TELEDYNE INDUSTRIES INC ALEXANDRIA VA EARTH SCIENCES DIV
Using a series of 19 explosions detonated within a 2500 sq km area of the Nevada Test Site, the effectiveness is demonstrated of applying predetermined travel-time anomalies to a limited network of teleseismic stations comprised of between 4 and 13 stations greater than 1900 km distance. Three different travel-time tables were used Jeffrey-Bullen Herrin, 1961 version and Herrin, November 1966 version and two different computer programs LOCATE and SHIFT, the former which minimizes the sum of squares of residuals and the latter which minimizes the sum of squares of relative residuals. The mean location error obtained without time anomalies is about 26 km, and with anomalies is less than 3 km, regardless of travel-time table and regardless of program. It is further demonstrated that neither the number of stations nor the distance aperture of the network has an effect on the location error, although the azimuth aperture does. Confidence estimates are made in three ways the standard confidence ellipses maximum-relative-error polygons and standard- deviation contours about the final solution. It is shown that by applying travel-time anomalies, the standard confidence ellipses can be reduced in area by factors of 15 to 1152 and still contain the true epicenter. A discussion is given of the stability of travel-time anomalies across the Nevada Test Site area, and of some problems involved in determining usable anomalies from earthquakes.