Che-Min Lin

The free-field strong earthquake stations of the Taiwan Strong Motion Instrument Program (TSMIP), distributed throughout Taiwan, have provided rich and high-quality seismic data over the years. Its data indirectly enhances Taiwan's earthquake-prevention and disaster-reduction capabilities. Due to the importance of seismic site effects in the research of strong-ground motion, the National Center for Research on Earthquake Engineering (NCREE) cooperated with the Central Weathering Administration (CWA) to carry out the Engineering Geological Database for TSMIP (EGDT) containing three site parameters, including Vs30, Z1.0, and κ0, and site classification. However, the CWA gradually renewed TSMIP stations since 2017, leading to the establishment and elimination of some stations. The new stations lack any earthquake site characteristics or classification information, which will make future research and the application of various strong earthquakes difficult. Besides, while Taiwan has produced numerous site characteristic studies based on microtremor or earthquake HVSR analysis, a comprehensive and systematic study of HVSR site characteristics for TSMIP strong-motion stations has been lacking. This study uses continuous seismic data from the GDMS-2020 database of CWA to extract seismic ambient noise and earthquake data for assessing site characteristics and the predominant frequencies of each TSMIP station using HVSR analysis. By establishing representative HVSRs for different site classifications using stations with well-known site parameters, we evaluated the site classifications of new stations without site information. Finally, an HVSR site database of the new TSMIP network is established and comprehensively discussed.

A timely, high-resolution earthquake catalog is crucial for estimating seismic evolution and assessing hazards. This study aims to introduce a deep-learning-based real-time microearthquake monitoring system (RT-MEMS) for Taiwan, designed to provide rapid and reliable earthquake catalogs. The system integrates continuous data from high-quality seismic networks via SeedLink with deep learning models and automated processing workflows. This approach enables the generation of an earthquake catalog with higher resolution and efficiency than the standard catalog announced by the Central Weather Administration, Taiwan. The RT-MEMS is designed to capture both background seismicity and earthquake sequences. The system employs the SeisBlue deep learning model, trained with a local dataset, to process continuous waveform data and pick P- and S-wave arrivals. Earthquake events are then associated and located using a modified version of PhasePAPY. Three stable RT-MEMS have been established in Taiwan: one for monitoring background seismicity along a creeping fault segment and two for monitoring mainshock–aftershock sequences. The system can provide timely information on changes in seismic activity following major earthquakes and generate long-term catalogs. The refined catalogs from RT-MEMS contribute to a more detailed understanding of seismotectonic structures and serve as valuable datasets for subsequent research.

The resonant frequency and site amplification of Horizontal-to-Vertical Spectral Ratio (HVSR) for a site have become common and useful site parameters. The present site database for the strong-motion stations of TSMIP (Taiwan Strong Motion Instrumentation Program) in Taiwan only included conventional parameters, including the averaged shear-wave velocity (VS) of the upper 30 meters (VS30), the depth to the horizons of shear-wave velocity larger than 1.0 km/s (Z1.0), and the high-frequency decay parameter (kappa; κ0). In order to complete the HVSR-based site parameters of the database, the earthquake-based HVSR (eHVSR) and microtremor-based HVSR (mHVSR) are conducted for the TSMIP stations within the site database based on the standard procedures suggested by recent studies. The characteristics of eHVSRs and mHVSRs for different VS30-based site classifications are evaluated and discussed. The relationships between HVSR-based parameters, including resonant frequency and amplification, and conventional parameters are systematically assessed. Although the differences between eHVSR and mHVSR are non-negligible for some stations, the results still show high correlations between their derived site parameters and realistic site conditions. The HVSR-based site parameters of the TSMIP stations would be reliable for further studies of strong motion prediction and simulation, seismic hazard analysis, etc, in Taiwan.

