Search

Seismic instrumentation for earthquake early warnings (EEWs) has improved significantly in the last few years, considering the station coverage, data quality, and the related applications. The official EEW system in Taiwan is operated by the Central Weather Bureau (CWB) and is responsible for issuing the regional warning for moderate-to-large earthquakes occurring in and around Taiwan. The low-cost micro-electro-mechanical system (MEMS)-based P-Alert EEW system is operational in Taiwan for on-site warnings and for producing shakemaps. Since 2010, this P-Alert system, installed by the National Taiwan University (NTU), has shown its importance during various earthquakes that caused damage in Taiwan. Although the system is capable of acting as a regional as well as an on-site warning system, it is particularly useful for on-site warning. Using real-time seismic signals, each P-Alert system can provide a 2-8 s-long warning time for the locations situated in the blind zone of the CWB regional warning system. The shakemaps plotted using this instrumentation help to assess the damage pattern and rupture directivity, a key feature in the risk mitigation process. These shakemaps are delivered to the intended users, including the disaster mitigation authorities, for possible relief purposes. Earlier, the network provided only peak ground acceleration (PGA) shakemaps, but has now been updated to include peak ground velocity (PGV), spectral acceleration (Sa) at different periods, and CWB intensity maps. The PGA and PGV shakemaps plotted using this network have proven helpful in establishing the fact that PGV is a better indicator of damage detection than PGA. This instrumentation is also useful in structural health-monitoring and estimating co-seismic deformations. Encouraged by the performance of the P-Alert network, more instruments are installed in Asia-Pacific countries.

Between September 17 and 18, 2022, a series of earthquakes struck the southern Longitudinal Valley of Taiwan, causing severe damage to infrastructure, particularly bridges, railroads, buildings, and highways in Taitung and Hualien. The primary event in this sequence was the Chihshang earthquake, measuring magnitude of M L 6.9, preceded by a significant foreshock of M L 6.4 the day before. Notably, the highest intensity reported in the epicentral region during this earthquake sequence reached 6 + , marking the highest intensity ever recorded since the Central Weather Bureau (CWB) revised its seismic intensity scale. In this study, we examined the strong ground motion data recorded by the low-cost P-alert network to investigate the cause of structural damage to buildings, railway stations, schools, and other structures in the affected area. Our analysis encompassed data from a total of 17 monitoring stations. The observed peak ground acceleration (PGA) ranges from 0.06 g and 0.78 g, the peak spectral acceleration or response spectrum (PSA) spanned from 0.20 g at 0.50 s to 3.87 g at 0.41 s, and site amplification varied from 4 at 0.6 Hz to 12.6 at 5.2 Hz. The reported damage of single storey house at Taitung City is correlated with the higher amplification 4.5 at 0.14 s; Bridge and factory at Chihshang Township is correlated with the higher amplification 5 at 0.17 s; Railway station at Zhuoxi Township with the higher amplification 7.5 at 0.3 s; of three storey building at Yuli Township with amplification 8 at 0.26 s and, also higher PSA 2.83 g at 0.29 s, and several schools at Kaoshiung City with amplification 7.2 at 0.37 s and with high PSA 3.87 g at 0.41 s. Additionally, we studied the static Coulomb stress changes resulting from the coseismic slip of the main shock, which contributed to an increase in aftershock occurrences. The occurrence of a high-intensity earthquake ( M L 6.9) on September 18, 2022, within a region of positive Coulomb stress changes, provides support for the hypothesis of aftershock triggering. Through analysis, we found that stronger foreshock ( M L 6.4) occurred in the rupture area of mainshock foreshock, positioned adjacent to the main fault. [ABSTRACT FROM AUTHOR]

Earthquake early warning systems assess the size of an earthquake based on initial ground motion and provide warnings before the arrival of large waves. We employ the iterative regression to simultaneously estimate the station correlations and the corresponding linear relations for magnitude. We report the standard error reductions of the magnitude estimation based on the initial peak P-wave acceleration (${P_a}$ P a ), velocity (${P_v}$ P v ), and displacement (${P_d}$ P d ) as $14.52\% $ 14.52 % , $7.63\% $ 7.63 % , and $7.58\% $ 7.58 % , respectively. Factor analysis of the stationary correction reveals its correlation with several factors. This study shows that station correction can help to improve the precision of magnitude estimation in the future. [ABSTRACT FROM AUTHOR]

The cumulative absolute absement (CAA) of the 3 s window after P-wave arrival can be used to estimate the magnitude ( M CAA ) of an earthquake. This method can achieve good results even when only the six stations nearest to the epicenter are used. The standard deviation between the estimated CAA magnitude ( M CAA ) and the moment magnitude ( M w ) is found to be 0.3 when using either 6 or 20 stations. This means that M CAA can be reliably predicted using the closest 6 stations. On the other hand, the magnitude ( M Pd ) derived from P d using the closest 20 stations has a standard deviation of 0.4 between the estimated M Pd and M w . This suggests that CAA is a better magnitude determination parameter for the EEW system than P d . [ABSTRACT FROM AUTHOR]

Recent studies in the earthquake magnitude‐frequency relation (b‐value) found that the b‐value can decrease with time before large earthquakes in the source area. Such a decrease even coincided with a burst of foreshocks immediately before the earthquakes implying a preslip migrating toward the nucleation zone. We systematically survey the b‐value variations in space and time in the source area of 17 earthquakes with magnitudes greater than ML 6.0 in Taiwan. We relocated the earthquakes released by the Central Weather Administration of Taiwan since 2012 as timing can provide a sufficiently low completeness magnitude (MC) of 1.5. We carefully determine the spatiotemporal search criteria on seismicity in the source area of all targeted earthquakes. We surveyed the temporal b‐value systematically using moving time windows with a given number of earthquake events with the magnitudes greater than MC. We found the b‐value decreased clearly for 2 days before the 2018 ML 6.3 Hualien earthquake, coinciding with a burst of ML 5.8 earthquake and its aftershocks nearby as foreshocks. The foreshocks migrated updip toward the hypocenter of the 2018 ML 6.3 Hualien earthquake and revealed a second decrease in the b‐value 10 hours before it occurred. The b‐value precursor primarily comes from a rapid reduction of small earthquakes in the foreshocks and agrees with recent key findings of preslip. However, there is no strong correlation between the b‐value precursor and coseismic slip. Apart from this, the b‐values did not change significantly in most cases of earthquakes in Taiwan before and after they occurred. Plain Language Summary: Direct evidence for the precursor of a large earthquake is rare and difficult to be quantified accurately. Earthquake magnitude‐frequency relation (b‐value) is a classic observation implying stress accumulation on a seismogenic fault. Scientists found that the b‐value can decrease within a few days before an impending large earthquake in the source area. Taiwan is one of the regions undergoing frequent M ≥ 6.0 earthquakes and can map the b‐value in high resolution based on the relocated earthquakes. This study systematically surveyed the b‐value in the source area of seventeen ML ≥ 6.0 earthquakes in Taiwan, probing into b‐value precursor in the 8‐year time series. We found that the b‐value had a robust decrease 2 days before the 2018 ML 6.3 Hualien earthquake, coinciding with the initiation of foreshocks nearby and at a downdip of the hypocenter of the 2018 ML 6.3 Hualien earthquake. The foreshocks showed clear evidence of precursor as they migrated updip toward the hypocenter with a second decrease in the b‐value 10 hours before the earthquake occurred. We found no evidence of a b‐value precursor in most of the targeted earthquakes that agrees with the knowledge of precursor as happening episodically and unexpectedly. Key Points: We found a b‐value precursor that coincides with foreshocks for 2 days before the 2018 ML 6.3 Hualien earthquake in the source areaThe b‐value precursor correlates with foreshock distribution in the source area and is likely uncorrelated to the high coseismic slipThere is no significant change in the b‐value before M ≥ 6.0 earthquakes in Taiwan in a majority of the cases that we analyzed [ABSTRACT FROM AUTHOR]

