Study On Frequency Spectrum Characteristics Of Different Type Events Record By Broadband Seismograph

With broadband digital seismograph widely used in earthquake monitoring system, which further improve the quality of the waveform record, for us to better analyze stations recorded event information characteristics provide the necessary conditions, but only from the broadband digital seismograph recorded the data characteristic unavoidably exist subjective visual interpretation of event.Traditional analysis and processing method of non-stationary signal mainly include: short-time Fourier transform and Wigner-Ville distribution and wavelet transform, etc., but these three methods are essentially is based on Fourier transform, as the Fourier analysis is mainly suitable for steady state signal processing, therefore, can not fundamentally solve the limitations of Fourier analysis. And Hilbert-Huang transform is put forward in 1998, a new method of non-stationary signal processing is essentially to smooth processing of non-stationary signal, the real signal in different scale fluctuation or trend into step by step, finally using the instantaneous frequency and energy to represent the frequency content of the original signal, instead of Fourier spectrum analysis of global frequency and energy, this avoids based on Fourier transform of the need to use many false harmonic expression of nonlinear and non-stationary signal.Because seismic wave data are non-stationary and nonlinear signals, this paper adopts the Hilbert-Huang transform broadband digital seismograph records to the data, using the Hilbert marginal spectrum in spectrum signature of the energy distribution of spectrum shows that different events, the analysis of the results can describe and distinguish between blasting, natural seismic events, collapse events, as well as the advantages of the station background noise frequency distribution.In this paper, we search, collect and select the Shaanxi, Gansu, Ningxia, Inner Mongolia border area has been largely determined by the magnitude equivalent artificial blasting event and waveform recording collapse events ML2.0 above, as well as a small portion of the waveform record small natural seismic events, by repeating the calculation analysis, determine the selection of each event in the epicenter near the waveform data(local earthquakes) station records. Selection seismic phase more clearer stations recording waveform, the interception of the original waveform from 5-10 seconds before lift off started 100 seconds recording length. The seismometer stations with sampling rate of 100 times / sec. The use of the Hilbert- Huang Transform different stations on the current broadband digital seismographs recorded the different types of events to process raw waveform data, found in the work, Hilbert marginal spectrum is relatively straightforward, although the reaction Angle is different, also can better response signal frequency components in the real situation. It is mainly comparative analysis of different events Hilbert marginal spectrum, the spectrum for its own characteristics were detailed in-depth analysis and possible causes of the phenomenon, the feature event of different comparative study further illustrate the spectrum signature of different types of events, summarized the blasting events, seismic events, spectrum signature distinguish collapse events in quantitative distinction between events and seismic events blasting collapse events provide more reliable evidence for future work accurately distinguish blasting, natural earthquake, collapse provided the necessary theoretical basis, but also for the further detailed study of low frequency pulsing characteristics laid a good foundation, the next step will focus on analysis of records for low frequency events.However, it is worth noting that most of the event analysis and calculation of the present study only in the stratum structure Shaanxi, Gansu, Ningxia, Inner Mongoliaborder area carried out analyze and calculate, there may be some limitations, it is necessary to further collect information in the future work, conduct more comprehensive calculations, and better comparative analysis of the spectrum signature of different event types whether as a result of changes in regional differences or not.