Search

Monitoring precursory decline in groundwater radon at the Antung hot spring is a useful means of forecasting the magnitude and precursor time of local disastrous earthquakes. With the help of a case study in southeastern Taiwan, this paper demonstrates the effect of tectonic setting in the subduction zone on the correlation between radon decline, precursory time and earthquake magnitude. Given a radon-monitoring site located near the plate boundary in the tectonic setting of advanced arc-continental collision, the observed radon decline and precursory time prior to the earthquakes in the tectonic setting of initial arc-continental collision are smaller than those observed prior to the earthquakes occurring on the plate boundary in the tectonic setting of advanced arc-continental collision. In the advanced arc-continental collision state, the coupling between the plates is strong and the stress transfer is efficient, whereas in the incipient collision state, the coupling and stress transfer are not as good. It also takes additional time lag and attenuation for the stress transfer from one tectonic setting to the other. This paper presents the difference in the precursory behavior of groundwater radon between earthquakes which occurred in two different tectonic settings: advanced and initial arc-continental collision. [ABSTRACT FROM AUTHOR]

In this paper we consider the statistics of the aftershock sequence of the m = 7.65 20 September 1999 Chi-Chi, Taiwan earthquake. We first consider the frequency-magnitude statistics. We find good agreement with Gutenberg-Richter scaling but find that the aftershock level is anomalously high. This level is quantified using the difference in magnitude between the main shock and the largest inferred aftershock $$ {{\Updelta}}m^{ *}. $$ Typically, $$ {{\Updelta}}m^{ *} $$ is in the range 0.8-1.5, but for the Chi-Chi earthquake the value is $$ {{\Updelta}}m^{ *} $$ = 0.03. We suggest that this may be due to an aseismic slow-earthquake component of rupture. We next consider the decay rate of aftershock activity following the earthquake. The rates are well approximated by the modified Omori's law. We show that the distribution of interoccurrence times between aftershocks follow a nonhomogeneous Poisson process. We introduce the concept of Omori times to study the merging of the aftershock activity with the background seismicity. The Omori time is defined to be the mean interoccurrence time over a fixed number of aftershocks. [ABSTRACT FROM AUTHOR]

- In this paper, the site characteristics of the Dahan downhole array are studied by analyzing the September 21, 1999M7.3 Chi-Chi earthquake sequence including the main shock and some aftershocks. The four-level array (0 m, 50 m, 100 m and 200 m) is located to the north of Hualien City in eastern Taiwan. Polarization analysis is used to check the orientation errors of the seismometers at different levels of depth. If the surface instrument is chosen as reference, the angle between the major polarization axes of the surface and any downhole records is the orientation error that must be corrected for the downhole accelerographs. The orientation errors at depths of 50 m, 100 m and 200 m are 32°, 120° and -84°. After the corrections, the coherency between the surface and downhole records is substantially improved. Spectral ratio analysis shows that the predominant frequency of the Chi-Chi main shock shifts to a lower frequency. We also simulate ground motions at different depths by using the Haskell method with a linear velocity structure model. The record at surface is chosen as the input motion. Compared with the observed data, ground acceleration can be well reproduced for the aftershocks (weak-motion events) of the September 21, 1999M7.3 Chi-Chi earthquake. However, for the Chi-Chi main shock, the synthetic waveform cannot match well with the observation neither in amplitude nor in phase. This indicates that large ground shaking probably induced the nonlinear site effect at that time, and the model used cannot support it. [ABSTRACT FROM AUTHOR]

We analyze high-resolution borehole strainmeter signals to constrain the fault plane solutions and locations of the 2009 Fengpin-Hualien earthquakes, in eastern Taiwan. The December M w 6.6 earthquake has ruptured a narrow section of the Collision Seismic Zone (within depth of 34 to 44 km), an active structure located offshore northeast Taiwan. The October M w 6.3 event has ruptured the deep section of the Longitudinal Valley fault (25–32 km depth) located between Fengpin and Fanglin. Coulomb stress changes induced by the October event in the hypocenter region of the December event are low (< +30 kPa), which suggests that the latter event was not induced by static stress transfer. The December event is characterized by an abundant seismicity for about 2.5 days following the mainshock. The temporal evolution between the aftershock rate and geodetic strain suggests that seismicity could have been controlled by a short-lived frictional afterslip. Besides, we infer that significant positive static Coulomb stress changes (> +100 kPa) induced by this sequence at shallow to moderate depths in the Longitudinal Valley (0–12 km) may have influenced the occurrence of large mainshocks in eastern Taiwan during the past decade. Conversely, limited stress (< +30 kPa) is transferred to the shallow, locked section of the southernmost Ryukyu trench. This study emphasizes the importance of borehole strain measurements in active regions to enhance detection of seismic and aseismic events unresolved by surface geodesy. [ABSTRACT FROM AUTHOR]

