Hybrid adaptive filtration for seismic event location

Abstract

Certain techniques of adaptive filtration, time delay estimation, and minimum variance parameter estimation were synthesized into a hybrid adaptive filtration technique and applied to a problem in seismic event location. Processing seismic wave data generated by strong impulsive source and recorded by a geophone array, we succeeded in doubling the range of detection previously achieved by conventional beamforming methods. The degree of accuracy of bearing estimation was also improved significantly. Additionally a method of observing the nature of the roots of the AR process can replace the conventional power spectrum method for observing wave fronts. The root migration was found to be very sensitive to signal structure.