Broadband Array In Taiwan For Seismology Research Articles

Controls on Seasonal Variations of Crustal Seismic Velocity in Taiwan Using Single‐Station Cross‐Component Analysis of Ambient Noise Interferometry

AbstractAmbient noise interferometry is a powerful technique to continuously measuring crustal seismic velocity changes (dv/v) and studying crustal behaviors over time. However, the interpretation of such dv/v variations is not straightforward since multiple causes including internal (tectonic/magmatic) processes of the crust and external (environmental) factors could both affect dv/v simultaneously. To differentiate the interplay between the internal and external processes in dv/v variations is an essential step toward accurate crustal monitoring. In this study, we apply the single‐station cross‐component (SC) method to 15 selected stations from the Broadband Array in Taiwan for Seismology (BATS) to investigate the temporal evolution of crustal seismic velocities across Taiwan. We process the continuous BATS seismic recording from 1998 to 2019, construct the daily SC correlation functions, and compute dv/v values by the stretching technique in a frequency band of 0.1–0.9 Hz. We observe both strong annual dv/v variations and co‐seismic velocity drops associated with regional moderate‐to‐large earthquakes. Systematic spectral and time‐series analyses with the weather data suggest that the rainfall‐induced pore‐pressure change plays a predominant role in driving the dv/v seasonality, reflecting a diffusion process from meteoric water into shallow crust. The effects of other factors are relatively local and secondary. We also demonstrate how understanding and correcting rainfall effects could critically improve the resolution and accuracy of internal crustal damage related to earthquakes.

Towards Automatic Landslide-Quake Identification Using a Random Forest Classifier

Landslide-generated seismic waves (landslide-quakes), exhibiting distinctive waveforms and frequency characteristics, can be recorded by nearby seismometers. Implementing an automatic classifier for landslide-quakes could help provide objective and accurate initiation times of landslides with efficiency. This study collected and analyzed the time information of 214 landslide seismic records due to 33 documented landslide events, from the Broadband Array in Taiwan for Seismology (BATS). In addition, equal numbers of earthquake and noise signals were also incorporated. The 642 seismic signals and time information were carefully examined using the random forest algorithm to create an automatic landslide-quake classifier. By validating the signal attributes of the landslide, earthquake, and noise events, specifically in the time and frequency domains, it was shown that the proposed classifier can reach an accuracy (the proportion of all correctly classified events to the total number of events) of 91.3%. To further evaluate the applicability of the automatic classifier, landslide-quakes generated during the devastating Typhoon Morakot (2009) and Typhoon Soudelor (2015) were also verified, showing that the sensitivity of the classifier is higher than 98%.

2012 Seismicity Quiescence in Taiwan a Result of Site‐Effect Artifacts

ABSTRACT A significant seismic quiescence in Taiwan was reported by Wu and Chiao (2006) before the occurrence of the 1999 M w 7.6 Chi‐Chi earthquake. A similar kind of activity was also observed in the 2012 earthquake catalog reported by Central Weather Bureau (CWB). Seven months in the 2012 catalog seem to have lower monthly seismicity rates than the one standard deviation below the mean (673±82) for earthquakes with M L ≥2.0. Here, we checked the seismic network in Taiwan and found that some new seismic stations, including the Broadband Array in Taiwan for Seismology (BATS), were incorporated within the Taiwan Central Weather Bureau Seismic Network (CWBSN) since 2012. Most of these BATS stations are located on hard‐rock sites, which may affect magnitude estimation because of their site characteristics. To account for the impact of site effect, earthquake catalog data during the period of 1994 to 2014 was collected and station correction was calculated for each station. A strong correlation is found between station corrections determined in this study and geological settings of the region. Stations located on soil sites have high amplifications with negative station corrections. On the other hand, stations located on hard‐rock sites have low amplifications with positive station corrections. After applying station correction, the mean seismicity rate is found to be 716 events per month with a standard deviation of 76 events for earthquakes with M L ≥2.0. We conclude that the reason behind the apparent low seismicity in 2012 is due to the installation of new seismic stations on rock sites, which lead to underestimation of M L .

An updated and refined catalog of earthquakes in Taiwan (1900–2014) with homogenized M w magnitudes

The main goal of this study was to develop an updated and refined catalog of earthquakes in Taiwan (1900–2014) with homogenized M w magnitudes that are compatible with the Harvard M w . We hope that such a catalog of earthquakes will provide a fundamental database for definitive studies of the distribution of earthquakes in Taiwan as a function of space, time, and magnitude, as well as for realistic assessments of seismic hazards in Taiwan. In this study, for completeness and consistency, we start with a previously published catalog of earthquakes from 1900 to 2006 with homogenized M w magnitudes. We update the earthquake data through 2014 and supplement the database with 188 additional events for the time period of 1900–1935 that were found in the literature. The additional data resulted in a lower magnitude from M w 5.5–5.0. The broadband-based Harvard M w , United States Geological Survey (USGS) M, and Broadband Array in Taiwan for Seismology (BATS) M w are preferred in this study. Accordingly, we use empirical relationships with the Harvard M w to transform our old converted M w values to new converted M w values and to transform the original BATS M w values to converted BATS M w values. For individual events, the adopted M w is chosen in the following order: Harvard M w > USGS M > converted BATS M w > new converted M w . Finally, we discover that use of the adopted M w removes a data gap at magnitudes greater than or equal to 5.0 in the original catalog during 1985–1991. The new catalog is now complete for M w ≥ 5.0 and significantly improves the quality of data for definitive study of seismicity patterns, as well as for realistic assessment of seismic hazards in Taiwan.

LINKS BETWEEN MICROSEISMS AND EXTREME WAVES DURING TYPHOONS

Ocean-Solid Earth coupling is the main forcing of generating seismic noise. It has been well known that seismic noise in the frequency band from 0.05 to about 1 Hz results from ocean waves. Components of noise spectrum, also called microseisms, are used to study extreme waves during typhoons. During the KUN–SHEN project (2011-2014), wave measurements have collected at the Cigu coast of Taiwan and compared with microseisms observations at inland seismometers from Broadband Array in Taiwan for Seismology (BATS) network. By comparing between background and storm induced microseisms, peaks in the noise spectrum can be divided in to two frequency bands. Comparison of microseism band between 0.1 and 0.2 Hz with buoys in the deep sea shows a strong correlation of seismic amplitude with storm generating waves, implying that this energy portion could originates at the remote regions. Microseism amplitudes above 0.2 Hz show a good correlation with wind-generating waves near coasts. Findings also show that maximum amplitudes of two different portions have the significant time lag, implying that source locations of generating microseisms can be identified. Results indicate that secondary microseisms observed at inland sites can be a potential tool of tracking typhoon motions and monitoring extreme waves near coasts in real time. Seismic noise in the 1-10Hz band has rough correlation with wave activities in swash zone. Further examination is required to investigate those differences.

Crustal and upper mantle S-wave velocity structures across the Taiwan Strait from ambient seismic noise and teleseismic Rayleigh wave analyses

Although orogeny tapers off in western Taiwan large and small earthquakes do occur in the Taiwan Strait, a region largely untouched in previous studies owing mostly to logistical reasons. But the overall crustal structure of this region is of particular interest as it may provide a hint of the proto-Taiwan before the orogeny.By combining time domain empirical Green’s function (TDEGF) from ambient seismic noise using station-pairs and traditional surface wave two-station method (TS) we are able to construct Rayleigh wave phase velocity dispersion curves between 5 and 120s. Using Broadband Array in Taiwan for Seismology (BATS) stations in Taiwan and in and across the Strait we are able to derive average 1-D Vs structures in different parts of this region. The results show significant shear velocity differences in the upper 15km crust as expected. In general, the highest Vs in the upper crust observed in the coastal area of Mainland China and the lowest Vs appears along the southwest offshore of the Taiwan Island; they differ by about 0.6–1.1km/s. For different parts of the Strait, the upper crust Vs structures are lower in the middle by about 0.1–0.2km/s relative to those in the northern and southern parts. The upper mantle Vs structure (Moho – 150km) beneath the Taiwan Strait is about 0.1–0.3km/s lower than the AK135 model. The overall crustal thickness is approximately 30km, much thinner and less variable than under the Taiwan Island. The inversion of seismic velocity structures using shorter period band dispersion data in the sea areas with water depth deeper than 1000m should take water layer into consideration except for the continental shelves.

