TWO EXAMPLES OF MAXIMUM-LIKELIHOOD FILTERING OF LASA SEISMOGRAMS.
TELEDYNE INDUSTRIES INC ALEXANDRIA VA EARTH SCIENCES DIV
Pagination or Media Count:
Relizable and symmetrical maximum-likelihood filtering was compared to time-shift-and-summation phased summation as array processing procedures, using records from two earthquakes made by seismometers at the Montana Large Aperture Seismic Array LASA. The records were prefiltered to sliminate long period microseismic noise. Within a single subarray seven kilometers diameter, the maximum-likelihood filter improved signal-to-noise ratio approximately 2 db more than the phased sum, if the noise was measured outside the fitting interval. As a method for combining subarray outputs there was no significant difference between the maximum-likelihood filters and phase summation. Approximately two hours of DCD 1604-B machine time were required to calculate the maximum-likelihood filter for each 19-channel subarray. Finally, complexity factors over subarrays and ober LASA were computed for one event and compared with those computed from three additional events and three shots. For the 25 November 1965, event, the complexity factor computed over LASA was 3.74, approximately the same as the mean subarray complexity factor. Author