Distribution of parameters characterizing soil response during the 1999 Chi-Chi, Taiwan, earthquake ( M w = 7.6) around the fault plane is studied. The results of stochastic finite-fault simulations performed in P avlenko and W en (2008) and constructed models of soil behavior at 31 soil sites were used for the estimation of amplification of seismic waves in soil layers, average stresses, strains, and shear moduli reduction in the upper 30 m of soil, as well as nonlinear components of soil response during the Chi-Chi earthquake. Amplification factors were found to increase with increasing distance from the fault (or, with decreasing the level of “input” motion to soil layers), whereas average stresses and strains, shear moduli reduction, and nonlinear components of soil response decrease with distance as ~ r −1 . The area of strong nonlinearity, where soil behavior is substantially nonlinear (the content of nonlinear components in soil response is more than ~40–50% of the intensity of the response), and spectra of oscillations on the surface take the smoothed form close to E(f) ~ f −n , is located within ~20–25 km from the fault plane (~ 1/4 of its length). Nonlinearity decreases with increasing distance from the fault, and at ~40–50 km from the fault (~ 1/2 of the fault length), soil response becomes virtually linear. Comparing soil behavior in near-fault zones during the 1999 Chi-Chi, the 1995 Kobe ( M w = 6.8), and the 2000 Tottori (Japan) ( M w = 6.7) earthquakes, we found similarity in the behavior of similar soils and predominance of the hard type of soil behavior. Resonant phenomena in upper soil layers were observed at many studied sites; however, during the Chi-Chi earthquake they involved deeper layers (down to ~ 40–60 m) than during lesser-magnitude Kobe and Tottori earthquakes. [ABSTRACT FROM AUTHOR]

A systematic analysis of the source duration (τ) and seismic moment (M0) for seismogenic earthquakes (MW 5.5–7.1) in the Taiwan region was completed by using a teleseismic P-wave inversion method. Irrespective of the source self-similarity, the M0–τ relationship derived in this study had a power-law form, namely M0 ∝ τ3, under the assumption that ΔσVr3 is constant following a circular fault model (Δσ: static stress drop; Vr: rupture velocity). For Taiwan's earthquakes, the derived M0–τ relationship not only provides information to predict the source duration of large earthquakes, but also probes the rupture features of seismogenic earthquakes. That is, there are different rupture patterns for earthquakes, but the product ΔσVr3 remains nearly constant. [ABSTRACT FROM AUTHOR]

Some researchers have suggested that significant electromagnetic changes that can be detected prior to strong earthquakes are associated with the tectonics and subsurface structures beneath monitoring stations. Numerical simulation seems to support this phenomenon. However, to date, few field measurement data are available to address this issue. In this work, we used the shifting correlation method to analyze 2-month recordings of 20 geoelectric stations in Taiwan and found at least two stations had recorded geoelectric signals significantly correlated with known seismic events. These two stations are installed on large reverse faults, which are probably characterized by high stress concentrations before earthquakes. In addition, they share the same resistivity variation pattern, namely, a conductor layer sandwiched between two high-resistivity layers. Moreover, both stations are above the electrical resistivity boundaries, which means they have an increased probability of recording the abnormal disturbance once EM signals have been emitted from EQs. Our analysis results support the viewpoint that lateral and vertical electrical resistivity variation and conductors in the subsurface may be able to amplify seismic electric signals. [ABSTRACT FROM AUTHOR]

