The apparent velocity and azimuth of P from the Novaya Zemlya nuclear explosion of October 27, 1966, have been measured for the Warramunga array and have been used to work out possible crustal structures for correcting measured apparent velocities and azimuths in the azimuth range 335 ø to 355 ø . Comparison of the measured apparent velocities and azimuths for seven earthquakes at different distances, but at similar azimuths to Novaya Zemlya, has shown that the computed structures give satisfactory corrections. Measurements of apparent velocity and azimuth have been used as an aid in identifying the phases PKiKP, PP, PcPPcP, PKKP, PcPPKP, and PKPPKP in the explosion record; similar measurements have been made for coherent signals not corresponding to any of the conventional phases. from teleseisms recorded at WRA show systematic differences from the true azimuth and the Jeffreys-Bullen (J-B) apparent velocities; these discrepancies are believed to be due to a dipping Mohorovicic discontinuity and other dipping interfaces within the crust.
An important application of an array of short-period vertical component seismometers is the determination of a velocity-depth distribution for P waves throughout the whole ofthe earth's mantle. The measurements of dA/dT need, however, to be corrected for the effects of structure beneath the array. The apparent velocities and azimuths have been calculated by means of a least-squares program designed by E. W. Carpenter and written by B. S. Gopala-Copyright ