Open AccessRecent Improvements in the Analysis of Surface Wave Observations
Author(s) -
Landisman M.,
Dziewonski A.,
Satô Y.
Publication year - 1969
Publication title -
geophysical journal of the royal astronomical society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.302
H-Index - 168
eISSN - 1365-246X
pISSN - 0016-8009
DOI - 10.1111/j.1365-246x.1969.tb00246.x
Subject(s) - seismogram , attenuation , acoustics , time domain , dispersion (optics) , optics , group velocity , refraction , phase velocity , filter (signal processing) , frequency domain , rayleigh wave , slowness , signal (programming language) , noise (video) , geology , physics , surface wave , mathematics , seismology , mathematical analysis , computer science , computer vision , programming language , artificial intelligence , image (mathematics)
Summary Moving window analysis of seismic transients can be used to measure group velocity in recordings of multi‐mode propagation. It converts an ordinary time series (amplitude vs time) into a two‐dimensional display of decibels on a velocity vs log period plane. The effects of excitation and attenuation may be studied for each of the modes observed. Application of this method of analysis to the recordings of mutually orthogonal detectors permits the measurement of lateral refraction, the separation of Love waves from Rayleigh waves and promises to be useful in the identification of the fundamental and higher modes. The modes received at the recording site may be separated from the observed composite signal by means of time‐variable filters. A significant improvement in signal‐to‐noise can be obtained if these filters are based upon the dispersion inferred from moving window analysis. Filtration in the frequency domain is described and compared to filtration in the time domain. Inter‐station phase velocity measurements from small events may be extended to periods previously measured only for great circle passages of wave trains from major events. In order to achieve these improvements (a) moving window analysis should be used to measure the dispersion, (b) a frequency‐domain time‐variable filter based upon the observed dispersion should be used to extract the desired mode from the records of both stations, and (c) the phase velocity should be calculated from the windowed crosscorrelogram of these filtered seismograms. The inter‐station cross‐correlogram computed from matched recordings may be considered an approximate representation of the response of the inter‐station elastic medium to an impulse in space and time, applied at the first station and recorded by a distorsionless displacement metre at the second station. Inter‐station group velocity thus can be evaluated by moving window analysis of the cross‐correlogram, without resort to numerical differentiation of the observed phase velocity curve.