Study On Space-time Statistical Distribution Models Of Earthquakes In Seismic Hazard Analysis

In recent decades, as the improvement of seismic station network, many countries have accumulated a large number of earthquake catalogs. One of the hot issues of the current seismological studies is how to directly apply earthquake catalogs into the seismic hazard analysis. Seismologists have proposed a variety of the s patially s moothed s eismicity models which d irectly uses ea rthquake catalogs into the s eismic hazard analysis. However, the research on the seismic space-time statistical distribution which is closely related to seismic hazard analysis is not thrived at all. In order to make seismic catalogs be used in the seismic hazard analysis scientifically and reasonably, we systematically and deeply analyze the space-time statistical distribution of Chinese mainland shallow events, and do a quantitative research on the effect of aftershock space-time clustering, which most significantly affects the characteristics of seismic space-time statistical distribution, in seismic hazard analysis.Firstly, w e use s tatistical approach t o analyze t he completeness of th e i nstrumental a nd h istorical destructive e arthquake catalogs, and w e ta ke t he c omplete catalogs as t he b asic data in t he s eismic space-time statistical distribution analysis.Secondly, for ch aracterizing the time s tatistical distribution of ear thquakes, w e ad dress t he study of seismic zo nes. W e use the s pace-time window and s tochastic d eclustering-two aftershock d etected methods, to remove aftershocks in the seismic sequence; and then use χ2test, K-S test and Run test to test the general time in dependence of the main s hock s ubcatalog; furthermore, w e u se D iffusion Entropy Analysis and time K-function to test the time correlation and time clustering of the main shock catalog. The results show that earthquake occurrence in Chinese mainland have the time independence, i.e. the time on which earthquakes occur follows the Poisson process--a typical non-stationary Poisson process.Thirdly, for the r esearch o n the characteristics of space s tatistical d istribution of earthquakes, w e use "cellular s eismology" m ethod, W ilcoxon-test, a nd S peatman Ran k C orrelation test to ve rify t he assumption that is "future earthquakes will occur near past earthquakes". The results show that spatial distribution of seismicity can be used as one of the indicators of zones where future earthquakes are likely to occur. This also means that the spatial distribution of seismicity appears clustering pattern. Furthermore, we adopt spatial K-function point pattern analysis tools and take Neyman-Scott point clustering model as the basic model to study seismic spatial point distribution model, and the result shows that the spatial distribution of the main shocks in China fits the generalized Thomas model.Finally, we calculate the effect of aftershock space-time clustering in seismic hazard analysis. We use the Monte C arlo method, take ETAS model as s eismicity model to s imulate the e arthquake ca talog, an d compute peak ground acceleration by using spatially smoothed seismicity method, to systematically study the effect of aftershock in the seismic hazard. The results show that the mean effect of the aftershock on the peak ground acceleration under the level of10%P. E. in50years is about6%and the maximum impact is about18%. Moreover, the higher the level of probability of exceedance in50years rises, the larger the impact of aftershocks is. In addition, we also study the role of seismicity parameter Mu (the upper1imit o f magnitude) a nd b value in th e im pact o f aftershocks. We find that as M u i ncreases, t he impact of aftershocks in seismic hazard decreases; with the changes of b value, the impact of aftershocks in seismic hazard is almost unchanged. The s tudies in this thesis ca n p rovide th eoretical basis f or s eismic h azard a nalysis directly u sing earthquake c atalog d ata i n o ur country, further deepen th e k nowledge an d u nderstanding of the characteristics of seismic space-time statistical distribution in our country, and give us an insight into the effect of the space-time clustering of aftershock in seismic hazard.