Building an earthquake early warning (EEW) network requires the installation of seismic instruments around the seismogenic zone. Using low-cost sensors to build a seismic network for EEW and to generate shakemaps is a cost-effective way in the field of seismology. The National Taiwan University (NTU) network employing 762 P-Alert low-cost sensors based on micro-electro-mechanical systems (MEMS) technology is operational for almost the last 10 years in Taiwan. This instrumentation is capable of recording the strong ground motions of up to ± 2 g and is dense enough to record the near-field ground motion. The NTU system has shown its importance during various earthquakes that caused damage in Taiwan. Although the system is capable of acting as a regional as well as an onsite warning system, it is particularly useful for onsite warning. Using real-time seismic signals, each P-Alert device provided a 2–8 s warning time for the near-source earthquake regions situated in the blind zone of the Central Weather Bureau (CWB) regional EEW system, during the 2016 M w 6.4 Meinong and 2018 M w 6.4 Hualien earthquakes. The shakemaps plotted by the P-Alert dense network help to assess the damage pattern and act as key features in the risk mitigation process. These shakemaps are delivered to the intended users, including the disaster mitigation authorities, for possible relief purposes. Currently, the P-Alert network can provide peak ground acceleration (PGA), peak ground velocity (PGV), spectral acceleration ( S a ) at different periods, and CWB intensity shakemaps. Using shakemaps, it is found that PGV is a better indicator of damage detection than PGA. Encouraged by the performance of the P-Alert network, more instruments are installed in Asia-Pacific countries. [ABSTRACT FROM AUTHOR]

Geodetic evidence for slow slip recurrence changed by stress perturbations was rare, especially from afterslip following a nearby large earthquake. The first observed slow slip events in the southernmost Ryukyu subduction had occurred in 2005, 2009, and 2015 following the nearby 2002 March 31 Mw 7.1 Hualien offshore earthquake. To investigate whether the Mw 7.1 earthquake had influenced the occurrence times of the slow slip, we calculated the coseismic slip distribution and afterslip distribution using the surface displacements from onshore Global Navigation Satellite System observations. The stress perturbation on the slow-slip regions caused by the coseismic slip was quantified using the Coulomb failure criteria. We also examined the aftershock distribution and the evolution with time to clarify the stress perturbations from the afterslip offshore. Our results show that the primary afterslip distribution may have overlapped the 2009 and 2015 slow-slip patch at the downdip of the earthquake. The coseismic stress perturbation may have influenced the SSEs area directly by a Coulomb stress increase of probably 0.10 MPa. However, the 2005 SSE patch in the updip depths had only a little coseismic slip and afterslip with the Coulomb stress increase of approximately 0.06 MPa. We find that most of the aftershocks had occurred in the 2009 and 2015 slow-slip region that evolved into a typical aftershock sequence at least 2.5 years after the earthquake. The surface geodetic displacements reveal that the afterslip may have lasted longer than 4.5 years after the earthquake. The evidence for the afterslip proves that the 2009 and 2015 slow-slip patch was influenced by the afterslip directly for years. We suggest that the ongoing afterslip may have modulated the coseismic stress perturbation. It may explain the delay of the 2009 slow slip occurrence compared with the recurrence interval between the 2009 and 2015 slow slip events. [ABSTRACT FROM AUTHOR]

The earthquake early warning (EEW) system in Taiwan is the outcome of rigorous research work carried out at various levels after the occurrence of the 1986 Hualien earthquake that caused destruction. After more than 25 years of development, three different EEW systems exist in Taiwan. Currently, the nation wide regional EEW system is operated by the Central Weather Bureau (CWB), whereas, a hybrid (regional & onsite) system based on Micro-Electro-Mechanical System sensors is run by National Taiwan University (NTU). The third EEW (onsite system) is run by the National Center for Research on Earthquake Engineering (NCREE). Both CWB and NTU systems are capable of reporting the EEW warnings within 20 seconds of earthquake occurrence. The CWB system is incharge of providing earthquake alerts in Taiwan via text message through mobile phone, TV, and directly broadcasting system to schools and is providing earthquake alarms to the general public since 2016. During recently damaging earthquakes in Taiwan, the NTU system provided 2–8 seconds onsite warning (lead time) in the blind zone around the epicenter. The NTU system also can generate near real-time shake maps for rapid response purposes. The NCREE system consists of about 98 stations and can provide several seconds lead time in the area near the epicenter. The NCREE system also can receive CWB regional EEW messages for warning the regions away from the epicenter. Individually every system has its advantage, however, the hybrid approach will be one of the future systems for real operation. [ABSTRACT FROM AUTHOR]

Recent studies suggested that transient and long‐term stress changes caused by Earth's surface processes (e.g., extreme weather events, annual variations on groundwater storages) can affect earthquake activities in the subsurface. However, these studies may be limited by the completeness of standard earthquake catalogs, especially during or right after extreme weather events. Here we apply the template matching method to build a more complete earthquake catalog in Taiwan spanning seven months before and 12 months after 2009 typhoon Morakot, which brought the highest rainfall in southern Taiwan in the past 60 years and triggered numerous landslides. We then use the enhanced catalog to investigate possible influences of typhoon‐driven Earth's surface processes (atmospheric pressure, precipitation, and erosion) on local seismicity. We find that the seismicity rate of a 40‐day earthquake sequence in northeastern Taiwan was reduced significantly right after the passage of typhoon Morakot's eye center. In the typhoon‐triggered landslide zone in southern Taiwan, we find a slight increase in background seismicity rate in the next year after Morakot, matching the results of a recent study. However, we do not observe a clear change in the Gutenberg‐Richter b‐value in this zone, which is different from the recent study. Station outages during and right after Morakot prevents us from better understanding short‐term precipitation effect on local seismicity. Overall, except for a reduction in seismicity rate near the typhoon's low‐pressure eye center in northeastern Taiwan, we do not observe other clear seismicity changes that can be attributed to surface changes induced by typhoon Morakot. Plain Language Summary: It is important to study cascading nature hazards (typhoons/hurricanes, landslides, earthquakes, etc.) because of their threats to human life and property. 2009 typhoon Morakot is the deadliest typhoon in Taiwan, which brought the highest rainfall in the past 60 years and led to numerous landslides. To investigate the possible influences of Morakot on subsurface earthquake activities, we build a more complete earthquake catalog for Taiwan spanning seven months before and 12 months after 2009 typhoon Morakot. This enhanced earthquake catalog includes many small earthquakes that cannot be felt by local residents in their ordinary life and were not in the standard earthquake catalog. In northeastern Taiwan, our results show that a 40‐day earthquake sequence was apparently shut down right after the passage of typhoon Morakot. If this observation is not a pure coincidence, it may suggest that the low‐pressure system of typhoon Morakot was capable of modulating the subsurface seismicity behavior. In southern Taiwan, the heavy‐rainfall and landslide region, we do not observe clear seismicity changes that can be attributed to surface changes induced by typhoon Morakot. Key Points: We use a template matching method to rebuild an earthquake catalog for Taiwan 7 months before and 12 months after 2009 typhoon MorakotThe seismicity rate of a 40‐day earthquake sequence in northeastern Taiwan was reduced significantly after the passage of MorakotThe seismicity rate increased slightly but without clear b‐value change after Morakot in typhoon‐driven landslide zone in southern Taiwan [ABSTRACT FROM AUTHOR]