The main purpose of this study is to propose an innovative methodology to forecast the cumulative probability of future larger earthquakes for any given magnitude. It is based on applying an innovative approach to explicitly incorporate the logarithmic mean annual seismicity rate and its standard deviation, replacing the conventional Gutenberg–Richter (G–R) relation, which is only expressed by the arithmetic mean. The new representation of the G–R relation can provide the median annual seismicity rate and upper and lower bounds of recurrence time period for future larger earthquakes in different regions of Taiwan. Subsequently, the logarithmic mean is found to have a more well-behaved lognormal distribution. The selected crustal earthquake data for 3.0 ≤ Mw ≤ 5.0 are used to obtain alternative Gutenberg–Richter relations for different regions. The results are as follows: log 10 N = 5.74 - 1.07 M w ± (- 0.18 + 0.12 M w) in and Taiwan; log 10 N = 5.08 - 1.07 M w ± (0.23 + 0.05 M w) for northeastern Taiwan offshore; log 10 N = 5.48 - 0.95 M w ± (- 0.32 + 0.14 M w) for eastern Taiwan offshore; log 10 N = 4.57 - 0.84 M w ± (0.07 + 0.07 M w) for southeastern Taiwan offshore. These results can be used for preventing and mitigating seismic hazards. [ABSTRACT FROM AUTHOR]

Water-level data recorded at a sensitive well next to a fault in central Japan between 1989 and 1998 showed many coseismic water-level drops and a large (60 cm) and long (6-month) pre-earthquake drop before a rare local earthquake of magnitude 5.8 on 17 March 1997, as well as 5 smaller pre-earthquake drops during a 7-year period prior to this earthquake. The pre-earthquake changes were previously attributed to leakage through the fault-gouge zone caused by small but broad-scaled crustal-stress increments. These increments now seem to be induced by some large slow-slip events. The coseismic changes are attributed to seismic shaking-induced fissures in the adjacent aquitards, in addition to leakage through the fault. The well’s high-sensitivity is attributed to its tapping a highly permeable aquifer, which is connected to the fractured side of the fault, and its near-critical condition for leakage, especially during the 7 years before the magnitude 5.8 earthquake. [ABSTRACT FROM AUTHOR]

Streamflow recorded by a stream gauge located 4 km from the epicenter of the 1999 M7.6 Chi-Chi earthquake in central Taiwan showed a large and rapid anomalous increase of 124 m3/s starting 4 days before the earthquake. This increase was followed by a comparable co-seismic drop to below the background level for 8 months. In addition, groundwater-levels recorded at a well 1.5 km east of the seismogenic fault showed an anomalous rise 2 days before the earthquake, and then a unique 4-cm drop beginning 3 h before the earthquake. The anomalous streamflow increase is attributed to gravity-driven groundwater discharge into the creek through the openings of existing fractures in the steep creek banks crossed by the upstream Shueilikun fault zone, as a result of pre-earthquake crustal buckling. The continued tectonic movement and buckling, together with the downward flow of water in the crust, may have triggered the occurrence of some shallow slow-slip events in the Shueilikun and other nearby fault zones. When these events propagate down-dip to decollement, where the faults merges with the seismogenic Chelungpu fault, they may have triggered other slow-slip events propagating toward the asperity at the hypocenter and the Chelungpu fault. These events may then have caused the observed groundwater-level anomaly and helped to trigger the earthquake. [ABSTRACT FROM AUTHOR]

A stress reconstruction was performed based on focal mechanisms around the Longmen Shan region prior to the 2008 Ms 8.0 Wenchuan earthquake using a newly developed algorithm (known as MCA). The method determines the stress tensor, including principal axes orientations, and quantitative stress values, such as the effective confining pressure and maximum shear stress. The results of the MCA application using data recorded by the regional network from 1989 to April 2008 show the background stress state around the Longmen Shan belt before the Wenchuan earthquake. The characteristics of the stress orientation reveal that the Longmen Shan region is primarily under the eastward extrusion of the eastern Tibetan plateau. Non-uniform quantitative stress distributions show low stress levels in the upper crust of the middle Longmen Shan segment, which is consistent with the observed high-angle reverse faulting associated with the 2008 Wenchuan earthquake. In contrast, other study areas, such as the Bayankela block and the NW strip extending to the Sichuan basin, show high stress intensity. This feature coincides with heterogeneity in the wave speed image of the upper crust in this region, which shows high S-wave speed in the high stress areas and comparatively low S-wave speed in low stress areas. Deformation features across the Longmen Shan belt with the slow rates of convergence determined by GPS and the distribution of surface deformation rates also are in keeping with our stress results. We propose a dynamic model in which sloping uplift under the Longmen Shan, which partly counteracts the pushing force from the eastern plateau, causes the low-quantitative stresses in the upper crust beneath the Longmen Shan. The decreasing gravitational potential energy beneath the Longmen Shan leads to earthquake thrust faulting and plays an important role in the geodynamics of the area that results from ductile thickening of the deep crust behind the Sichuan basin, creating a narrow, steep margin. [ABSTRACT FROM AUTHOR]

The 2014 ML 5.9 Fanglin earthquake occurred at the northern end of the aftershock distribution of the 2013 ML 6.4 Ruisui event and caused strong ground shaking and some damage in the northern part of the Longitudinal Valley. We carried out the strong-motion simulation of the 2014 Fanglin event in the broadband frequency range (0.4–10 Hz) using the empirical Green’s function method and then integrated the source models to investigate the source characteristics of the 2013 Ruisui and 2014 Fanglin events. The results show that the dimension of strong motion generation area of the 2013 Ruisui event is smaller, whereas that of the 2014 Fanglin event is comparable with the empirical estimation of inland crustal earthquakes, which indicates the different faulting behaviors. Furthermore, the localized high PGV patch might be caused by the radiation energy amplified by the local low-velocity structure in the northern Longitudinal Valley. Additional study issues are required for building up the knowledge of the potential seismic hazard related to moderate-large events for various seismogenic areas in Taiwan. [ABSTRACT FROM AUTHOR]

We use preseismic, coseismic, and postseismic GPS data of the 1999 Chi-Chi earthquake to infer spatio-temporal variation of fault slip and frictional behavior on the Chelungpu fault. The geodetic data shows that coseismic slip during the Chi-Chi earthquake occurred within a patch that was locked in the period preceding the earthquake, and that afterslip occurred dominantly downdip from the ruptured area. To first-order, the observed pattern and the temporal evolution of afterslip is consistent with models of the seismic cycle based on rate-and-state friction. Comparison with the distribution of temperature on the fault derived from thermo-kinematic modeling shows that aseismic slip becomes dominant where temperature is estimated to exceed 200° at depth. This inference is consistent with the temperature induced transition from velocity-weakening to velocity-strengthening friction that is observed in laboratory experiments on quartzo-feldspathic rocks. The time evolution of afterslip is consistent with afterslip being governed by velocity-strengthening frictional sliding. The dependency of friction, μ, on the sliding velocity, V, is estimated to be $${{\partial \mu }/{\partial \, {\rm ln}\, V}} = 8 \times 10^{ - 3}$$. We report an azimuthal difference of about 10–20° between preseismic and postseismic GPS velocities, which we interpret to reflect the very low shear stress on the creeping portion of the décollement beneath the Central Range, of the order of 1–3 MPa, implying a very low friction of about 0.01. This study highlights the importance of temperature and pore pressure in determining fault frictional sliding. [ABSTRACT FROM AUTHOR]

Gas variations of many mud volcanoes and hot springs distributed along the tectonic sutures in southwestern Taiwan are considered to be sensitive to the earthquake activity. Therefore, a multi-parameter automatic gas station was built on the bank of one of the largest mud-pools at an active fault zone of southwestern Taiwan, for continuous monitoring of CO2, CH4, N2 and H2O, the major constituents of its bubbling gases. During the year round monitoring from October 2001 to October 2002, the gas composition, especially, CH4 and CO2, of the mud pool showed significant variations. Taking the CO2/CH4 ratio as the main indicator, anomalous variations can be recognized from a few days to a few weeks before earthquakes and correlated well with those with a local magnitude >4.0 and local intensities >2. It is concluded that the gas composition in the area is sensitive to the local crustal stress/strain and is worthy to conduct real-time monitoring for the seismo-geochemical precursors. [ABSTRACT FROM AUTHOR]