020a). Pu, H. C. et al. Active volcanism revealed from a clustering seismicity conduit beneath the long-resting Tatun volcano group in northern Taiwan, Scientific Reports 10 , 6153 (2020b). Pu, H. C. et al. The volcanic earthquake swarm of October 20, 2009 in the Tatun Area of Northern Taiwan. Terr. Atmos. Ocean. 25 , 625-635 (2014). Lin, C. H. & Pu,H. C. Very-Long-Period Seismic Signals at the Tatun Volcano Group, Northern Taiwan, Journal of Volcanology and Geothermal Research , 328 , 230–236 (2016). Lin, C. H. Probable dynamic triggering of phreatic eruption at the Tatun volcano group of Taiwan, Journal of Asian Earth Sciences 149 , 78-85 (2017a). Lin, C. H. Dynamic triggering of drumbeat seismicity at Tatun volcano group in Taiwan, Geophysical Journal International 210 , 354-359 (2017b). Pu, H. C. et al. Ascending volcanic fluids portended by spatiotemporal variations of the earthquake mechanisms in the Tatun volcano group in northern Taiwan, Geophy. Res. Letters 48 , e2020GL091686 (2021). Zhang, H. J. & Thurber, C. H. Double-difference tomography: The method and its application to the Hayward fault, California, Bull. Seismo. Soc. Am. 93 , 1875-1889 (2003). Thurber, C. H. Earthquake locations and three-dimensional crustal structure in the Coyote Lake area, central California, J. Geophys. Res. 88 , 8226-8236 (1983). Thurber, C. H. & Eberhart-Phillips, D. Local earthquake tomography with flexible gridding, Computers and Geosciences 25 , 809-818 (1999). Waldhauser, F., & Ellsworth, W.L. A double-difference earthquake location algorithm: method and application to the northern Hayward Fault, California, Bull. Seismo. Soc. Am. 90 , 1353–1368 (2000). Jaya, M.S. et al. Temperature dependence of seismic properties in geothermal rocks at reservoir conditions, Geothermics 39 , 115–123 (2010). Gao, K. & Gibson Jr., R. L. Pressure-dependent seismic velocities based on effective compliance theory and an asperity deformation model, Geophysics 77 (6), 229–243 (2012). Eberhart‐Phillips, D., Han, D-H. & Zoback, M. D. Empirical relationships among seismic velocity, effective pressure, porosity, and clay content in sandstone, Geophysics 54 (1), 82-89 (1989). Mukeri, T. & Mavko, G. Pore fluid effects on seismic velocity in anisotropic rocks, Geophysics 59 (2), 233-244 (1994). Kiddle, E. J. The Structure of the Crust and Magmatic System at Montserrat, Lesser Antilles. Ph.D. Thesis, University of Bristol, Bristol, UK (2011). Kristinsdóttir, L. et al. Electrical conductivity and P-wave velocity in rock samples from high-temperature Icelandic geothermal fields. Geothermics 39 , 94–105 (2010). Boitnott, G. N. Core Analysis for the Development and Constraint of Physical Models of Geothermal Reservoirs; DE-FG07–99ID13761; New England Research, Inc.: White River Junction, VT, USA, 52. (2003). Murase M., Lin, C. H., Kimata, F., Mori, H. & Pu, H. C. Volcano-hydrothermal activity detected by precise levelling surveys at the Tatun volcano group in northern Taiwan during 2006-2013. J. Volcanol. Geotherm. Res. 286 , 30-40 (2014). Patane, D., Barberi, G., Cocina, O., De Gori, O., Chiarabba, C. Time-resolved seismic tomography detects magma intrusions at Mount Etna, Science 313 , 5788 (2006). Additional Declarations No competing interests reported. Supplementary Files Appendix.docx Cite Share Download PDF Status: Published Journal Publication published 25 May, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Accepted 13 May, 2025 Reviews received at journal 12 May, 2025 Reviewers agreed at journal 12 May, 2025 Reviewers invited by journal 06 May, 2025 Submission checks completed at journal 02 May, 2025 First submitted to journal 21 Apr, 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-5675134","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":445646989,"identity":"7df72b70-33c1-4fba-93e6-1517e6803372","order_by":0,"name":"Hsin-Chieh Pu","email":"","orcid":"","institution":"Central Weather Administration","correspondingAuthor":false,"prefix":"","firstName":"Hsin-Chieh","middleName":"","lastName":"Pu","suffix":""},{"id":445646990,"identity":"b1aa3288-1a81-4ad6-b6af-972d981e7276","order_by":1,"name":"Cheng-Horng Lin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7ElEQVRIiWNgGAWjYBACPmYGxgcghgEQHwCTDFASF2BjZmAGqZBA05KARwsQScC0wAABLey8xyo+1NyrM2dgf3i4oMBOnoG9eZsE44/DeBzGl3ZzxrFiCcsGHoPDMwySDRt4jpVJMCTg08JjdpuHLUHC4AAPw2EegwOMDRI5ZkAtt/FqKf7zD6SF/QFIi32D/BvCWpgZ20BaGAxAWhIbJHgIajGW7O1LkNwAVA9EycltPGnFFglp/3Fq4ec/Y/jhx7cEfoPj7Y8/8/yxs+1nP7zxxgebNJxaEIAZZi+ISCBCwygYBaNgFIwC3AAAo61HpbmajR8AAAAASUVORK5CYII=","orcid":"","institution":"Academia Sinica","correspondingAuthor":true,"prefix":"","firstName":"Cheng-Horng","middleName":"","lastName":"Lin","suffix":""},{"id":445646991,"identity":"46aa805f-2f7f-4545-b7a0-e757a6e5556e","order_by":2,"name":"Ya-Chuan Lai","email":"","orcid":"","institution":"National Center for Research on Earthquake Engineering","correspondingAuthor":false,"prefix":"","firstName":"Ya-Chuan","middleName":"","lastName":"Lai","suffix":""},{"id":445646992,"identity":"4be21185-996d-4a0b-9203-5f98dac837cb","order_by":3,"name":"Min-Hung Shih","email":"","orcid":"","institution":"National Center for Research on Earthquake Engineering","correspondingAuthor":false,"prefix":"","firstName":"Min-Hung","middleName":"","lastName":"Shih","suffix":""},{"id":445646993,"identity":"4f77cbc2-1d2b-4fee-9da1-95a7f64eebab","order_by":4,"name":"Yi-Heng Li","email":"","orcid":"","institution":"Industrial Technology Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Yi-Heng","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2024-12-19 08:53:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5675134/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5675134/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-02345-9","type":"published","date":"2025-05-25T15:58:09+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81801255,"identity":"4ac61ed2-fb5b-45e0-af5b-65c6ac2452c0","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":465609,"visible":true,"origin":"","legend":"\u003cp\u003eBackground seismicity (represented by small circles) at the Tatun Volcano Group (TVG) located at the northern tip of Taiwan from 2014 to 2021. The locations of Mt. Chihsin and the Dayoukeng fumarole are indicated by triangles. The administrative district boundary between Taipei and New Taipei City is shown with a thick dashed line.\u003c/p\u003e","description":"","filename":"image1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/a40f3d318ce7ad715d01e670.jpeg"},{"id":81801258,"identity":"aa5274ef-925a-4d6f-a360-d11541da9b8d","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":281742,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal variations of monthly seismicity at the Tatun Volcano Group (TVG) from 2014 to 2021.\u003c/p\u003e","description":"","filename":"image2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/31481e0d6be511e283fc8bb2.jpeg"},{"id":81801261,"identity":"ab0bcdfc-78b5-42da-b440-a8eb54e7fee3","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":122323,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Spatial and (b) temporal variations of larger earthquakes (M \u0026gt; 3) at the TVG from 2014 to 2021.\u003c/p\u003e","description":"","filename":"image3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/d8828e98e8b07d70c6721673.jpeg"},{"id":81801259,"identity":"b9913230-18d4-4dad-aa56-3042024ea355","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":238948,"visible":true,"origin":"","legend":"\u003cp\u003eP-wave velocity perturbations at a depth of 3 km from 2014 to 2021. Two major positive perturbations in 2018 around both Mt. Chihsin and the Dayoukeng fumarole are indicated by H1 and H2. The locations of Mt. Chihsin, the Dayoukeng fumarole, and the Shanchiao fault are marked by yellow triangles, pink triangles, and a dashed line, respectively.\u003c/p\u003e","description":"","filename":"image4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/3f4f783d55bf49bc0e222a67.jpeg"},{"id":81801905,"identity":"061052aa-5e0f-483b-a383-759351cea9d6","added_by":"auto","created_at":"2025-05-02 06:04:13","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":655283,"visible":true,"origin":"","legend":"\u003cp\u003eP-wave velocity differences between 2018 and each of the other years along the A-A’ profile. A representative anomaly of velocity difference is a nearly vertical channel (R1) observed between 2017 and 2018.\u003c/p\u003e","description":"","filename":"image5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/e521ce8b7c0d48031cd9aa4d.jpeg"},{"id":81801263,"identity":"be28e25b-7cb0-401e-aafc-c16555f2a720","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":328712,"visible":true,"origin":"","legend":"\u003cp\u003eP-wave velocity differences between 2017 and 2018 along two depth profiles (A-A’ and B-B’), which are almost perpendicular to each other. A nearly vertical narrow channel, marked as R1 along the A-A’ profile, is clearly identified near the Dayoukeng fumarole.\u003c/p\u003e","description":"","filename":"image6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/6f2f57278832aff5214d6661.jpeg"},{"id":81801268,"identity":"c026e171-fbbc-4c78-b2fe-d2b9a08bd9a2","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":707275,"visible":true,"origin":"","legend":"\u003cp\u003eP-wave velocity differences between 2018 and each of the other years along the C-C’ profile.\u003c/p\u003e","description":"","filename":"image7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/66a3cdfc66b5866b60a3e859.jpeg"},{"id":81801270,"identity":"0b3e8b9d-4ff4-43e6-9f29-d3deaa279c84","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":325452,"visible":true,"origin":"","legend":"\u003cp\u003eP-wave velocity differences between 2017 and 2018 along two depth profiles (C-C’ and D-D’), which are almost perpendicular to each other. A major anomaly, marked as R2, is clearly identified beneath the Mt. Chihsin area at a depth of 3-4 km.\u003c/p\u003e","description":"","filename":"image8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/67b266a89166249277cbe8a8.jpeg"},{"id":81801272,"identity":"9236f6d4-7f34-4f2b-8f21-b24ed1b4e755","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":108663,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic plots showing temporal evolution models of seismicity, inflation pressure (stress), and seismic velocity in and around the seismicity conduit beneath the Dayoukeng fumarole in the TVG. (a) Stable degassing at the fumarole site through the volcanic conduit in 2014, (b) gradual sealing of the top of the volcanic conduit and an increase in inflation pressure within the conduit from 2015 to 2018, and (c) eventual induction of strong volcanic seismicity in and after 2019.\u003c/p\u003e","description":"","filename":"image9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/db408af2f92582e0956961b8.jpeg"},{"id":81801906,"identity":"b1c452a2-fbec-4310-909a-7abed12f55c2","added_by":"auto","created_at":"2025-05-02 06:04:13","extension":"jpeg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":194388,"visible":true,"origin":"","legend":"\u003cp\u003eLocations of seismic stations (black triangles), grids (small plus signs), Dayoukeng (pink triangle), and Mt. Chihsin (yellow triangle) on the topographic map of the Tatun Volcano Group (TVG) in northern Taiwan.\u003c/p\u003e","description":"","filename":"image10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/a7372ec71ac467e7e6d98ed3.jpeg"},{"id":81801264,"identity":"1675cf95-8e05-4faf-8b48-3c43250c3a37","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpg","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":182025,"visible":true,"origin":"","legend":"\u003cp\u003eCheckerboard test showing the reliability of results at different depths based on seismic data collected in 2014.\u003c/p\u003e","description":"","filename":"image11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/85cedb6e98879e26510bfa5c.jpg"},{"id":81801909,"identity":"0d696a7e-d366-467e-8105-345422e35a90","added_by":"auto","created_at":"2025-05-02 06:04:15","extension":"jpeg","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":626278,"visible":true,"origin":"","legend":"\u003cp\u003eCheckerboard tests at the top seven layers based on seismic data from each year between 2014 and 2021.\u003c/p\u003e","description":"","filename":"image12.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/33c58bccb63db74e813944b1.jpeg"},{"id":81801267,"identity":"d4272e68-d01b-4736-99e7-a879b760e877","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":874870,"visible":true,"origin":"","legend":"\u003cp\u003eThe Vp differences of the recovered checkerboards at different depths between 2018 and all other years. Profiles A-A’, B-B’, C-C’ and D-D’ are also plotted.\u003c/p\u003e","description":"","filename":"image13.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/cc08dd209928eaa4340113bc.jpeg"},{"id":81802283,"identity":"74b90224-2d2a-4f4c-87a8-1312b3c48b1c","added_by":"auto","created_at":"2025-05-02 06:12:13","extension":"jpeg","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":436284,"visible":true,"origin":"","legend":"\u003cp\u003eThe Vp differences of the recovered checkerboards between 2018 and all other years along Profile A-A’.\u003c/p\u003e","description":"","filename":"image14.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/a26afcbac0765029195b3dc5.jpeg"},{"id":81801269,"identity":"665974a0-49f9-479a-a647-737653bf0e3f","added_by":"auto","created_at":"2025-05-02 05:40:13","extension":"jpeg","order_by":15,"title":"Figure 15","display":"","copyAsset":false,"role":"figure","size":389751,"visible":true,"origin":"","legend":"\u003cp\u003eThe Vp differences of the recovered checkerboards between 2018 and all other years along Profile C-C’.\u003c/p\u003e","description":"","filename":"image15.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/36a7fdf9ccebe3c3adcd2ead.jpeg"},{"id":83460666,"identity":"c3d00572-0696-43d7-aa0b-3f89c42c2999","added_by":"auto","created_at":"2025-05-26 16:13:26","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6537018,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/d37c8352-6e67-42b3-9fa0-cbbb689c0621.pdf"},{"id":81801904,"identity":"c9029045-adfa-4fe2-8076-35c8e939ae33","added_by":"auto","created_at":"2025-05-02 06:04:13","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2108826,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix.docx","url":"https://assets-eu.researchsquare.com/files/rs-5675134/v1/c7314fb571d5e06d3fe64098.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"4D seismic tomography unveiling velocity increase prior to seismic swarms at the Tatun volcano group of Taiwan","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe Tatun Volcano Group (TVG) is located at the northern tip of Taiwan (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) and is directly linked to the Taipei metropolitan area, home to over 7\u0026nbsp;million residents across the cities of Taipei and New Taipei. The horizontal distance from Mt. Chihsin, the highest peak in the TVG and likely the site of its most recent eruption\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, to the Taipei 101 skyscraper in downtown Taipei is less than 15 km. In fact, the northernmost parts of Taipei City, including the Beitou and Shilin districts, partially overlap with the TVG. As a result, Taipei City, similar to Naples in Italy, Auckland in New Zealand, and Cartago in Costa Rica, can be considered one of the major cities located on an active volcano, where a volcanic eruption could potentially cause widespread loss of life and regional disruption\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. One of the most well-known cities affected by a volcanic eruption is ancient Pompeii in Italy, which was completely buried by the eruption of Mt. Vesuvius in 79 A.D.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. In fact, the potential global economic impact of a future eruption of the TVG has been seriously considered\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAlthough there is no recorded volcanic eruption in the TVG throughout human history, recent geological and seismic observations suggest that it may still be active. The most significant evidence from these recent observations includes the dating of volcanic ash layers\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e and seismic detections of a magma reservoir\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. First, radiocarbon dating of various volcaniclastic deposits in the TVG suggests that the last eruption occurred approximately 6,000 years ago at Mt. Chihsin\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Besides, a later study by Zellmer et al.\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e suggests that the last eruption may have occurred as recently as 1,350 years ago. Both studies consistently indicate that the last eruption occurred less than 10,000 years ago. Second, a magma reservoir at mid-crust depth was first detected through observations of both S-wave shadowing and P-wave delays recorded by a dense seismic array in the TVG\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. This magma reservoir was further confirmed by tomographic images showing low-velocity zones\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e and strong seismic attenuation\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e following the successful deployment of a dense seismic array in northern Taiwan in 2018\u003csup\u003e9\u0026ndash;11\u003c/sup\u003e. Taken together, the dating results and seismic observations strongly suggest that the TVG is an active volcano according to established criteria\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Therefore, disaster prevention and mitigation measures for potential volcanic hazards are critical issues for the Taipei metropolitan area.\u003c/p\u003e\n\u003ch3\u003eSeismicity at the TVG\u003c/h3\u003e\n\u003cp\u003eSince 2003, a dense seismic array of 40 broadband stations has been gradually deployed to monitor volcanic activity in the TVG\u003csup\u003e\u003cspan additionalcitationids=\"CR14 CR15\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Many micro-earthquakes have been detected by the Taiwan Volcano Observatory at Tatun. Volcanic seismicity is generally concentrated around the administrative boundary between Taipei and New Taipei cities, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Several key studies have examined the volcanic activity in the TVG, including reports of significant seismic events\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan additionalcitationids=\"CR17 CR18 CR19 CR20\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Notably, between 2014 and 2018, more than 1,000 earthquakes occurred at depths of 0 to 2 km beneath the Dayoukeng fumarole, forming a distinct seismic conduit\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Additionally, several \"heartbeat-like\" earthquake sequences have been observed in the same region\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. These seismic phenomena strongly suggest ongoing degassing processes beneath the Dayoukeng fumarole\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. The presence of both the seismic conduit and the heartbeat-like sequences indicates that the area is a likely pathway for future volcanic eruptions.\u003c/p\u003e \u003cp\u003eAn analysis of the temporal variation in background seismicity from 2014 to 2021 reveals that the number of earthquakes typically ranged from 100 to 300 per month. However, seismic swarms occasionally occurred, with several hundred to over a thousand earthquakes occurring over just a few days or within a single month (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Based on the slope of the accumulated earthquake count, the background seismicity can be roughly divided into three periods: (1) 2014 to early 2015, (2) early 2015 to 2018, and (3) 2019 to 2021. On average, the number of earthquakes per month began at a relatively low level of ~\u0026thinsp;115 events during the first period, increased to ~\u0026thinsp;198 events during the second period, and surged to ~\u0026thinsp;300 events in the final period. Notably, there was a significant spike in the total accumulated events in 2019, primarily driven by a major seismic event. This jump resulted from more than 1,400 earthquakes triggered by a magnitude 4.2 earthquake on January 28, 2019.\u003c/p\u003e \u003cp\u003eIn addition to the significant increase in the total number of volcanic earthquakes observed from 2019 onward, larger earthquakes (M\u0026thinsp;\u0026gt;\u0026thinsp;3) occurred more frequently during the third period compared to previous years (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In 2014, only one such earthquake was recorded, and there were none from 2015 to 2018. However, it is noteworthy that more than 20 larger earthquakes (M\u0026thinsp;\u0026gt;\u0026thinsp;3) were detected between 2019 and 2021 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Interestingly, most of these larger earthquakes (M\u0026thinsp;\u0026gt;\u0026thinsp;3) were concentrated around the Dayoukeng and Mt. Chihsin areas (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), which are two major degassing sites in the Tatun Volcanic Group (TVG)\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Among these events, one of the largest earthquakes, with a magnitude of 4.2, occurred near Dayoukeng on January 28, 2019. Additionally, two significant earthquakes with magnitudes of 3.6 and 3.7 took place near Mt. Chihsin on February 9 and August 6, 2019, respectively. These dramatic increases in volcanic seismicity around the Dayoukeng fumarole and Mt. Chihsin strongly suggest that temporal changes in physical or chemical conditions within the shallow crust beneath the TVG may have occurred.\u003c/p\u003e \u003cp\u003eTo investigate potential temporal changes in the shallow crustal velocity structures beneath the TVG (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), we conducted a 4D seismic tomography analysis\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e using the extensive seismic data recorded at the TVG between 2014 and 2021. We carefully selected non-redundant earthquake data to ensure reliable tomographic images for each year. We then calculated the differences between the velocity structures obtained in 2018 and those from other years to assess the dramatic increase in volcanic earthquakes observed in the TVG. By examining the strong correlations between velocity changes and seismicity variations, both temporally and spatially, we propose that temporal velocity variations in the TVG could serve as a potential warning signal for seismic swarms or volcanic activity.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eVelocity structures at a depth of 3 km\u003c/h2\u003e \u003cp\u003eThe results from the 4D seismic tomography reveal significant perturbations in P-wave seismic velocity (Vp) in 2018. For discussing the reliable inverted results in the Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, we only plotted the derivative weight sum (DWS\u003csup\u003e\u003cspan additionalcitationids=\"CR23 CR24\" citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e) of the velocity node is higher than 5, which means the ray density at the velocity node is five times higher than the average ray density during the inversion. One of the most notable layers of perturbations occurs at a depth of 3 km, with data from 2014 to 2021 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). In general, lateral velocity perturbations during the first five years (2014\u0026ndash;2018) are more pronounced than those in the last three years (2019\u0026ndash;2021). Among these, the strongest lateral perturbations were observed in 2018, particularly around the Dayoukeng fumarole (marked by a pink triangle) and Mt. Chihsin (marked by a yellow triangle). Beneath the Dayoukeng fumarole, a significant positive velocity perturbation (H1, marked in blue) is clearly identifiable. This major positive anomaly is surrounded by a zone of strong negative velocity perturbations (marked by a dashed circle). Similarly, around Mt. Chihsin, another prominent positive velocity perturbation (H2) is encircled by negative perturbations (marked by another dashed circle). To further highlight the intensity of the velocity perturbations in the Dayoukeng and Mt. Chihsin areas during 2018, we present velocity differences between 2018 and the other years in four depth profiles below.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eVelocity anomaly beneath Dayoukeng\u003c/h3\u003e\n\u003cp\u003eTo examine the strong perturbations beneath the Dayoukeng fumarole, we selected reliable P-wave velocity structures, validated by the checkerboard test, along two vertical profiles (Profiles A-A\u0026rsquo; and B-B\u0026rsquo;) that intersect near Dayoukeng (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). These two profiles are nearly perpendicular to each other. First, we plotted the P-wave velocity differences between 2018 and each of the other years along Profile A-A\u0026rsquo;. Notably, the most prominent and consistent anomaly across most profiles is a nearly vertical channel characterized by negative velocity differences extending from sea level down to a depth of approximately 3 to 4 km beneath the Dayoukeng fumarole. One of the most representative profiles displaying this nearly vertical channel (R1) is shown in Profile A-A\u0026rsquo; for the velocity difference between 2017 and 2018. Interestingly, similar negative anomalies (R1) beneath the Dayoukeng fumarole are also observed along Profile BB\u0026rsquo; in the other years (Fig. A1). To further emphasize the significant change in P-wave velocities within the nearly vertical channel beneath the Dayoukeng fumarole (R1), we plotted the velocity differences between 2017 and 2018 along both Profiles A-A\u0026rsquo; and B-B\u0026rsquo; (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Overall, both profiles indicate that the P-wave velocities within the nearly vertical channel (R1) in 2018 are consistently higher compared to the other years.\u003c/p\u003e \n\u003ch3\u003eVelocity anomaly beneath Mt. Chihsin\u003c/h3\u003e\n\u003cp\u003eTo examine the strong perturbations beneath Mt. Chihsin, we analyzed two additional vertical profiles (Profiles C-C\u0026rsquo; and D-D\u0026rsquo;) that intersect at Mt. Chihsin and are nearly perpendicular to each other. For Profile C-C\u0026rsquo; (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e), a major negative anomaly (R2) is consistently observed at depths of 3\u0026ndash;4 km beneath Mt. Chihsin in all years, despite the presence of other individual anomalies in different years. The most representative anomaly (R2) is clearly visible in the velocity difference between 2017 and 2018. These negative anomalies are not only observed along Profile C-C\u0026rsquo; (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e), but also along Profile D-D\u0026rsquo; (Fig. A2). For example, Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e shows that the representative anomaly beneath Mt. Chihsin is clearly identified along both Profiles C-C\u0026rsquo; and D-D\u0026rsquo;. Overall, the negative anomalies consistently observed along both profiles indicate a significant change in P-wave velocities at depths of 3\u0026ndash;4 km beneath Mt. Chihsin between 2018 and the other years. Specifically, the velocities at R2 in 2018 are significantly higher than those in the other years.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIt is particularly intriguing to note that the dramatic increase in seismicity observed in the TVG in 2019 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) occurred shortly after a significant increase in P-wave seismic velocity beneath both the Dayoukeng fumarole and Mt. Chihsin areas in 2018 (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e–\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). On one hand, the sharp rise in seismic activity in 2019 is evident in both the total number of volcanic earthquakes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) and the occurrence of larger earthquakes (M \u0026gt; 3.0) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). This marked increase in seismicity was the strongest observed in the TVG since the deployment of a dense seismic array in 2003\u003csup\u003e13\u003c/sup\u003e. On the other hand, the notable increase in P-wave seismic velocity beneath both the Dayoukeng and Mt. Chihsin areas in 2018 occurred just prior to the surge in seismic activity in 2019 (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e–\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). This temporal sequence of changes in velocity structures and seismicity suggests a potentially strong causal relationship between the two phenomena.\u003c/p\u003e \u003cp\u003eThe significant alterations in Vp structures, along with the clustering of larger earthquakes (M \u0026gt; 3), beneath both the Dayoukeng fumarole and Mt. Chihsin, are generally consistent with findings from previous studies\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Notably, background seismicity in the TVG often clusters around these two areas\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Beneath Mt. Chihsin, earthquakes primarily occur within a sphere-like zone at depths ranging from 0.5 km to 2.0 km. Interestingly, earthquake clustering beneath the Dayoukeng fumarole appears to be confined to a nearly vertical conduit, approximately 500 meters in diameter, extending from sea level down to 2 km depth\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Detailed stress analyses of focal mechanisms from numerous micro-earthquakes, coupled with surface observations of the sulfur-to-carbon dioxide ratio, suggest that the seismic conduit beneath Dayoukeng may gradually seal with smectite-rich rocks at its upper levels, leading to the accumulation of inflationary pressure over time\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. In other words, both the sphere-like and conduit seismic zones are likely influenced by the ascent of volcanic fluids from a deep magma reservoir. Periodically, these zones may seal at the top, causing inflationary pressure to build within the seismic zones\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe sequential changes in seismic velocities in 2018 and seismicity in 2019 may suggest a plausible link, wherein the increase in P-wave seismic velocity (Vp) over the years could result from higher pressure in the rocks or sediments beneath the TVG. While Vp can be influenced by various factors, including temperature, pressure, chemical composition, and water content\u003csup\u003e\u003cspan additionalcitationids=\"CR27 CR28 CR29 CR30 CR31\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e–\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, the most likely explanation for the Vp increase in this region is an increase in rock confining pressure, as other factors appear less likely to account for the change. Previous studies and laboratory experiments consistently show that Vp increases with confining pressure in rocks\u003csup\u003e\u003cspan additionalcitationids=\"CR28 CR29 CR30\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e–\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. In fact, increased pressure has already been observed in the TVG, as indicated by spatial-temporal variations in the focal mechanisms of numerous micro-earthquakes\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e and precise leveling surveys\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. In contrast, increases in temperature or water content typically lead to a reduction in Vp\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e–\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Additionally, changes in seismic velocity occurring over the course of months or even years are unlikely to result from chemical variations in the rocks, as such changes would generally require more time to manifest.\u003c/p\u003e \u003cp\u003eThe processes underlying the sequential changes in seismic velocity and earthquake swarms associated with confining pressure along major volcanic pathways in the TVG can be divided into three distinct periods from 2014 to 2021. The major pathway beneath the Dayoukeng fumarole is well delineated as a nearly vertical conduit\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. A simplified conceptual model representing the changes in seismicity, inflation pressure, and seismic velocity over these three periods is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePeriod 1 (2014)\u003c/strong\u003e \u003c/p\u003e\u003cp\u003eDuring this period, seismicity is primarily driven by stable volcanic degassing beneath the fumarole site, following the smooth volcanic pathway. Earthquake activity is largely confined to clustering within the pathway itself\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePeriod 2 (2015–2018)\u003c/strong\u003e \u003c/p\u003e\u003cp\u003eOver this period, the top of the major volcanic degassing pathway begins to gradually seal, leading to an increase in confining pressure within the pathway. It may be like a pressure cooker on a flame. Consequently, a dramatic rise in seismic velocity is observed in 2018 (Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). In addition to the continued clustering of seismicity within the pathway, some earthquakes occur in the surrounding area, likely due to the inflationary stress exerted by the sealing of the pathway.\u003c/p\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePeriod 3 (2019 and beyond)\u003c/strong\u003e \u003c/p\u003e\u003cp\u003eBeginning in 2019, strong seismic swarms are unexpectedly observed in and around the pathway. The significant dynamic stress from larger earthquakes (M \u0026gt; 3) likely breaks the seal at the top of the pathway. It is just similar to a pressure cooker explosion. Following this, seismicity, seismic velocity, and confining pressure may return to a temporarily stable state, similar to the conditions in Period 1, until the sealing of the volcanic degassing pathway occurs again.\u003c/p\u003e \u003cp\u003e\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe sequential changes in both seismicity and seismic velocity around the two major volcanic degassing pathways—Mt. Chihsin and Dayoukeng fumarole—suggest that monitoring temporal variations in seismic velocity could provide a valuable early warning system for seismic swarms or volcanic activity, as illustrated in the conceptual model in Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e. First, major degassing zones such as Mt. Chihsin and Dayoukeng fumarole can be clearly identified from clustering seismic events, which reflect the spatial distribution of seismicity\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Second, the migration of volcanic fluids or pressure variations along the primary pathways can cause transient changes in seismic clustering\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. These variations in seismicity can be routinely monitored through local earthquake detection systems operated by the Taiwan Volcano Observatory at Tatun. Third, temporal variations in seismic velocity can be tracked using 4D seismic tomography, provided that sufficient seismic data is collected within a given time window. By comparing these temporal variations in seismicity and velocity structures, it may be possible to issue early warnings for strong seismic swarms or potential volcanic activity in the future, based on the detailed analysis proposed in the model in Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e.\u003c/p\u003e "},{"header":"Method and Data Selection","content":"\u003cp\u003eTo obtain 3-D velocity structures that change over time beneath the TVG, we applied the double-difference tomography method\u003csup\u003e22\u003c/sup\u003e to seismic data collected by the Taiwan Volcano Observatory at Tatun from 2014 to 2021 (Fig. 1). During the tomographic inversion process, various parameter values were tested based on recommendations from Zhang and Thurber\u003csup\u003e22\u003c/sup\u003e. A detailed description of the inversion procedure can be found in one of our previous studies\u003csup\u003e16\u003c/sup\u003e. The final inversion parameters used in this study are as follows: the number of iterations was 18, and the damping factor for both velocity structure inversion and earthquake relocation was set to 125, which was optimized after trying different values ranging from 25 to 200 (Fig. A3). Such an optimal result is based on the trade-off between the damping factor and CND (Condition Number) of ~100, which is the ratio of the largest to smallest eigenvalue\u003csup\u003e22\u003c/sup\u003e. For constraining the order smooth model, we calculated the recovery percentage of the checkerboard model under a series of smoothing constraints, ranging from 0.5 to 5.0 in increments of 0.5 (Fig. A4). We ultimately selected a smoothing constraint of 1.5, as the improvement in recovery percentage became marginal for smoothing constraints greater than 1.5. The weighting ratio between absolute and differential arrival times was reduced from 100 to 0.01. The large number of relative arrival times for both P- and S-waves, approximately four times the number of absolute arrivals, significantly minimized systematic errors, thus improving the TVG velocity model.\u003c/p\u003e\n\u003cp\u003eThe seismic data used in this study were recorded by a dense network of 40 seismic stations in the TVG\u003csup\u003e6\u003c/sup\u003e (Fig. 10). Each station is equipped with a three-component broadband seismometer (Guralp CMG-6TD or Maredian Compact). Seismic data are continuously recorded at a sampling rate of 100 Hz and transmitted in near-real-time to the Taiwan Volcano Observatory at Tatun and the Institute of Earth Sciences, Academia Sinica, Taiwan, via telephone lines or wireless radio. The arrival times of both the first P- and S-waves generated by local earthquakes are automatically picked by the software and later manually verified to determine the earthquake parameters, including hypocenter and magnitude. In total, 22,290 local earthquakes were detected between 2014 and 2021 (Fig. 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAmong the 22,290 earthquakes recorded at 40 seismic stations in the TVG between 2014 and 2021, we carefully selected 14,221 events to avoid redundancy in a model grid during the tomographic inversion. The dense gridding velocity model has been designed for an area of approximately 100 km\u0026sup2; to cover the TVG (Fig. 10). The horizontal grid spacing is evenly set to 0.25 km, and the vertical grid is starting 0.5 km above sea level and increasing by 0.5 km increments down to a depth of 5.0 km. The denser grids at depths less than 5 km are designed to accommodate the high density of local events observed in the uppermost crust (Fig. 1). The velocity values at each grid node were slightly adjusted from the previous study by one of the previous studies\u003csup\u003e16\u003c/sup\u003e, who successfully inverted 3D seismic images based on earlier seismic data. For each grid\u0026nbsp;cell, with a volume of 0.25 km \u0026times; 0.25 km \u0026times; 0.5 km (Fig. 10), only two earthquakes with the smallest residuals and location errors were selected for the inversion. Thus, the model distributions are expected to be same between the different time steps\u003csup\u003e34\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eSince seismicity is not evenly distributed across the study area (Fig. 1), many redundant arrival times had to be adjusted for the tomographic inversion. In regions with dense clustering of earthquakes, such as around Mt. Chihsin and the Dayoukeng fumarole, we applied stricter criteria to select the most reliable seismic data. These criteria included not only smaller uncertainties in earthquake locations but also a greater number of seismic stations recording each event. In contrast, we aimed to retain all arrival times for ray paths passing through regions where seismic ray coverage was more limited. This approach ensures that the tomographic images are more evenly reliable across the entire study area.\u003c/p\u003e\n\u003cp\u003eTo examine potential temporal variations in the 3D velocity structures, the seismic data were divided into eight sets, one for each year from 2014 to 2021. To maintain consistency in the inversion process, the number of earthquakes selected for each year was roughly comparable (Table 1), ensuring that the ray-path coverage in each year provided similar resolution for the seismic images. This approach allows us to derive annual tomographic images and analyze the temporal variations in the 3D velocity structures.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e Total number of earthquakes selected for tomographic inversion in each year.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2014\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2015\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2016\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2017\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2018\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2021\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003eNo.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e1428\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e1519\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e1913\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e1865\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e1230\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2309\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e2326\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 55px;\"\u003e\n \u003cp\u003e1631\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo assess the effectiveness of the inversion, a checkerboard test was conducted using the following procedure. First, a checkerboard velocity model with velocity perturbations of \u0026plusmn;10% and a grid spacing of 1.0 km was applied to create a simulated dataset. Second, simulated arrivals of both P- and S-waves, generated from actual seismic events and stations in the TVG, were used for the double-difference tomographic inversion. Finally, a comparison between the checkerboard model and the inverted results showed that the P-wave velocity structures (Vp) were successfully recovered for most of the study area at depths shallower than 3.5 km (Fig. 11). This indicates that the inverted velocity structures in the shallow crust are highly reliable, although the area with reliable results gradually shrinks with depth due to the limited availability of deeper earthquakes. Besides, the results of the checkerboard tests from 2014 to 2021 for the top five layers are shown Fig. 12. The inverted images from the checkerboard velocity model were clear and consistent across most of the study area each year.\u0026nbsp;For instance, the differences in checkerboard tests between years were minimal in most areas shallower than 2 km in depth (Fig. 13).\u0026nbsp;The velocity perturbations (\u0026gt; 1.0 km/s) at the areas where we have discussed at Figures 5-8 are significantly greater than the residuals around \u0026plusmn; 0.25 to 0.5 km/s observed by differencing checkerboards (Figs. 14 and 15). It indicates that the major velocity changes between time steps from 2014 to 2021 can be well resolved. To further demonstrate the resolution similarity between the different time steps, the values of the derivative weight sum (DWS\u003csup\u003e22\u003c/sup\u003e) for the model space in each year have been plotted for showing the seismic ray-path density is comparable (Fig. A5). It concludes that the inverted results are not only reliable, but also almost consistent year to year. Therefore, the temporal variations in the inverted results are meaningful and warrant further discussion.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eY.C., M.H. and Y. H. collected seismic data and prepared some figures; H. C. analyzed the data and joined the writing as well as detailed discussion; C. H. analyzed the data and wrote the manuscript text.All authors reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgments:\u003c/h2\u003e \u003cp\u003eWe would like to express our gratitude to our colleagues at the Institute of Earth Sciences, Academia Sinica as well as the Taiwan Volcano Observatory at Tatun, National Centre for Research on Earthquake Engineering in Taipei, Taiwan for their efforts in deploying and maintaining the seismic arrays since 2013. This work was funded by the National Science and Technology Council of Taiwan. We are appreciated at the valuable comments and suggestions by two anonymous reviewers to improve this work.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eSeismic data used in this study have been collected and carefully analyzed by the Taiwan Volcano Observatory at Tatun (TVO) in Taiwan. Due to their size and data regulation, completely seismic records can be made available upon reasonable request from the corresponding author (C.H. Lin). Some plots were made using the Generic Mapping Tools (version 4.5.2; URL:gmt.soest.hawaii.edu) and the SAC software (version 101.5; URL:ds.iris.edu).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBelousov, A., Belousova, M., Chen, C. H. \u0026amp; Zellmer, G. F. Deposits character and timing of recent eruptions and gravitational collapses in Tatun volcanic group, northern Taiwan: hazard-related issues.\u003cem\u003e J. Volcanol. Geotherm. Res.\u003c/em\u003e \u003cstrong\u003e191\u003c/strong\u003e, 205-221 (2010).\u003c/li\u003e\n\u003cli\u003eZellmer, G. F., Rubin, K. H., Miller, C. A., Belousov, A. \u0026amp; Belousova, M. Resolving discordant U\u0026ndash;Th\u0026ndash;Ra ages: constraints on petrogenetic processes of recent effusive eruptions at Tatun Volcano Group, northern Taiwan. \u003cem\u003eGeological Society \u003c/em\u003e\u003cstrong\u003e422\u003c/strong\u003e (2015).\u003c/li\u003e\n\u003cli\u003eChester, D. K., Degg, J. M., Duncan, A. M. \u0026amp; Guest, J. E. The increasing exposure of cities to the effects of volcanic eruptions: a global survey, Global Environmental Change Part B, \u003cem\u003eEnvironmental Hazards\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e(3), 89-103 (2000). \u003c/li\u003e\n\u003cli\u003eGiacomelli, L., Perrotta, A., Scandone, R. \u0026amp; Scarpati, C. The eruption of Vesuvius of 79 AD and its impact on human environment in Pompeii, \u003cem\u003eEpisodes\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 235-238 (2003).\u003c/li\u003e\n\u003cli\u003eMani, L., Tzachor, A. \u0026amp; Cole, P. Global catastrophic risk from lower magnitude volcanic eruptions. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 4756 (2021).\u003c/li\u003e\n\u003cli\u003eLin, C. H.Evidence for a magma reservoir beneath the Taipei metropolis of Taiwan from both S-wave shadows and P-wave delays, \u003cem\u003eScientific Reports\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 39500 (2016).\u003c/li\u003e\n\u003cli\u003eHuang, H. H. et al. Unveiling Tatun volcanic plumbing structure induced by post-collisional extension of Taiwan mountain belt, \u003cem\u003eScientific Reports\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 5286 (2021).\u003c/li\u003e\n\u003cli\u003eLin, Y. P., Ko, J. Y. T., Huang, B. S., Lin, C. H., \u0026amp; Shih, M. H. Unveiling attenuation structures in the northern Taiwan volcanic zone, \u003cem\u003eScientific Reports \u003c/em\u003e\u003cstrong\u003e14\u003c/strong\u003e\u003cem\u003e, 4716 \u003c/em\u003e(2024).\u003c/li\u003e\n\u003cli\u003eLin, C. H., Shih, M. H., \u0026amp; Lai, Y. C. A strong seismic reflector in the mantle wedge above the Ryukyu subduction system beneath northern Taiwan, \u003cem\u003eSeism. Res. Lett., \u003c/em\u003e\u003cstrong\u003e91\u003c/strong\u003e(1), 310-316 (2020). \u003c/li\u003e\n\u003cli\u003eLin, C. H., Shih, M. H., \u0026amp; Lai Y. C. Mantle wedge diapirs detected by a dense seismic array in northern Taiwan, \u003cem\u003eScientific Reports\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1561 (2021).\u003c/li\u003e\n\u003cli\u003eLiu, C. N. et al. High resolution 3-D shear velocity model of northern Taiwan via Bayesian joint inversion of Rayleigh wave ellipticity and phase velocity with Formosa Array, \u003cem\u003eJ. Geophy. Res: Solid Earth\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e(5), e2020JB21610 (2021).\u003c/li\u003e\n\u003cli\u003eSzakacs, A. Redefining active volcanoes: A discussion, \u003cem\u003eBull. Vol- canol.\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 321\u0026ndash;325 (1994). \u003c/li\u003e\n\u003cli\u003eLin, C. H. et al. Preliminary analysis of tectonic earthquakes and volcanoseismic signals recorded at the Tatun volcanic group, northern Taiwan, \u003cem\u003eGeophysical Research Letters\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e(10), L10313. (2005a).\u003c/li\u003e\n\u003cli\u003eLin, C. H. et al. Preliminary results of seismic monitoring at Tatun volcanic area of northern Taiwan. \u003cem\u003eTerr. Atm. Ocean\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 563-577 (2005b).\u003c/li\u003e\n\u003cli\u003ePu, H.C. et al. Volcano-hydrothermal inflation revealed through spatial variation in stress field in Tatun volcano group, northern Taiwan, \u003cem\u003eJ. Volcanology Geothermal Research\u003c/em\u003e \u003cstrong\u003e390\u003c/strong\u003e, 160712 (2020a).\u003c/li\u003e\n\u003cli\u003ePu, H. C. et al. Active volcanism revealed from a clustering seismicity conduit beneath the long-resting Tatun volcano group in northern Taiwan, \u003cem\u003eScientific Reports\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 6153 (2020b).\u003c/li\u003e\n\u003cli\u003ePu, H. C. et al. The volcanic earthquake swarm of October 20, 2009 in the Tatun Area of Northern Taiwan. \u003cem\u003eTerr. Atmos. Ocean.\u003c/em\u003e\u003cstrong\u003e 25\u003c/strong\u003e, 625-635 (2014).\u003c/li\u003e\n\u003cli\u003eLin, C. H. \u0026amp; Pu,H. C. Very-Long-Period Seismic Signals at the Tatun Volcano Group, Northern Taiwan, \u003cem\u003eJournal of Volcanology and Geothermal Research\u003c/em\u003e, \u003cstrong\u003e328\u003c/strong\u003e, 230\u0026ndash;236 (2016).\u003c/li\u003e\n\u003cli\u003eLin, C. H. Probable dynamic triggering of phreatic eruption at the Tatun volcano group of Taiwan, \u003cem\u003eJournal of Asian Earth Sciences\u003c/em\u003e \u003cstrong\u003e149\u003c/strong\u003e, 78-85 (2017a).\u003c/li\u003e\n\u003cli\u003eLin, C. H. Dynamic triggering of drumbeat seismicity at Tatun volcano group in Taiwan, \u003cem\u003eGeophysical Journal International\u003c/em\u003e \u003cstrong\u003e210\u003c/strong\u003e, 354-359 (2017b).\u003c/li\u003e\n\u003cli\u003ePu, H. C. et al. Ascending volcanic fluids portended by spatiotemporal variations of the earthquake mechanisms in the Tatun volcano group in northern Taiwan, \u003cem\u003eGeophy. Res. Letters\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, e2020GL091686 (2021).\u003c/li\u003e\n\u003cli\u003eZhang, H. J. \u0026amp; Thurber, C. H. Double-difference tomography: The method and its application to the Hayward fault, California, \u003cem\u003eBull. Seismo. Soc. Am.\u003c/em\u003e \u003cstrong\u003e93\u003c/strong\u003e, 1875-1889 (2003).\u003c/li\u003e\n\u003cli\u003eThurber, C. H. Earthquake locations and three-dimensional crustal structure in the Coyote Lake area, central California,\u003cem\u003e J. Geophys. Res.\u003c/em\u003e \u003cstrong\u003e88\u003c/strong\u003e, 8226-8236 (1983).\u003c/li\u003e\n\u003cli\u003eThurber, C. H. \u0026amp; Eberhart-Phillips, D. Local earthquake tomography with flexible gridding, \u003cem\u003eComputers and Geosciences\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 809-818 (1999).\u003c/li\u003e\n\u003cli\u003eWaldhauser, F., \u0026amp; Ellsworth, W.L. A double-difference earthquake location algorithm: method and application to the northern Hayward Fault, California, \u003cem\u003eBull. Seismo. Soc. Am. \u003c/em\u003e\u003cstrong\u003e90\u003c/strong\u003e, 1353\u0026ndash;1368 (2000).\u003c/li\u003e\n\u003cli\u003eJaya, M.S. et al. Temperature dependence of seismic properties in geothermal rocks at reservoir conditions, \u003cem\u003eGeothermics\u003c/em\u003e\u003cstrong\u003e 39\u003c/strong\u003e, 115\u0026ndash;123 (2010).\u003c/li\u003e\n\u003cli\u003eGao, K. \u0026amp; Gibson Jr., R. L. Pressure-dependent seismic velocities based on effective compliance theory and an asperity deformation model, \u003cem\u003eGeophysics\u003c/em\u003e \u003cstrong\u003e77\u003c/strong\u003e(6), 229\u0026ndash;243 (2012).\u003c/li\u003e\n\u003cli\u003eEberhart‐Phillips, D., Han, D-H. \u0026amp; Zoback, M. D. Empirical relationships among seismic velocity, effective pressure, porosity, and clay content in sandstone, \u003cem\u003eGeophysics\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e (1), 82-89 (1989).\u003c/li\u003e\n\u003cli\u003eMukeri, T. \u0026amp; Mavko, G. Pore fluid effects on seismic velocity in anisotropic rocks, \u003cem\u003eGeophysics\u003c/em\u003e \u003cstrong\u003e59\u003c/strong\u003e(2), 233-244 (1994).\u003c/li\u003e\n\u003cli\u003eKiddle, E. J. The Structure of the Crust and Magmatic System at Montserrat, Lesser Antilles. Ph.D. Thesis, University of Bristol, Bristol, UK (2011).\u003c/li\u003e\n\u003cli\u003eKristinsd\u0026oacute;ttir, L. et al. Electrical conductivity and P-wave velocity in rock samples from high-temperature Icelandic geothermal fields. \u003cem\u003eGeothermics\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 94\u0026ndash;105 (2010).\u003c/li\u003e\n\u003cli\u003eBoitnott, G. N. Core Analysis for the Development and Constraint of Physical Models of Geothermal Reservoirs; DE-FG07\u0026ndash;99ID13761; New England Research, Inc.: White River Junction, VT, USA, 52. (2003).\u003c/li\u003e\n\u003cli\u003eMurase M., Lin, C. H., Kimata, F., Mori, H. \u0026amp; Pu, H. C. Volcano-hydrothermal activity detected by precise levelling surveys at the Tatun volcano group in northern Taiwan during 2006-2013. \u003cem\u003eJ. Volcanol. Geotherm. Res. \u003c/em\u003e\u003cstrong\u003e286\u003c/strong\u003e, 30-40 (2014).\u003c/li\u003e\n\u003cli\u003ePatane, D., Barberi, G., Cocina, O., De Gori, O., Chiarabba, C. Time-resolved seismic tomography detects magma intrusions at Mount Etna,\u003cem\u003e Science\u003c/em\u003e \u003cstrong\u003e313\u003c/strong\u003e, 5788 (2006).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5675134/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5675134/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe Tatun Volcano Group (TVG) is in close proximity to the Taipei metropolitan area, which is home to over 7\u0026nbsp;million residents. Although there has been no recorded volcanic eruption in the TVG throughout human history, recent seismic observations suggest that it may still be active. To investigate possible volcanic seismicity and activity in the TVG, we conducted a 4-dimensional seismic tomography study using abundant seismic data collected from 2014 to 2021. We obtained 3D seismic velocity structures to examine both the temporal and spatial variations in seismicity each year. Our results show that the dramatic increase in seismicity in 2019 followed an increase in P-wave seismic velocity in the Dayoukeng fumarole area and Mt. Chihsin in 2018. This increase in seismic velocity may have been caused by rocks or sediments subjected to higher pressure beneath the TVG, resembling a pressure cooker on a flame. Thus, the sequential rise in both seismic velocity and seismicity strongly suggests that careful monitoring of temporal velocity variations in the volcanic area might provide an early warning of potential seismic swarms or volcanic activity in the future.\u003c/p\u003e","manuscriptTitle":"4D seismic tomography unveiling velocity increase prior to seismic swarms at the Tatun volcano group of Taiwan","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-02 05:40:08","doi":"10.21203/rs.3.rs-5675134/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Accepted","date":"2025-05-13T05:22:22+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-12T18:04:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"50772046181770493816434177753104741912","date":"2025-05-12T17:57:20+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-05-07T01:59:14+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-05-02T15:39:27+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-04-21T08:40:24+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0dd16c53-4097-4e03-9bc6-4097b85aa0d2","owner":[],"postedDate":"May 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47438291,"name":"Earth and environmental sciences/Natural hazards"},{"id":47438292,"name":"Earth and environmental sciences/Solid earth sciences"},{"id":47438293,"name":"Earth and environmental sciences/Solid earth sciences/Seismology"},{"id":47438294,"name":"Earth and environmental sciences/Solid earth sciences/Volcanology"}],"tags":[],"updatedAt":"2025-05-26T16:09:56+00:00","versionOfRecord":{"articleIdentity":"rs-5675134","link":"https://doi.org/10.1038/s41598-025-02345-9","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-05-25 15:58:09","publishedOnDateReadable":"May 25th, 2025"},"versionCreatedAt":"2025-05-02 05:40:08","video":"","vorDoi":"10.1038/s41598-025-02345-9","vorDoiUrl":"https://doi.org/10.1038/s41598-025-02345-9","workflowStages":[]},"version":"v1","identity":"rs-5675134","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5675134","identity":"rs-5675134","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}