Publishing timely warnings of areas potentially affected by landslides and subsequent events remains challenging in hazard assessment. Recently, seismic monitoring has emerged as a promising technique for landslide detection and location determination. However, the effectiveness of this approach is limited by location errors. To overcome this problem, we propose a hybrid method (a three-step process) through seismic analysis and consideration of prior geomorphological characteristics of areas and aspects in a landslide inventory to constrain the potential landslide locations. The first analysis entails a grid-based single-force inversion of 0.02–0.05 Hz seismic signals by spatial grid spacings of 0.2° and 0.05° to retrieve the single-force direction, the estimated landslide area, and a wide search range for a given source. The second analysis involves two location methods for high-frequency seismic signals (>1 Hz) with a finer spatial grid spacing of 0.01°, gradually reducing the landslide search region. By integrating seismology-determined information with a landslide inventory, potential landslide locations and warning areas could be identified effectively. Then we apply the proposed the three-step analysis method to the 2021 Silabaku landslide in southern Taiwan. The results showed that the seismic techniques we used progressively condensed the initial search range from 1831 to 770 km2, converging to a small area of 99 km2. The estimated landslide area was 0.80±0.04 km2, and the force direction was 154.09° ± 6.76°. Subsequently, according to the earlier information, we identified three possible landslide source locations from the 2020 landslide inventory. These findings have implications for warning for two road sections, one of which was subsequently damaged by debris flow after the Silabaku landslide. This study highlights the effectiveness of the hybrid three-step algorithm in constraining areas potentially affected by landslides, facilitating timely assessments for warnings in southern Taiwan.

A new methodology is proposed for developing a scenario‐based site‐specific response spectrum (RS) considering near‐fault effects in Taiwan. First, source parameters, together with reference rock site conditions, are defined according to the available geological and geophysical information at a target site close to a potential active fault in northern Taiwan. Secondly, the scenario‐based response spectrum for a reference rock site condition is developed theoretically through an empirical approach by using a ground motion prediction equation (GMPE). The effect of the pulse period and the occurrence probability of near‐fault pulse‐like ground motion on RS is evaluated by using the ground motion simulation (GMS) technique, in which the stochastic finite‐fault simulation method is validated and applied for evaluating velocity pulse. Third, site‐specific site amplification is incorporated into RS through a site transfer function calculated from the measured horizontal‐to‐vertical Fourier spectral ratio through the microtremor (MHVR) of the target site. Finally, the design spectrum of the target site is compared with the derived site‐specific RS to evaluate the impact of the neighbor fault on the structure of the target site.

Studies have indicated that the power of seismic waves, rather than the acceleration and velocity, strongly correlates with the level of structural damage and the shutdown of high‐tech equipment during high‐precision process. However, current earthquake early warning (EEW) systems in Taiwan issue a warning only for a seismic wave with high acceleration or velocity but not for that with high power. To extend the application of the EEW system to high‐tech fields, conventional and novel hybrid prediction models were developed in this study to predict the ground motion power and to issue EEWs for high‐power ground motions. A conventional prediction model that relies on earthquake source parameters was first developed to determine ground motion characteristics, and a hybrid prediction model that relies not only on earthquake source parameters but also on the on‐site ground motion parameter of P‐waves was developed to improve the prediction accuracy. The two‐step maximum likelihood method was used to derive model coefficients for considering the mixed effect of the ground motion records. The source, path, and site effects on the short‐period power and the long‐period power of ground motions were illustrated by the proposed models. A method for determining a broadcast scenario of EEW systems for high‐power ground motions with a target success rate and damage threshold was also developed. These novel prediction models can be used to develop a customized EEW system to provide an early warning for high‐power seismic waves in Taiwan.

Broadband seismometers, ground-based Global Navigation Satellite Systems (GNSS), and magnetometers that were located within an epicentral distance of approximately 150 km consistently observed the novel anomalous behaviors of the common-mode ground vibrations approximately 5–10 days before the M6.6 Meinong earthquake in Taiwan. The common-mode ground vibrations with amplitudes near 0.1 m at frequencies ranging from 8 × 10⁻⁵ to 2 × 10⁻⁴ Hz were generated near the region close to the epicenter of the impending earthquake. The common-mode vibrations were consistently observed in seismic and GNSS data associated with five other earthquakes in four distinct areas. The results reveal that the common-mode vibrations could be a typical behavior before earthquakes. The causal mechanism of common-mode vibrations can be attributed to crustal resonance excitations before fault dislocations occur. Potential relationships with other pre-earthquake anomalies suggest that the common-mode vibrations could be ground motion before earthquakes, which was investigated for a significant length of time.