No abstract

<p>Modeling the amplitude spectra based on the source term, the path one and site ones for 54 sites located in and around the Taipei basin is the aim of this study. The site term includes the amplification function varied with frequency and the site-specific parameter (k0). The amplification functions for Class-C, -D, and -E site are from Huang et al. (2007) for the central Taiwan. Meanwhile, the amplification function for Class-B site can be referred to Boore and Joyner (1997). The root-mean-squared spectral amplitudes of two horizontal shear waves after three-point smoothing from the observed seismograms are compared to the synthetic amplitude spectra. The goodness of fit coefficient (GFC) and the residual errors (ERR) are calculated for concluding the fitness of the modeling amplitude spectra. Results show both the GFC and ERR of stations are varied with the earthquake magnitude and hypo-central distance. The averaged GFC are larger than 0.8 for 42 stations. Meanwhile, there are 12 station with averaged GFC smaller than 0.8. Besides, the ERRs of 28 stations are less than 0.5. Meanwhile, there are 18 stations with ERRs in the range of 0.5-0.6. The obtained results may be used for modeling the amplitude spectra for the Taipei area. The more accurate amplitude spectra can be improved by updating the parameters utilized in the source-, the path- and the site terms.</p>

Taiwan is located on the boundary between the Eurasia Plate and Philippine Sea Plate, which is a very high seismicity rate area. We begin calculating the earthquake occurrence probability using four recurrence models to mitigate seismic disasters. We focus on estimating the occurrence probabilities for regional earthquake sources based on the catalog released by the Central Weather Bureau over the period from 1900 to 2011. According to the tectonic and seismicity characteristics areas in and around Taiwan are divided into several zones for shallow and deep earthquakes. We utilize four recurrence models to estimate the earthquake occurrence probabilities over the next 30 and 50 years, respectively. In addition, the grid-based probabilities in 0.1° × 0.1° spatial size are calculated using the micro-zoning method. The results obtained from four recurrence models show that areas with high values over the next 30 and 50 years are correlated with two subduction zones and a suture zone. High probabilities in the western foothills appear highly active tectonic. Moreover, the high values appear at in Eastern Taiwan, offshore Hualien County. There are discrepancies between the results from the four models. The highest grid-based probability is about 3.0, 3.5, 2.5 and 3.5% for the Lognormal, Gamma, Exponential, and Weibull models, respectively. The inland probabilities are below 0.5% for the results from Lognormal, Gamma, and Weibull models. Even so, the results from the Exponential model are upmost in the range from 0.5 -1.0%.

No abstract