Ambient 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. Plain Language Summary: The measured wave speeds from small ground motions such as ocean, wind, traffic, etc help us to understand changes in subterranean regions. The causes of the medium changes are multifold such as the fault zone damage by earthquakes (internal) or groundwater rise by rainfall (external). To discriminate the causes it is important to use weak ground motions for monitoring internal medium changes. For this purpose, we collect and study 20 years of ground motion data from 15 seismic stations distributed across Taiwan. We measure the propagation wave speed of weak ground motions and find earthquake‐related changes and strong seasonal changes at most stations. By analyzing and comparing wave speed changes with weather data (e.g., rainfall, temperature, air pressure, and wind speed), we find that rainfall varying groundwater level and the subterranean pore pressure best fit the seasonal wave speed changes in both space and time. After removing this rainfall‐related signals, we can estimate more precisely the extent of earthquake‐related underground damages that may not be obvious or completely masked by the strong seasonal changes. Key Points: We observe both coseismic dv/v drops and strong seasonal dv/v variations through a 20‐year systematic analysis in TaiwanThe rainfall‐induced pore pressure change is likely the primary control of the seasonal dv/v variations in the crustEvaluating the environmental influences on dv/v allows us to isolate crustal damage related to earthquakes [ABSTRACT FROM AUTHOR]

The rheology of the crust and mantle and the interaction of viscoelastic flow with seismic/aseismic slip on faults control the state of stress in the lithosphere over multiple seismic cycles. The rheological behavior of rocks is well constrained in a laboratory setting, but the in situ properties of the lithosphere and its lateral variations remain poorly known. Here, we access the lower‐crustal rheology in Southern California by exploiting 8 years of geodetic postseismic deformation following the 2010 El Mayor‐Cucapah earthquake. The data illuminate viscoelastic flow in the lower crust with lateral variations of effective viscosity correlated with the geological province. We show that a Burgers assembly with dashpots following a nonlinear constitutive law can approximate the temporal evolution of stress and strain rate, indicating the activation of nonlinear transient creep before steady‐state dislocation creep. The transient and background viscosities in the lower crust of the Salton Trough are on the order of ~1018 and ~1019 Pa s, respectively, about an order of magnitude lower than those in the surrounding regions. We highlight the importance of transient creep, nonlinear flow laws, and lateral variations of rheological properties to capture the entire history of postseismic relaxation following the El Mayor‐Cucapah earthquake. Key Points: Coupled viscoelastic flow and afterslip explain the 8 years of postseismic deformation following the El Mayor‐Cucapah earthquakeThe effective viscosity in the lower crust is heterogeneous over space and timeThe mechanical response of the lower crust includes nonlinear transient creep transitioning to steady‐state dislocation creep [ABSTRACT FROM AUTHOR]

Earthquake early warning system uses high-speed computer network to transmit earthquake information to population center ahead of the arrival of destructive earthquake waves. This short (10 s of seconds) lead time will allow emergency responses such as turning off gas pipeline valves to be activated to mitigate potential disaster and casualties. However, the excessive false alarm rate of such a system imposes heavy cost in terms of loss of services, undue panics, and diminishing credibility of such a warning system. At the current, the decision algorithm to issue an early warning of the onset of an earthquake is often based on empirically chosen features and heuristically set thresholds and suffers from excessive false alarm rate. In this paper, we experimented with three advanced machine learning algorithms, namely, $K$ -nearest neighbor (KNN), classification tree, and support vector machine (SVM) and compared their performance against a traditional criterion-based method. Using the seismic data collected by an experimental strong motion detection network in Taiwan for these experiments, we observed that the machine learning algorithms exhibit higher detection accuracy with much reduced false alarm rate. [ABSTRACT FROM AUTHOR]

Uttarakhand Himalayas are highly sensitive to seismic hazard with possible occurrence of high-magnitude earthquakes. Fewer waveforms are available from previously recorded earthquakes, which are insufficient for carrying out seismic hazard studies. The recently installed strong motion instrumentation network (SMIN) in India, particularly, in Indian Himalayas is providing useful data. Using recorded data from SMIN, time-dependent peak ground acceleration and observed peak ground velocity shake maps are drawn for two earthquakes widely recorded by SMIN in Uttarakhand region of Indian Himalayan belt. Open-source Earthworm software with new algorithms is used for drawing these shake maps. The source mechanism is computed for April 4, 2011 earthquake using waveform inversion technique to relate it to the trend of shake maps. The computed focal mechanism shows one of the nodal planes in NW–SE, which are consistent with shake maps for the same earthquake. These time-dependent plotted shake maps provide useful information on the initial rupture, as well as the potential directivity of the rupture. [ABSTRACT FROM AUTHOR]

The main goal of present study is to test the functionality of an earthquake early warning (EEW) system (a life-saving tool), in India using synthesized data and recorded earthquake data from Taiwan. In recent time, India set up an EEW system in the central seismic gap along the Himalayan Belt, consisting of about 100 low-cost P-Alert instruments. The area, where these instruments are installed, is highly sensitive to the seismic risk with the potential of strong, major and great earthquakes. In the absence of recorded data from the Himalayas required for analysis of such system, we take advantage of recorded waveforms from Taiwan, to test the EEW system. We selected Taiwanese stations in good accordance with the Indian sensor network, to have a best fit in terms of inter station spacing. Finally, the recorded waveforms are passed through Earthworm software using tankplayer module. The system performs very well in terms of earthquake detection, P-wave picking, earthquake magnitude and location (using previously estimated regressions). Pd algorithm has been tested where the peak amplitude of vertical displacement is used for estimating magnitudes using previously regressed empirical relationship data. For the earthquakes located between Main Boundary Thrust and Main Central Thrust along with a matching instrumentation window, a good estimate of location, as well as magnitude is observed. The approach based on Pd for estimating magnitude works perfectly as compared to τc approach, which is more sensitive to signal-to-noise ratio. To make it more region specific, we generated synthetic seismograms from the epicenters of historical Chamoli (1999) and Uttarkashi (1991) earthquakes at EEW stations in India and checked the functionality of EEW. While placing these earthquakes within the instrumentation window, a good approximation of earthquake location and magnitude is obtained by passing these generated waveforms. The parameters used to judge the performance of EEW system included the time taken by the system in issuing warning after the confirmation of the occurrence of damaging earthquake and the lead time (time interval between the issuing of warning and arrival of damaging earthquake ground motion at a particular location). High lead times have been obtained for the plainer regions including thickly populated regions of Gangetic plains, such as Delhi National Capital Region according to the distance from the epicenter, which are the main target of EEW system. [ABSTRACT FROM AUTHOR]