Seismic moments of Taiwan’s earthquakes evaluated from a regional broadband array

Abstract We have taken the seismic moment (M o) values of 79 earthquakes occurring in the Taiwan region that have been published in the Global centroid-moment tensor (CMT) and regional Broadband Array in Taiwan for Seismology (BATS) catalogues for the period 1996–2005 and compared the values determined from the global and regional networks, respectively. M oG and M oB are used to denote the M o values published in the Global CMT and regional BATS catalogues, respectively. Our results show that M oB linearly correlates with M oG and that M oB is, on average, approximately equal to 0.37M oG. This difference may be caused by the use of shorter period seismic waves in BATS for estimating M oB. The moment magnitude evaluated from regional BATS seismograms is about 0.3 less than that estimated from global data.

Preface to the 2006 Pingtung Earthquake Doublet Special Issue

On 26 December 2006, two closely linked earthquakes, both ML 7.0, occurred only 8 minutes apart at 12:26:21 and 12:34:15 UT, respectively, in southwestern offshore of Taiwan near Pingtung. These two earthquakes were referred to as the Pingtung Earthquake Doublet that is unique not only because of being an earthquake doublet, but also both events were the largest in southwestern Taiwan during the past one hundred years. The occurrence of this doublet was in a region, where the seismic activity has been relatively low. According to the Central Weather Bureau (CWB) catalog, the epicenter of the first event is located at 21.69 N and 120.56 E with a focal depth 44 km and a magnitude of ML = 7.0 (Mw = 7.0 Global CMT), the second earthquake was at 21.97 N and 120.42 E, about 36 km to the northnortheast of the first event, with a magnitude of ML = 7.0 (Mw = 6.9 Global CMT) and a focal depth of 50 km. This earthquake doublet caused 44 injuries, including 2 fatal ones, 3 building collapses, and a massive failure of submarine communication cables, resulting in a significant economic loss. In addition, the Taipower’s Nuclear Power Plant No. 3 (NPP3), which is within about 50 km from the epicenters of the earthquake doublet, registered the largest motions ever since the strong-motion recording commenced at the plant site. As a result, the Unit 2 of NPP3 was shut down while Unit 1 remained operatal at the time of the doublet event. The occurrence of this earthquake doublet has caused a debate on the possible association to an active fault, and its linkage to the subducting slab of Manila trench. The focal mechanisms determined from Broadband Array in Taiwan for Seismology (BATS) were a normal faulting for the first, and a strike-slip faulting for the second event. The associated rupture planes of the earthquake doublet were difficult to determine due to the sparse distribution of the aftershocks for the first one. The aftershock distribution from the CWB near the hypocentral area of the second one suggests a west-dipping northwest-southeast striking ruptured plane. Tectonically, this earthquake doublet gives the insight toward understanding of the seismic activity in southwestern Taiwan, and thus, brings special attention to the potential hazard from large earthquakes in southern Taiwan. In view of the characteristics of the earthquake doublet, in this Special Issue, we have compiled valuable data observed from this earthquake doublet and studies investigating earthquake occurrence and rupture characteristics of this earthquake doublet. We believe that this Special Issue should be a useful documentation to the Pintung earthquake doublet, and will be beneficial to the hazard mitigation of the future large earthquakes associated with the southwestern Taiwan tectonics. This Pingtung Earthquake Doublet Special Issue consists of 16 regular articles and 2 notes. It began with three articles on the discussion of the initial rupture of the earthquake by Haung et al. (2008a), the fault rupture plane determination and spatial slip distribution analysis from the regional strong motion and BATS waveform data by Lee et al. (2008) and Yen et al. (2008) to obtain the geometry of fault ruptures and to discuss the interaction of the earthquake doublet from the spatial slip distributions based on teleseismic data. Chen et al. (2008d) used the teleseismic data to refine the source depths of this earthquake doublet. Furthermore, Lin et al. (2008) studied the possible static triggering of earthquake doubles in convergent zones of Taiwan. The valuable strong motion data from the CWB’s Taiwan Strong-Motion Instrumentation Program (TSMIP) stations for this earthquake doublet was compiled by four articles. The most comprehensive one is by Wu et al. (2008a), giving the data processing procedures for the efficient use of the data for other researchers, and a comparison of ground motions recorded by multiple accelerographs at a common site in Taipei, and attenuation of the horizontal ground motions. Wen et al. (2008) compiled the damage features and strong motion records, especially near the Taipower’s Nuclear Power Plant No. 3, to give the seismic hazard assessment for southwestern Taiwan. Chen et al. (2008b) gave special attention to the long-period ground motion observations along paths of the earthquake doublet traversing the Central Range, and the Longitudinal Valley, respectively. Lee and Tsai (2008), furthermore, provided the near site structure for mapping average shear-wave velocity of the upper 30 meters of soil profile (Vs30) in Taiwan. In addition to seismic data, GPS observations and analysis associated with this earthquake doublet were also carried out by Chen et al. (2008a). For offshore large earthquakes, several articles address their corresponding tsunami potential and the possible association of the earthquake doublet with the offshore tectonic structure. Chen et al. (2008c) evaluated the earthquake probabilities and energy characteristics in southwestern offshore Taiwan for the slow tsunami earthquake, and Wu et al.

Taiwan shaken by an uncommon earthquake

Exactly 2 years to the date after the great Sumatra earthquake, the island of Taiwan was shaken by an uncommon magnitude 7.1 earthquake and a series of aftershocks centered just off her southwestern coast on the continental shelf. The event is the largest in a century in that particular area, technically controlled by the subduction of the Eurasia plate under the Philippine plate, a rare and peculiar case of a lighter continental plate subducting under a heavier oceanic plate.The main shock, at 1226 UT, 26 December 2006, had a moment magnitude of 7.1 as determined by the U.S. Geological Survey's National Earthquake Information Center (NEIC/USGS), followed by a 7.0 aftershock 8 minutes later, and three more subsequent aftershocks above magnitude 5 within a day. Relatively minor damage was reported on land, while underwater communication cables were severed at as many as seven locations at different times. A nearby nuclear power plant experienced a g‐force as large as 0.16g. A tsunami warning was issued in Vietnam and the Philippines but was later called off. The Broadband Array in Taiwan for Seismology (BATS) put the main shock depth at somewhat over 20 kilometers and deeper for the large aftershocks. The water depth at the epicenter region is no more than 600 meters.

Earthquake-induced gravitational potential energy change in the active Taiwan orogenic belt

SUMMARY The Philippine Sea Plate is converging against the Eurasian Plate near Taiwan at a velocity of 7‐8 cm yr −1 ; this has caused the Taiwan orogenesis and induced abundant earthquakes. In this study we examine the corresponding change of gravitational potential energy (�GPE) using 757 earthquakes from the earthquake catalogue of the Broadband Array in Taiwan for Seismology (BATS) from 1995 July to 2003 December. Our results show that the variation of

Stress Tensor Inversion for the Chi-Chi Earthquake Sequence and its Implications on Regional Collision

We study the stress regime associated with the 1999 Chi-Chi, Taiwan, earthquake sequence. Broadband waveforms of 115 events recorded by the Broadband Array in Taiwan for Seismology (BATS) are analyzed to determine their focal mechanisms, depths, and seismic moments. These solutions are then used to invert for the reduced stress tensor using the algorithm of Angelier (1998). We perform the stress inversion in three different ways. First, we aim at defining, in the first approximation, a uniform stress field by taking the entire dataset as a whole. The inversion result shows a stress regime with the maximum compression along the direction of N119°E, remarkably consistent with the direction of the relative motion between the Philippine Sea plate and Eurasia. This solution is stable and varies little in terms of the orientations of stress axes (within 8°) and the ratio of stress differences (between 0.26 and 0.44). In the second approximation, we aim at explaining more subtle features of our observations by considering differently built subsets. We divide the dataset according to different seismogenic structures as identified by Kao and Chen (2000). Our results yield rather similar results (axes of maximum compressive stress fall in the range of N99-138°E in terms of trends, and in the range of 1-17° in plunges) for all subsets except a small group in the footwall. Furthermore, we observe a systematic clockwise change in the direction of compression from south to north, representing the more general fan-shaped pattern of stress trajectories that characterizes the whole collision zone. Finally, the inversion is performed for the three main types of faulting, that is, reverse, strike-slip, and normal types. The axis of maximum compression for the reverse mechanisms has the same direction as that for the strike slip, whereas the normal type subset yields a N54°E extension that trends obliquely to the compression. The fact that the stress regime associated with the Chi-Chi aftershock sequence is similar to that derived from geological and seismic data in the region implies that the stress regime corresponding to the regional collision in Taiwan is responsible for the occurrence of the Chi-Chi earthquake. Moreover, systematic, secondary variation in the state of stress is observed for different types of focal mechanisms and seismogenic structures. Such variation may play an important role in producing local stress concentration that leads to the nucleation of disastrous earthquakes in central Taiwan. Manuscript received 12 January 2001.

