P-Wave Seismograms From Three Seismic Sources in SW USSR

Abstract

Summary An attempt is made to compute P-wave seismograms to match the seismograms observed at three array stations from three seismic sources (two earthquakes and one possible explosion) in the SW USSR. A rough model of the seismic source is estimated from a visual analysis of the observed and deconvolved seismograms and this initial estimate is refined by trial and error. The method of computation assumes a simplified model of the Earth and seismic source, so exact agreement between computed and observed seismograms cannot be expected; all that can be hoped for from such a model is that the amplitudes, polarities and possibly pulse shapes of the main arrivals (usually P, pP and sP) will be predicted or that, for observed seismograms that are complex showing many overlapping arrivals, the computed seismograms will show similar complexity. Within these limitations of the model, agreement between computed and observed seismograms is obtained for all three sources using a double-couple source model for the earthquake sources and a pure compressional source for the presumed explosion. For one of the earthquakes however, agreement between theory and observation is shown to be improved if the seismograms are assumed to be from two earthquakes separated by about 15 km in epicentre and 0.4 s in origin time. Although the presumed explosion can be modelled by a pure compressional source it is also shown that these theoretical seismograms are virtually indistinguishable from those for a 45 dip slip double-couple source with the same focus. The results suggest that modelling P seismograms to obtain details of earth structure and source orientation is practicable and that the method has advantages over first motion methods for earthquakes of magnitude around in,, 5.