A series of earthquakes that struck Taiwan's southern Longitudinal Valley on September 17 and 18, 2022 severely damaged several buildings in Taitung and Hualien. The Chishang earthquake, which had a magnitude of ML 6.8 and a large foreshock with a magnitude of ML 6.6 the day before, was the mainshock in this sequence. The strongest intensity reported in the epicentral region during this earthquake sequence, which was 6 + , is the highest ever recorded since the Central Weather Bureau (CWB, renamed as the Central Weather Administration since September 15, 2023) revised its seismic intensity scale. National Taiwan University (NTU) has operated a low-cost earthquake early warning (EEW) system known as the P-Alert for a decade. In this study, we demonstrate the performance of the P-Alert network during the 2022 Chishang earthquake and the largest foreshock. The P-Alert network plotted shake maps during these earthquakes that displayed various values within 5 min. The high shaking areas on these maps were in good agreement with observed damages during this earthquake, providing valuable insights into rupture directivity, a crucial component of earthquake engineering. Individual P-Alert stations acted as on-site EEW systems and provided a lead time of 3–10 s within the blind zone of CWB. For the ML 6.8 mainshock, there was a lead time of at least 5 s, even up to 10 s, demonstrating their effectiveness in the blind zone. The P-Alert regional EEW system provided the first report about 9 s and 7 s after the mainshock and the largest foreshock occurrence, respectively, with estimated magnitudes of 5.74 and 5.67. The CWB system estimated magnitudes of 6.72 and 6.16 in the first report, respectively, about 7 s and 9 s after the earthquake occurrence. The timeliness of the two systems were not significantly different. Despite the effectiveness of the P-Alert network, data loss due to connection interruptions prompted us to develop a new compact data logger for improved data availability. [ABSTRACT FROM AUTHOR]

In present work, seismic hazard from future earthquake is worked out for Delhi region in terms of different strong motion parameters such as peak ground acceleration (PGA), characteristics frequency and spectral acceleration (Sa). The earthquake of March 5, 2012, is taken as key earthquake for synthesis. Stochastic finite modeling technique based on dynamic corner frequency initially is used to produce and match the ground motion histories where 2012 earthquake was recorded. The matching is attained in terms of PGA, response spectra and duration. Once a good match is found, the ground motion is estimated for higher magnitude earthquakes (i.e., Mw 6.0 and Mw 6.5). Our work demonstrates that a Mw 6.0 magnitude earthquake in proximity of Delhi will deliver PGA estimations of 20-209 gal (1 cm/s = 1 gal), the lower values occurring at hard rock sites like NDI (IMD) and DJB. Similarly Mw 6.5 earthquake may produce PGA values ranging between 30 and 323 gal. Finally seismic hazard in Delhi and surrounding regions is estimated from Mw 6.5 magnitude earthquake in terms of PGA, Sa and predominant period. Our computation specifies that at short period, the small structures toward eastern and north-western part of Delhi city may be affected by the earthquakes. For a case of 0.5 s period, Sa values are distributed uniformly at all the places in Delhi, indicating that the buildings with five floors or so may be in danger from future higher magnitude earthquakes. The Sa maps acquired in this study can be utilized to survey the seismic danger of the region and identify vulnerably susceptible areas in and around Delhi from future higher magnitude earthquake. [ABSTRACT FROM AUTHOR]

Since the year 1973, more than 54,000 M ≥ 3.0 earthquakes have occurred around Taiwan, and their magnitude-frequency relationship was found following with the Gutenberg-Richter recurrence law with b value equal to 0.923 from the least-square calculation. However, using this b value with the McGuire-Arabasz algorithm results in some disagreement between observations and expectations in magnitude probability. This study introduces a simple approach to optimize the b value for better modeling of the magnitude probability, and its effectiveness is demonstrated in this paper. The result shows that the optimal b value can better model the observed magnitude distribution, compared with two customary methods. For example, given magnitude threshold = 5.0 and maximum magnitude = 8.0, the optimal b value of 0.835 is better than 0.923 from the least-square calculation and 0.913 from maximum likelihood estimation for simulating the earthquake's magnitude probability distribution around Taiwan. [ABSTRACT FROM AUTHOR]

Earthquake prediction is by all means controversial and challenging, given the fact that some recent catastrophic earthquakes went unpredicted. Not surprisingly, statistical approaches have been utilized to model earthquake randomness in time or space. One of the suggestions is that the earthquake's temporal probability distribution should follow the Poisson model, which is suitable for rare events by definition. As a result, the customarily used hypothesis should be largely associated with the prior judgment that earthquakes are rare, but not as a result of abundant quantitative evidence or theoretical derivation. Therefore, this study aims to offer new empirical evidence to the hypothesis based on 110-year-long earthquake data around Taiwan. From the series of statistical tests, the first statistical inference is indeed in line with the model's proposition: the level of fitting between observation and theory is better for earthquakes with a lower mean rate. To be more specific, it shows that the Poissonian hypothesis applied to local magnitude earthquakes around Taiwan with a mean annual rate as high as 1,600 is clearly rejected, but as far as earthquakes with a mean rate of 0.35 per year are concerned, the same hypothesis is statistically accepted for modeling their temporal randomness. Also, according to the tests on a variety of conditions, the annual rate of approximately 0.1 per year (or 10-year return period) was suggested as a reasonable empirical estimate for Poissonian rareness. Accordingly, from a practical point of view, it should be a robust analytical presumption to use the Poisson model in daily earthquake engineering analyses because the return period of design earthquakes is longer than 10 years, if not much longer. [ABSTRACT FROM AUTHOR]

This study first presents the series of peak ground acceleration (PGA) in the three major cities in Taiwan. The PGAs are back-calculated from an earthquake catalog with the use of ground motion models. The maximums of the 84th percentile (mean + one standard deviation) PGA since 1900 are 1.03, 0.36, and 0.10 g, in Taipei, Taichung, and Kaohsiung, respectively. Statistical goodness-of-fit testing shows that the series of PGA follow a double-lognormal distribution. Using the verified probability distribution, a probabilistic analysis was developed in this paper, and used to evaluate probability-based seismic hazard. Accordingly, given a PGA equal to 0.5 g, the annual exceedance probabilities are 0.56, 0.46, and 0.23 % in Taipei, Taichung, and Kaohsiung, respectively; for PGA equal to 1.0 g, the probabilities become 0.18, 0.14, and 0.09 %. As a result, this analysis indicates the city in South Taiwan is associated with relatively lower seismic hazard, compared with those in Central and North Taiwan. [ABSTRACT FROM AUTHOR]

Abstract: In order to better understand the mode of subduction polarity reversal in the Taiwan–Luzon arc-continent collision zone, we mapped its crust–mantle boundaries using local earthquake tomography. By contouring surfaces of constant Vp=7.5kms−1, we identified three Moho discontinuities and the plate interface that juxtaposes Eurasian lower crust against mantle lithosphere of the Philippine Sea plate. The plate interface dips to the east under southeastern Taiwan and steepens progressively towards north until it becomes vertical at 23.7°N. From there it continues northward in a vertical orientation, until the limit of the tomographic model inhibited further mapping. For the Moho, additional depth constraints were derived from 1D models using P-wave arrivals of local earthquakes. The Mohos of the Eurasian and Philippine Sea plates are disconnected across the plate interface. Beneath southern Taiwan, the Eurasian Moho dips to the east at 50–60°, following the orientation of the plate boundary and continuous with the Benioff zone. Towards north, the Eurasian Moho steepens into subvertical, again together with the plate boundary. The Philippine Sea plate Moho exhibits a synform-like crustal root, interpreted as the base of the magmatic Luzon arc. Towards the north, this root deepens from 30 to 70km underneath the Ryukyu trench. In northernmost Taiwan, the hinge of the vertically subducting Eurasian slab steps westward out of the thrust belt, leaving the deformation front to the east inactive and giving way to the north-dipping Philippine Sea plate. A subhorizontal Moho at 30–35km depth overlies the north-dipping Philippine Sea slab and is interpreted as a newly formed Moho, established after westward rollback and delamination of the subducting Eurasian slab. In combination, these data support a model of progressive subduction polarity reversal, in which a tear within the Eurasian lithosphere propagated southwestward, deactivating the deformation front. [Copyright &y& Elsevier]

