Digital Filter And Polarization Analysis’s Application Of Data Processing Of Strong Motion Records

Earthquake early warning(EEW) is a emerging method of earthquake preventionand disaster reduction, which has been attracted more attention in recent years.Highsignal-to-noise ratio(SNR) of seismic recording data can greatly achieve the function ofthe non-seismic exclusion, recognition of seismic phase and arrival time and earthquakelocation in EEW system. Therefore, preprocessing the strong motion observation datacan lower noise, improve SNR, playing an important role in the achievement of EEW.This paper focuses on the urgent needs in EEW of getting seismic signal which hashigh SNR. This paper has studied the using of digital filtering and polarization analysisin strong motion observation data noise reduction, epicenter azimuth estimateing.Themajor works can be summarized as follows:1. Introduced the basic principle of a digital filter, studied the common digital filterdesign methods and order, the filter type (acausal, causal) and cut-off frequency, whoseimpact on the filtering effect of strong motion records and SNR. The results indicatethat:2-pole acausal Butterworth filter can be used in strong motion observationrecording processing;2-pole causal Butterworth filter can be used in the EEW datarecords preprocessing; Methods to determine the filter cutoff frequency can be noisespectrum; Digital filtering can play an inportant role in increasing the SNR of recordingsignal.2. The polarization filtering is studied, the method to determine time window isproposed and polarization filtering overall flow chart is given. Proposed a nova idea fornoise reduction by combined using digital filtering and polarization filtering methods.By using artificial signal polarization filtering noise reduction is verified and thesuperposition of train noise and noise on seismic records were verified. The resultsindicated a significant improvement of the SNR afer polarization filtering.3. Based on the Japanese KiK-net strong ground motion records, to analyze theinfluencing factors of computing azimuth using the polarization analysis method. Still todeliberate the recording parameters (acceleration, velocity, displacement), and filteringband impact on the result of calculation azimuth in this way. The studies show that abetter result of epicenter azimuth can be got by the polarization analysis method withthe displacement parameter of records, the1s time window and filtering band of0.1Hzto1Hz.