Search

The Formosa array, with 137 broadband seismometers and ∼5 km station spacing, was deployed recently in Northern Taiwan. Here by using eight months of continuous ambient noise records, we construct the first high‐resolution three‐dimensional (3‐D) shear wave velocity model of the crust in the area. We first calculate multi‐component cross‐correlations to extract robust Rayleigh wave signals. We then determine phase velocity maps between 3 and 10 s periods using Eikonal tomography and measure Rayleigh wave ellipticity at each station location between 2 and 13 s periods. For each location, we jointly invert the two types of Rayleigh wave measurements with a Bayesian‐based inversion method for a one‐dimensional shear wave velocity model. All piecewise continuous one‐dimensional models are then used to construct the final 3‐D model. Our 3‐D model reveals upper crustal structures that correlate well with surface geological features. Near the surface, the model delineates the low‐velocity Taipei and Ilan Basins from the adjacent fast‐velocity mountainous areas, with basin geometries consistent with the results of previous geophysical exploration and geological studies. At a greater depth, low velocity anomalies are observed associated with the Linkou Tableland, Tatun Volcano Group, and a possible dyke intrusion beneath the Southern Ilan Basin. The model also provides new geometrical constraints on the major active fault systems in the area, which are important to understand the basin formation, orogeny dynamics, and regional seismic hazard. The new 3‐D shear wave velocity model allows a comprehensive investigation of shallow geologic structures in the Northern Taiwan. Plain Language Summary: Around 137 broadband seismometers have recently been installed with ∼5 km uniform station spacing across the entire Northern Taiwan. The new array, named Formosa array, provides new opportunities in studying detailed subsurface structure associated with geological features in the areas such as basins and faults. Here we first extract coherent seismic surface waves propagating between all station pairs from ambient noise records using seismic interferometry. We then measure the propagation wave speed and the ellipticity of the ground particle motion across the entire array. By using the complimentary sensitivity of the measurements, we construct a high‐resolution 3‐D crustal shear wave velocity model between the surface and ∼12 km depth. Our model illuminates detailed subsurface geometry of the Taipei and Ilan Basins, locations and dips of major active faults in the areas, and possible magmatic structures associated with the Tatun Volcano Group and the active geothermal area in Southern Ilan. Getting clearer images of such shallow structures are important not only to understand the geological processes in an extensional tectonic setting but also to better assess the regional geohazards. Key Points: A new Three‐dimensional shear wave velocity model of Northern Taiwan is constructed using Rayleigh wave H/V ratios and phase velocitiesThe addition of H/V ratio measurements allows detailed shallow crustal structure to be resolved with sub‐kilometer depth resolutionThe resolved fine crustal structures provide new geometrical constraints on basin formation and orogeny dynamics [ABSTRACT FROM AUTHOR]

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]