Abstract: Following the 4 March 2010 MW 6.4 Jiashian earthquake the seismicity rate in southern Taiwan was determined to be significantly higher than before the quake and aroused seismic hazard awareness. In this work, seismic hazards were investigated in terms of earthquake activity, the Coulomb stress change, the rate-and-state friction model, and short-term seismic hazard assessments. The significantly higher seismicity rate that followed the 2010 Jiashian earthquake was found to mainly be attributed to aftershock decay, in terms of the modified Omori formula. The results suggest that aftershock duration may continue until the end of 2012. The spatial migration of seismicity was modeled using the Coulomb stress changes of large earthquakes. Most of the consequent events were distributed in the vicinity of large earthquakes. The observations corresponded to a remarkable stress increase within the same area. Additionally, large events were located within regions with stress increases promoted by previous earthquakes. The results confirm interactive relationships between large events. By considering time-dependency, the seismicity rate evolution was estimated using the rate-and-state friction model. The results indicated that a high seismic rate will persist at least until the end of 2012. Short-term probabilistic seismic hazard assessments were also applied in terms of the probability of strong ground motion. Using this application, a sudden jump in seismic hazards in southern Taiwan was accompanied by each large earthquake. At the end of 2012 it is expected that hazards will return to a background level. Our results may be valuable in the future to decision-makers and public officials engaged in seismic hazard mitigation. [Copyright &y& Elsevier]

Abstract: We investigate the spatial and temporal variations of b-values before twenty-three earthquakes with M L ≥6.0 in the Taiwan region from 1999 to 2009. We estimate the spatial distribution of b-values within a one-year period before the occurrence of investigated earthquakes. It shows that the epicenters of those earthquakes are located predominately in the regions with low b-values relative to the entire study area. In terms of temporal evolutions, the b-values are slightly lower during the year prior to the investigated earthquakes in comparison to the periods two to five years earlier. We propose that these anomalies may offer spatiotemporal constraints for the earthquake forecasting in Taiwan in the future. [Copyright &y& Elsevier]

The annual maximum earthquake magnitudes around Taiwan from 1900 to 2009 are presented in this paper. Using the distribution of the AMEM, a probabilistic framework to estimate the recurrence probability of a large-size earthquake is also proposed and an illustration was made in this paper. The mean value of the 110-AMEM is 6.433, and the coefficient of variation is around 10%. The results of two goodness-of-fit tests show that the Gamma and lognormal distributions are relatively suitable to represent the AMEM around Taiwan among five common probability distributions. Using the proposed approach, the recurrence probability is 4% for an earthquake with magnitude greater than 7.5 in a 1-year period around Taiwan. More site-specifically, the probability is around 5% in Central Taiwan for such an earthquake to occur in a 50-year period. [ABSTRACT FROM AUTHOR]

Abstract: A very common but important observation of collided and accreted volcanic arcs is that the forearc regions of these arcs are largely or even entirely missing. The processes and mechanisms responsible for the removal and transport of the forearc materials from the collisional belts are thus important issues in understanding tectonics and crustal growth. The young and ongoing collision between the Luzon volcanic arc and the Eurasian continental margin that forms the island of Taiwan provides a rare opportunity to examine these processes and mechanisms as they occur. From observations of a new detailed 3-D tomography combined with relocated hypocenters of two earthquake sequences occurred in 2003 and 2006, we found that the Luzon forearc lithosphere initially underthrusts westward after the collision began. As the collision proceeds, the forearc basement then subducts eastward beneath the colliding and accreting Luzon arc along a major fault system in eastern Taiwan. Thus the Luzon forearc lithosphere appears to be removed by tectonic erosion and is being transported eastward into the mantle. Our results from the active Taiwan orogen will provide important insights for interpreting rock records from many old arc-continent collisional belts in the world. [Copyright &y& Elsevier]

Abstract: The use of characteristic period τc and peak displacement amplitude Pd of the initial P wave in earthquake early warning (EEW) was proposed by Wu and Kanamori . Here we apply this approach to strong-motion records from a building sensor array installed in Taitung County, Taiwan. This building was damaged during the 2006 Mw =6.1 Taitung earthquake with a peak ground velocity (PGV) of up to 38.4cm/s at an epicentral distance of 14.5km. According to our analysis, the peak displacement amplitude Pd is a better indicator for the destructiveness of an earthquake than τc because τc is more sensitive to the signal-to-noise ratio (SNR) than Pd. In accordance with previous studies, only the structurally damaging Taitung earthquake generated a Pd value larger than 0.5cm (a threshold for identifying damaging events). Using Pd as an indicator for destructive earthquakes does not lead to missing or false alarms for EEW purposes. [ABSTRACT FROM AUTHOR]

Abstract: A new prototype earthquake early warning (EEW) system is being developed and tested using a real-time seismographic network currently in operation in Taiwan. This system is based on the Earthworm environment which carries out integrated analysis of real-time broadband, strong-motion and short-period signals. The peak amplitude of displacement in the three seconds after the P arrival, dubbed Pd , is used for the magnitude determination. Incoming signals are processed in real time. When a large earthquake occurs, P-wave arrival times and Pd will be estimated for location and magnitude determinations for EEW purpose. In a test of 54 felt earthquakes, this system can report earthquake information in 18.8±4.1s after the earthquake occurrence with an average difference in epicenter locations of 6.3±5.7km, and an average difference in depths of 7.9±6.6km from catalogues. The magnitudes approach a 1:1 relationship to the reported magnitudes with a standard deviation of 0.51. Therefore, this system can provide early warning before the arrival of S-wave for metropolitan areas located 70km away from the epicenter. This new system is still under development and being improved, with the hope of replacing the current operational EEW system in the future. [ABSTRACT FROM AUTHOR]

Abstract: Modern technology allows real-time seismic monitoring facilities to evolve into earthquake early warning (EEW) systems, capable of reducing deaths, injuries, and economic losses, as well as of speeding up rescue response and damage recovery. The objective of an EEW system is to estimate in a fast and reliable way the earthquake’s damage potential, before the strong shaking hits a given target. The necessary framework for EEW implementation is provided by the observed relationships between different parameters measured on the signal onsets and the final earthquake size. The implication of these observations on the physics of fracture processes has given rise to a significant debate in the seismological community. Currently, EEW systems are implemented or under testing in many countries of the world, and different methodologies and procedures have been studied and developed. The leading experience of countries like Japan or Mexico shows that, with a proper education of population and end-users, and with the design of real-time systems for the reduction of vulnerability/exposure, EEW can be an effective approach to the mitigation of the seismic risk at short time-scales. [ABSTRACT FROM AUTHOR]

