Discussion On Dynamic Corner Frequency Of Source Spectrum

The study on strong ground motion prediction involves the understanding and description of the source,the propagation of seismic waves in the medium and the influence of local site conditions.For the synthesis of broadband near-fault ground motion,it is one of the key points to establish a more reasonable source spectral model,and the corner frequency is one of the parameters.Thus,the improvement of the dynamic corner frequency is discussed in this thesis,and the influence of the corresponding energy conservation factor?or scaling factor?on the synthetic ground motion is analyzed,which is compared with strong ground motion records.The main work is as follows.?1?In the northeast region of Japan?36°N-40°N,138°E-143°E?as the research area,the data were divided into M_W=3.5-4.5,M_W?3.5 and M_W?4.5 by collecting F-net and K-NET data.Tn the inversion strategy,the velocity Fourier amplitude spectrum is used as the target spectrum,and the parameters required for the discussion on the dynamic corner frequency are inversed,including the corner frequency f₀ and parameter aof the source spectral model,?₀,the parameters of quality factor Q₀ and?.?2?The improvement of dynamic corner frequency.The parameter of the corner frequencyN_Rij is redefined as the number of rupturing sub-faults,rather than that of ruptured ones.Sub-fault dimensions are given as 5km×5km,7.5km×7.5km,10km×10km and 20km×20km.Ground motion on 51 K-NET bedrock stations are synthesized randomly by a finite fault model.The results show that the effect of sub-faults size is reduced to a certain extent,compared with our previous studies,the improvement is more obvious in the periods of 0.1-1s;by comparing with the response spectra of strong ground motion records,the long-period parts fit better,and the short-period parts with period less than 0.1s are lower than the observed spectra,while for several stations,the synthesized spectra fit with the observed ones over the periods to 10s.?3?An energy conservation factor is considered further.The introduction and development process of this factor are summarized,and the models of Motazedian and Atkinson?2005?and Boore?2009?are added to the source spectral model in Chapter 3.The results show that the dependence on sub-fault size is further reduced,and the synthesized spectra fit with the observed ones on the periods of less than 0.1s.Finally,we summarize all the work and draw relevant conclusions,and point out the shortcomings in the work and the prospects for future work.