Application of Distributed Object Technology to Seismic Waveform Distribution

We develop a networked data distribution system for broadband seismic waveform data using Java RMI (Remote Method Invocation). Users can download data from various networks using a unified interface. There are two technical problems in building such a user-friendly system with a high capability for expansion: communication through firewalls and information synchronization between request interfaces. We solve these two problems using cgi (common gateway interface) filtering and the Directory Service, respectively. We are now distributing data from various Japanese broadband seismic networks (including PACIFIC21) together with data from BATS (Broadband Array in Taiwan for Seismology) and IRIS through the system. Digital seismic waveform data are being accumulated at an accelerating rate by various networks. Data analysis using these huge data sets is becoming increasingly common. For example, in the Western Pacific region, a large number of digital broadband seismographic stations have recently been installed by various institutions (Figure 1). If the data from these networks are appropriately combined, they effectively constitute a data set from a very dense broadband network. Such a data set should be very useful for determining seismic structure and earthquake source properties with high resolution. However, in the current waveform distribution system, such studies require somewhat tedious procedures for collecting the waveform data from different data centers, each of which has a different user interface. Looking at the field of computer science, distributed object technology (technology for distributing software modules to multiple computers on a network; The Object Management Group, 2000) is one of the most important technologies in modern system-programming practice, as it can handle a huge system in a straightforward manner. Java RMI (Gosling and Steele, 1996; Sun Microsystems, 1999) provides one of the most successful technologies to build a distributed object management system because of its high capability for expansion and security. …

Source Parameters of Regional Earthquakes in Taiwan: 1999-2000 Including the Ch-Chi Earthquake Sequence

We report source parameters of 277 earthquakes that occurred during 99.2000 in the Taiwan region, a time period dominated by the Chi-Chi earthquake sequence. We used an improved CMT inversion algorithm to handle high background noise and structural heterogeneity from complex tectonic settings near Taiwan. We also searched the best fitting solutions at an increment of 1 km of focal depths for all events that occurred after the :hi-Chi main shock on September 20, 1999. In general, central and northern Taiwan was seismically quiet before the Chi-Chi main shock. Aftershocks of this sequence dominated the regional seismicity for approximately 3 months afterward. The Chi-Chi sequence consists of a wide range of focal mechanisms, but the overall pattern is a west-verging thrust block bound strike-slip tear faults near the northern and southern termini of the surface rupture – the Chelungpu fault. Seismicity outside of the source zone of the Chi-Chi sequence generally agrees with known active tectonics from previous studies. Nonetheless, under high elevations of the eastern Central Ranges, there are several intriguing cases of changing focal mechanisms from normal faulting near the surface to thrust at depth. To facilitate further research of the Chi-Chi sequence, results of our inversion and original waveforms from the Broadband Array in Taiwan for Seismology (BATS) of 170 events within this sequence are ready for distribution on CD-ROM. The dataset is also online at the BATS web site (http://bats.earth.sinica.edu.tw/Chi-Chi_CMT).

Seismogenic patterns in the Taiwan region: insights from source parameter inversion of BATS data

We systematically invert source parameters of 96 earthquakes that occurred in the Taiwan region between July 1995 and June 1998, using waveforms recorded by the newly established Broadband Array in Taiwan for Seismology (BATS). The results are utilized to delineate the seismogenic patterns associated with the regional tectonic processes between the Eurasia and Philippine Sea plates. In general, high seismicity is observed at five locations, including the Nan-ao basin (∼24°N, ∼122.4°E), near the Hualien area (∼24.2°N, ∼121.7°E), east of the Longitudinal Valley and within the Philippine Sea plate, the forearc region to the west of Lanshu, and the Okinawa trough. Events with focal depths less than 25km scatter across a wide region from the Taiwan Strait to the Ryukyu and Luzon arcs, whereas events between 25 and 50km concentrate offshore east of Taiwan. Most earthquakes that occurred beneath the Nan-ao basin are consistent with the northward subduction of the Philippine Sea plate along the plate interface. Events offshore east of the Longitudinal Valley and within the Philippine Sea plate are dominated by a compressional regime along the NW/NWW directions, whereas the Lanshu–Lutao forearc is characterized by E–W compression. Normal faults showing N–S extension are observed in the Okinawa trough where we observed a successive rotation of T-axes from N–S to NW–SE as the epicenters approach Taiwan, indicating a possible interaction between the extensional and compressional strain regimes there. Most events between 50 and 65km depth within the subducted Philippine Sea plate beneath Ryukyu are in downdip extension with lateral compression approximately parallel to the strike of the slab. The depth where the state of strain switches from downdip extension to downdip compression is constrained at 110–125km for the southernmost Ryukyu slab. Based on our results, we propose that the subducted Philippine Sea plate beneath NE Taiwan plays a significant role in the overall orogenic process of Taiwan. The lithospheric collision in Taiwan should consist of two major components: one is the well-recognized “arc–continent collision” that dominates the central and southern Taiwan and another the “slab–continent collision” that reigns in northern Taiwan. A cluster of earthquakes near Hualien suggests that an incipient west- or northwestward subduction of the Philippine Sea plate is currently taking place there. Alternatively, they can be interpreted to suggest that the steeply east-dipping Longitudinal Valley fault may have transformed, near Hualien, into a diffuse thrust zone, which eventually connects itself to the plate interplate beneath the southernmost Ryukyu arc. Earthquakes beneath Taiwan and the Taiwan Strait are compatible with the overall collision convergence and lateral extrusion. Nonetheless, the lateral extrusion is probably only a secondary feature in the tectonic processes of Taiwan.

Source Parameters of Regional Earthquakes in Taiwan: January-December, 1997

Establishment of the ”Broadband Array in Taiwan for Seismology (BATS)” has enabled the systematic determination of reliable source parameters for regional earthquakes in Taiwan through waveform inversion technique. Using an improved algorithm by Kao et at (1998) to overcome the generally higher background noise as well as the heterogeneous velocity structure resulted from the complex tectonic interactions near Taiwan, we report source parameters of 46 events that occurred in 1997. The inversion quality is classified by a combination of a letter (A-F) and a digit (1-4) reflecting the waveform misfit and the compensated linear vector dipole (CLVD) component, respectively. To make the results more accessible and useful to the academic community, both the table that summarizes the reported Source parameters and the complete set of inversion results are available electronically from the BATS web site at http://bats.earth.sinica.edu.tw/CMT_Solutions/cmtF1997.html.

Source Parameters of Regional Earthquakes in Taiwan: July 1995-December 1996

Systematic determination of reliable source parameters for regional earthquakes in Taiwan, particularly offshore events with m(subscript b)＜5.5, has been a difficult task because of poor coverage by the local network and the lack of signals at teleseismic distances. Establishment of the ”Broadband Array in Taiwan for Seismology (BATS)” has greatly improved such a task through moment-tensor inversion of regional waveforms. Our inversion algorithm differs from previous studies in two major aspects. First, we evaluate the characteristics of background noise for individual stations. The results are then used to determine the lower corner of the frequency band used in the inversion to maximize the long-period information in waveforms. The higher corner is set at 0.06-0.08 Hz to avoid the effects of strong lateral heterogeneity and possible epicentral mislocation. Second, to further reduce the uncertainty caused by complex structures, a two-step procedure is adapted to select the best velocity models for different stations in calculating the synthetics. The inversion quality is classified by a combination of a letter (A-F) and a digit (1-4) reflecting the waveform misfit and the compensated linear vector dipole (CLVD) component, respectively. In total, source parameters of 36 events that occurred between July 1995 and December 1996 are reported in this study. For the few events that are big enough to be studied teleseismically, most of our solutions are consistent with those reported by other institutions. We intend to make our inversion results available on a routine basis that they will not only be able to provide precise source parameters for smaller regional earthquakes, but will also serve as an alternative to independently examine solutions of large and moderate-sized events reported from other sources.