Abstract: The total amount of 4,761 focal mechanisms was determined based on P-wave first motion polarities from 1991 to 2007 in Taiwan region. This dataset offers us a good opportunity to examine temporal and spatial variability of the stress field. We find that the spatial variations of stress axes are mainly controlled by tectonic structures while the temporal changes are greatly influenced by the Chi-Chi earthquake. The orientation of the maximum horizontal compressive stress axes (SH) shows a general agreement with the direction of plate motion between a depth range of 0–30km. The 20° anticlockwise rotation of SH from the Longitudinal Valley (LV) to western Taiwan is probably caused by the left-lateral motion on the Longitudinal Valley Fault (LVF) that has consumed part of the oblique motion of plate convergence. On the other hand, part of the oblique convergence is transferred into the Central Range and the Hsuehshan Range judging from counterclockwise rotation of SH from east to west and strike-slip faulting in the Hsuehshan Range. Most events with a depth greater than 30km occur offshore eastern Taiwan and the azimuth of SH is close to E–W directed, different from NW–SE directed at shallow depths. This may infer the existence of the transition of lithosphere rheology in offshore eastern Taiwan. The trends of SH in the depth of 0–10km are strongly affected by the coseismic stress change of the Chi-Chi earthquake. In the northern half of the Chi-Chi rupture area, the trends of SH rotate 30° clockwise and the stress ratio increased by a factor of six after the mainshock. The orientations of SH still differ by 30° in 2007 comparing to that in the period before the Chi-Chi earthquake. The variation of SH trend is more diverse in the southern half of the rupture area, showing 20° counterclockwise rotation immediately after the 1999 mainshock followed by a clockwise rotation. The trend of SH returns to the pre-seismic direction of 110° in 2001. These notable changes of SH before and after the Chi-Chi mainshock suggests that the magnitude of background stress in the rupture area is close to the coseismic stress drop. We also recognize a significant 10° counterclockwise rotation of SH in the entire Chi-Chi rupture area between 1991 and 1999 before the earthquake took place. However, to the south of the Chi-Chi rupture, the trends of SH remain little changed before and after the Chi-Chi earthquake. [Copyright &y& Elsevier]

Abstract: We combine the arrival time data from eleven ocean bottom seismometers (OBS) and from permanent stations of the Central Weather Bureau Seismic Network (CWBSN) and the Taiwan Strong Motion Instrumentation Program (TSMIP) to improve the three-dimensional Vp and Vp/Vs structures in the Taiwan region, relocate hypocenters of earthquakes in southern Taiwan including the 2006 Pingtung Earthquake sequence, and determine the focal mechanisms of some of the larger events. Our tomography result suggests that in the region south of Taiwan the Eurasian continental lithosphere is being subducted beneath the oceanic Philippine Sea plate, leading to a region of low P-wave speed with seismicity on top of a relatively high P-wave speed zone. The hypocentral distribution of the Pingtung Earthquake sequence indicates that the first event in the initial two largest shocks was a normal-faulting quake caused by the bending of the subducting lithosphere. The second event, perhaps triggered by the first one, was a strike-slip earthquake. [Copyright &y& Elsevier]

Abstract: We investigate the variations in seismicity pattern in the Taiwan region before the 2003 Chengkung, Taiwan, earthquake (M w =6.8) by calculating the standard normal deviate of the Z-values and b-values from the Gutenberg–Richter relation. The Mogi donut-shaped variations in the seismicity can be identified in the Z-value map surrounding the Chengkung earthquake source region. Noticeable decreases in the b-values have also been found around the mainshock region before the Chengkung event. The relatively low-seismicity rate and the decrease in the b-values may be the precursory phenomena associated with the quiescence in overall seismicity and the activation of moderate-sized events occurred around the mainshock region before the Chengkung event. [Copyright &y& Elsevier]

Abstract: The Mw=6.8 Chengkung earthquake on December 10, 2003 in eastern Taiwan was associated with abundant aftershocks. To explore the seismogenic mechanism and its causative fault system, we relocated the aftershocks, determined the focal mechanisms from P-wave polarities and used the Finite Dimension Source Model method to identify the fault plane. Based on our results, the earthquake can be divided into three seismic clusters: one is located in the east of the southern Longitudinal Valley, one on the Chihshang fault, and the other on the western side of the Luyeh fault to the south. Both the seismic distribution and Finite Dimension Source Model method show that east-dipping fault planes with thrust mechanisms coincide with previous studies based on background seismicity. The third seismic cluster suggests thrust mechanism along west-dipping fault plane that was not recognized by background seismicity. This seismic cluster occurred along the Central Range Fault previously proposed as active by geomorphic and geologic evidence. [Copyright &y& Elsevier]

Abstract: We in this study have calculated the standard normal deviate Z-value to investigate the variations in seismicity patterns in the Taiwan region before and after the Chi-Chi earthquake. We have found that the areas with relatively high seismicity in the eastern Taiwan became abnormally quiet before the Chi-Chi earthquake while the area in the central Taiwan with relatively low seismicity showed unusually active. Such a spatially changing pattern in seismicity strikingly demonstrates the phenomenon of “seismic reversal,” and we here thus present a complete, representative cycle of “seismic reversal” embedding in the changes of seismicity patterns before and after the Chi-Chi earthquake. [Copyright &y& Elsevier]

We introduce an empirical method of near real-time, near-field magnitude determination for large (M>6.5) crustal earthquakes. Time integration over the strong shaking duration on the absolute values of the acceleration records is carried out for nearby stations surrounding many large earthquake sources in Taiwan. The integrated quantity, here denoted as total effective shaking, is used in a regression process to derive an empirical relationship for a quick Mw determination useful for a reliable real-time operation in earthquake rapid reporting and earthquake early warning systems (EWSs). [Copyright &y& Elsevier]

On February 6, 2018, a Mw 6.4 earthquake struck near Hualien city, in eastern Taiwan with a focal depth of 10.4 km. The earthquake caused strong shaking and severe damage to many buildings in Hualien. The maximum intensity during this earthquake reached VII (> 0.4 g) in the epicentral region, which is extreme in Taiwan, and capable of causing damage in built structures. Seventeen people died and approximately 285 were injured. Taiwan was one of the first countries to implement an earthquake early warning (EEW) system that is capable of issuing a warning prior to strong shaking. In addition to the official EEW run by the Central Weather Bureau (CWB), a low-cost EEW system (P-Alert) has been deployed by National Taiwan University (NTU). The P-Alert network is currently operational and is capable of providing on-site EEW as well as a map of expected ground shaking. In present work, we demonstrate the performance of the P-Alert network during the 2018 Hualien earthquake. The shake maps generated by the P-Alert network were available within 2 minutes and are in good agreement with the patterns of observed damage in the area. These shake maps provide insights into rupture directivity which are crucial for earthquake engineering. During this earthquake individual P-Alert stations acted as onsite EEW systems and provided 2-8 s lead-time in the blind zone around the epicenter. The co-seismic deformation (Cd) is estimated using the records of P-Alert stations. The higher Cd values (Cd > 2) in epicentral region are very helpful for authorities to respond for the purpose of damage mitigation. [ABSTRACT FROM AUTHOR]

Many earthquake empirical models were developed based on the statistics in the past. However, it is commonly seen that a non-local model was applied to a local study without any adjustment. In this paper, a new algorithm using maximum likelihood estimation (MLE) to adjust a non-local model for local applications was presented, including a case study assessing the probability of major earthquake occurrences in Taipei. Specifically, considering the fault length of 36 km and slip rate of 2 mm/yr, it suggests the Sanchiao (or Shanchiao) fault could induce a major earthquake with magnitude M w 7.14±0.17, based on a non-local model integrated with limited local data using the MLE algorithms. [ABSTRACT FROM AUTHOR]

Earthquake early warning (EEW) has been implemented in several regions around the world. However, because of natural randomness and uncertainty, false alarm and missed alarm can be expected in EEW. The key scope of this study is to evaluate the reliability of an on-site EEW in Taiwan, by testing the system's algorithm with 17,836 earthquake data from 1999 to 2013. The analysis shows that the on-site EEW system, empirically speaking, should have a false-alarm probability of 2.5%, and a missed-alarm probability of 14.1%. Considering missed alarm should be more critical to EEW, a new algorithm that could reduce the system's missed-alarm occurrences to 6% is also discussed in this paper. [ABSTRACT FROM AUTHOR]

On Mar. 27, 2013, a M L 6.2 earthquake occurred in the Nantou area of central Taiwan, which caused one death and nearly 90 injured. Two months later, another M L 6.3 earthquake struck the same region on June 2, 2013, the epicenter of which is close to the March M L 6.2 earthquake. Seismicity is a sensitive indicator of stress rate and inelastic deformation process in crust. Therefore, examination of temporal changes in seismicity is important to understand the preparatory processes of damaging inland earthquakes. In this study, we applied the Epidemic-Type Aftershock-Sequences model (ETAS model) to the earthquake data covering broader Taiwan region, which is maintained by the Central Weather Bureau (CWB) of Taiwan, to investigate precursory temporal changes in seismicity for the M L 6.2 Nantou earthquake. We regard the March M L 6.2 and June M L 6.3 earthquakes as an event sequence and especially focus on temporal changes in seismicity prior to the M L 6.2 event. Application of more than one model to an earthquake catalog would be informative in elucidating the relationships between seismicity precursors and the preparatory processes of large earthquakes. Based on this motivation, we further applied two different approaches: the pattern informatics (PI) method and the ZMAP method, which is a gridding technique based on the standard deviate (Z-value) test to the same earthquake data of CWB. As a result, we found that the epicenter of the 2013 M L 6.2 Nantou earthquake was surrounded by three main seismic quiescence regions prior to its occurrence. The assumption that this is due to precursory slip (stress drop) on fault plane or its deeper extent of the M L 6.2 Nantou earthquake is supported by previous researches based on seismicity data, geodedic data, and numerical simulations using rate- and state-dependent friction laws. [ABSTRACT FROM AUTHOR]

Abstract: The first attenuation relationships of peak ground acceleration (PGA) and peak ground velocity (PGV) for northern Vietnam are obtained in this study. Ground motion data are collected by a portable broadband seismic network in northern Vietnam as a part of cooperation between the Institute of Geophysics, Vietnamese Academy of Science and Technology, Vietnam and Institute of Earth Sciences, Academia Sinica, Taiwan. The database comprises a total of 330 amplitude records by 14 broadband stations from 53 shallow earthquakes, which were occurred in and around northern Vietnam in the period between 01/2006 and 12/2009. These earthquakes are of local magnitudes between 1.6 and 4.6, focal depths less than 30km, and epicentral distances less than 500km. The new attenuation relationships for PGA and PGV are: where PGA is in cm/s2, PGV is in cm/s, and R is the epicentral distance in kilometers. The site corrections are also derived in this study. These site corrections are very suitable with the station corrections for ML and imply the qualification of the resulting attenuation relationships. [Copyright &y& Elsevier]

Abstract: Research at plate boundaries generally focuses on the kinematics of plate motions, the genesis of large earthquake, and natural resources. However, often the present state of knowledge is still insufficient to reveal the details of such dynamic systems. To define in more detail the dynamics of, a major plate boundary fault the Coastal Range Fault, eastern Taiwan, we use seismologic and geodetic data to unravel its subsurface geometry and relevant fault behavior. By recomputing the archive triangulation datasets (1917–1978), and using a dislocation algorithm, we determine that it consists of three segments. Our model results are consistent with the observation of the 1951 M7.3 Hualien–Taitung earthquake sequence: that the southern two segments are both oblique-slip faults with significant thrust components, while the northern segment is a pure strike-slip fault. According to our model each segment has the potential to produce an earthquake larger than M7. However, based on GPS observations and seismic records, slip on the southern segment may take the form of aseismic creep and multiple moderate earthquakes, rather than large events with long recurrence interval as occur on the two more northerly segments. With regard to the central segment, the epicenter distribution of catalogued earthquakes occurring after 1951 reveals the presence of a seismic gap, indicating that slip on this segment is predominantly coseismic, with a correspondingly higher potential for large earthquakes in the future. [Copyright &y& Elsevier]

Earthquake early warning (EEW) plays an important role in earthquake disaster mitigation. There are two types of EEW system, regional and on-site. On-site EEW systems analyze initial part of the seismic waves from the P-waves to predict later ground motion from the S-waves and surface waves. In recent years, an on-site EEW method based on fixed peak displacement (Pd) threshold is developed. Although the method consistently provides effective warnings in Taiwan, several studies suggest that its strong filter dependence might introduce extra biases to the system. Also, the fixed Pd threshold suffers from the inevitable trade-off between a false alarm and a missed alarm. In order to overcome the abovementioned problems of fixed Pd threshold method, we utilize techniques in machine learning and develop a new method of on-site early warning. Owing to the property that convoluted-neural-network (CNN) will automatic sampling on different frequencies, the unfiltered seismic signal itself is sufficient to derive a warning threshold. Also, multi-layered neural-network is capable of forming a complex non-linear threshold to reduce both the false alarms and missed alarms. As an example, we collect 21 medium to large inland earthquakes in Taiwan. We compare the performance between the proposed method and the fixed Pd threshold method. The result shows that the proposed method outperforms in both false alarm rate (0.01 percent) and missed alarm rate (30 percent). The proposed method can provide not only significant improvements for on-site EEW but also a window into the initial P-waves. Studying the trained models might even reveal the hidden indicators inside the initial P-waves.Keywords: earthquake disaster mitigation, on-site type earthquake early warning, machine learning [ABSTRACT FROM AUTHOR]

Recent studies have shown that structural damage exhibits stronger correlations with cumulative absolute velocity (CAV) than other ground motion intensity measures (e.g., PGA). This paper presents a CAV site-effect assessment for the Taipei areas for the first time. The study was based on more than 1200 strong-motion data from 47 major earthquakes that had occurred around Taiwan. The results show that the site effects are more conspicuous in the western Taipei than the eastern areas, and it is also obvious that the site effect is strong in locations close to the rims of the basin, where seismic waves could be easily reflected, refracted and superimposed. Subsequently, a map showing the areas in Taipei subject to severe CAV amplification and/or high variability was developed for site-effect microzonation for the study area, on the basis of CAV that was considered better correlated with structural damage under earthquake condition. [ABSTRACT FROM AUTHOR]

In this paper, we study (1) the spatial anomalies and (2) the temporal anomalies of soil gas in northern Taiwan. The spatial anomalies of soil gas are related to tectonic faults, while the temporal anomalies of soil gas are associated with pre-earthquake activities. Detailed soil gas sampling was systematically performed, and the analysis of the collected gas species shows that high helium and nitrogen concentrations appear in samples from specific sites, which coincide with the structural setting of the area studied. This analysis indicates the possibility of using these soil gases to determine fault zones in the studied area. Based on the soil gas data, a station (Tapingti) for automatic soil gas monitoring was constructed on an appropriate site at the fault zone. Some anomalous high radon concentrations at certain times can be identified from the dataset, which was generated by the continuous monitoring of soil gas for over a year. Notably, many of these anomalies were observed several hours to a few days before the earthquakes (M L > 3) that occurred in northern Taiwan. By combining the information of epicenters and fault plane solutions of these earthquakes, we find that the shallow earthquakes (<15 km) were mainly strike-slip and normal-type earthquakes, and concentrated within a distance of 30 km to the monitoring site (Group A). The deep earthquakes (>20 km) were mainly thrust-type earthquakes and distributed in greater distances (>45 km) east of the monitoring site (Group B). Such focal mechanisms of earthquakes suggest an extensional and compressional structural domain in the continental crust for Group A and Group B earthquakes, respectively. It is suggested that the pre-earthquake activities associated with the seismicity of Group B may be transmitted along the major decollement in the region below the Tapingti station, leading to the observed soil gas enhancements. [ABSTRACT FROM AUTHOR]

