| Earthquake hazard is one of the most life-threatening situations for people all over the world. Collapse of buildings and other geological hazards are mainly caused by near-fault strong ground motion during the onset of an earthquake. As the country's political economical and cultural center Beijing area is a heavily populated metropolitan region; meanwhile, it is also sitting in a seismically and neotectonically active region. Thus, the task of seismic hazard mitigation is facing a great challenge.In Beijing area, numerous strong earthquakes had occurred during the historic time and had caused enormous economic and life loss. There are many earthquakes reports in our country's acient time, and the faults caused the earthquakes should be still active nowadays. It is possible that some kinds of similar earthquake will happen in the future on the same active fault.This thesis consists of 3 major parts: 1) construction of a three-dimensional (3D) velocity structure model based on a number of geo-information databases. 2) A brief description of the principles and approaches of numerical simulation of strong ground motion using pseudo-spectral time domain method; and 3) through the case study of the simulation of the strong ground motion caused by the historic 1730 Winter Palace Earthquake in Beijing area, to analyze the influence of the distribution of sediments and velocity structure. With the application of earthquake hazard analysis to this simulated scenario results, some scientific findings of earthquake disaster prevention and withstanding are provided.As the first part of thesis, a 3D crust and the top of upper mantle P-wave velocity and geological surfaces structure of Beijing area (115.5°E—117.5°E,39°N—41°N) has been built up using GoCAD, with data from GIS, numeric geological maps, microtremor measurements and wide-angel refraction/ reflection Deep Seismic Sounding (DSS) profiles. I analyzed the geological statistics character of DSS profiles used in our model, and designed a new kind of Smooth Estimate Variation, which is suitable to the constraint weight of Discrete Smooth Interpolation (DSI), to estimate the quality of interpolations in the modeling. By analyzing characters of crust and the uppermost mantle P-wave velocity and comparing our model with seismic tomography results, our model is fine in embodying the structure characters of the crust in Beijing area.Seismic numerical modeling simulates seismic wave propagation and generates ground motion synthetic data at the receiving points on the ground surface, based on the assumed medium and source parameters.As the second part of the thesis I use the pseudospectral time domain (PSTD) method to simulate strong ground motion in Beijing area. The PSTD method uses the Fourier transform to approach spatial partial derivative and use stagger grid in time marching. This method has high precision in spatial domain and only needs two grids per minimum wave length, much less than the FDTD method, which can save computer memory and CPU space. PSTD needs smoothness in source time function and the medium, so I tried to use three kinds of wavelet functions in source modulation and have chosen one of these three for the detailed study developed in Part three of this thesis.In the third part I simulated the 1730 Winter Palace Earthquake in northwest Beijing, which have most abundant reports in the history. At first in accordance with two possible source faults (Qinghe fault in NE direction and Dongbeiwang-Xiaotangshan fault in NW direction), calculate peak ground acceleration (PGA) respectively to estimate the earthquake intensity, and compare the iso-seismal to the PGA distribution. I can conclude that Qinghe fault is the seismic genetic fault in this earthquake. Than by simulation in different underground velocity structure, I confirm the significant influence of sedimentation to seismic wave propagation, and the result shows the sediment amplification to the acceleration of seismic wave, and the depth of Quaternary deposit is very important to the magnification.
|
PDF Full Text Request