This paper presents the first whole Taiwan-scale spatial variation of the seismogenic zone using a high-quality crustal seismicity catalog. The seismicity onset and cutoff depths (i.e., seismogenic depths) are determined by the earthquake depth–moment distribution and used to define the upper and lower boundaries of the seismogenic zone, respectively. Together with the published fault geometries and fault area–moment magnitude relations, the depth difference in the onset and cutoff depths (i.e., seismogenic thickness) is used as the fault width to determine the moment magnitudes of potential earthquakes for the major seismogenic faults. Results show that the largest ( Mw 7.9–8.0) potential earthquake may occur along the Changhua fault in western Taiwan, where the seismic risk is relatively high and seismic hazard mitigation should be a matter of urgent concern. In addition, the first-motion focal mechanism catalog is used to examine the relation between the seismogenic depths and earthquake source parameters. For crustal earthquakes (≤ 50 km), the shallowest onset and cutoff depths are observed for normal and strike-slip events, respectively. This observation is different from the prediction of the conventional continental-rheology model, which states that thrust events have the shallowest cutoff depth. Thus, a more sophisticated rheology model is necessary to explain our observed dependence of the seismogenic depths on faulting types. Meanwhile, for intermediate to large crustal ( Mw ≥ 4; depth ≤ 50 km) earthquakes, thrust events tend to occur at the bottom region of the seismogenic zone, but normal and strike-slip events distribute at a large depth range. [ABSTRACT FROM AUTHOR]

Abstract: Variations in the fractal dimension of earthquakes have been suggested to be a precursor of a large earthquake. However, the physical characteristics and seismicity are always different along a large fault system, and it is difficult to segment a large fault for further investigating. Therefore, the fractal dimension of earthquakes on a given fault only reflects the average seismic characteristics of the area and may be unrelated to precursory activity. In this paper, the evolution of the temporal generalized fractal dimension (Dq) of a seismic cluster within a small fault system associated with the Chiayi earthquake (ML =6.4), Taiwan, is investigated. The earthquakes of this sequence are confined in small source volume and reflect the behavior of the local fault system. Our results show that the Dq curve of the background stage is smooth with a low multifractal degree (0.36) and the seismicity is nearly monofractal in the Chiayi region. During the foreshock stage, the seismicity becomes active such that not only the Dq curve exhibits greater variability especially for the steep slope of Dq curve at q=0 but also the temporal fractal dimension changes from nearly monofractal to multifractal. In the aftershock stage, the fluctuation of Dq is large and the multifractal degree is up to 1.0. The temporal multifractal property becomes more pronounced as well. Finally, when the seismicity returns to the background stage, the Dq curve becomes smooth and nearly monofractal again. Therefore, the fluctuations in the Dq spectrum of an earthquake cluster associated with a large earthquake not only give a clear picture of the temporal alterations in the seismogenic regimes but also illustrate their variability through time for a large earthquake. In addition, the variation of the slope of Dq curve at q=0 in the foreshocks may be used as a precursor of a large earthquake. [Copyright &y& Elsevier]

Abstract: We propose a new variant of the sandpile model, the long-range connective sandpile model, by means of introducing randomly internal connections between two separated distant cells. The long-range connective sandpile model demonstrates various self-organized critical states with different scaling exponents in the power-law frequency-size distributions. We found that a sandpile with higher degree of randomly internal long-range connections is characterized by a higher value of the scaling exponent for the distribution, whereas the nearest neighbor sandpile is possessed of a lower scaling exponent. Our numerical experiments on the long-range connective sandpile models imply that higher degree of random long-range connections makes the earthquake fault system more relaxant that releases accumulated energy more easily and produces fewer catastrophic events, whereas lower degree of long-range connections possibly caused by fracture healing very likely motivates accelerating seismicity of moderate events. [Copyright &y& Elsevier]

[Display omitted] • Two earthquakes having almost the same magnitude and focal mechanism in Hualien County, Taiwan, provide different damage scenarios. • The peak ground velocity (PGV) is found a better indicator of damage scenario as compared to peak ground acceleration (PGA) using shakemaps. • Our system provides PGV shakemap that complements the PGA issued by the official agency in Taiwan. • The dense P-Alert network in Taiwan plots near real-time PGA and PGV shakemaps, and these maps are posted regularly on social media. Two earthquakes having almost the same magnitude occurred in the Hualien area of Taiwan in 2018 and 2019. The 2018 earthquake had a magnitude M L 6.2 produced severe destruction; however, the 2019 earthquake (M L = 6.3) did not cause any severe damage. The P-Alert Strong Motion Network provides real-time shakemaps, in addition, to earthquake early warning (EEW) in terms of lead-time. Each instrument provides a different lead-time using peak ground acceleration (PGA) and peak ground velocity (PGV). During both the events, the instruments reported a lead-time of 1.5 to 8.0 s in the epicentral region. This network system also generated high-quality shakemaps during both earthquakes. The shakemaps showed that the higher PGAs are concentrated in the epicentral region for the 2018 and 2019 earthquakes. The lower PGA contour (≥25 Gal) extended to a broader area, including Taipei, during the 2019 earthquake compared to the 2018 earthquake. However, PGV shakemaps display a different pattern. The higher PGV values (more than 17 cm/s) are observed in the epicentral region during the 2018 earthquake (locations suffering building collapse) compared to the 2019 earthquake, suggesting that PGV correlates better with damage distribution as compared to the PGA. The PGV shakemap, currently only available for the P-Alert network, provides crucial information that complements the PGA issued by the official agency in Taiwan. [ABSTRACT FROM AUTHOR]

Earthquake Early Warning (EEW, a life-saving tool) system in India is in its early stage. In recent time, India set up an EEW system in the central seismic gap (CSG) along the Himalayan Belt, which is highly sensitive to the seismic risk with the potential of strong, major and great earthquakes. Almost 200 low-cost P-Alert sensors have been installed for the same. Despite the functioning of this network for last 4 years, it has not recorded any significant earthquake. In absence of any data from this network, we take advantage of recorded earthquake data from Taiwan and synthesized data to test the functionality of EEW in India. In first phase, 100 stations were installed in Garhwal Himalayas of Himalayan belt. We selected Taiwanese stations in good accordance with the Indian sensor network, to have a best fit in terms of inter station spacing. Finally, the recorded waveforms are passed through Earthworm software using tankplayer module. The system performs very well in terms of earthquake detection, P-wave picking, earthquake magnitude and location (using previously estimated regressions). Pd algorithm has been tested where the peak amplitude of vertical displacement are used for estimating magnitudes using previously regressed empirical relationship data. For the earthquakes located in Himalays close to epicenter of historical earthquakes like Uttarkashi (1991) and Chamoli (1999) earthquake, a good estimate of location, as well as magnitude is observed. The approach based on P_d for estimating magnitude works perfectly as compared to τc approach, which is more sensitive to signal to noise ratio. To make it more region specific, we generated synthetic seismograms from the epicenters of historical Chamoli (1999) and Uttarkashi (1991) earthquakes at EEW stations in India and checked the functionality of EEW. While placing these earthquakes within the instrumentation window, a good approximation of earthquake location and magnitude is obtained by passing these generated waveforms. The parameters used to judge the performance of EEW system included the time taken by the system in issuing warning after the confirmation of the occurrence of damaging earthquake and the lead time (time interval between the issuing of warning and arrival of damaging earthquake ground motion at a particular location). High lead times have been obtained for the plainer regions including thickly populated regions of Gangetic plains, such as, Delhi-National Capital Region (NCR) according to the distance from the epicenter, which are the main target of EEW system. [ABSTRACT FROM AUTHOR]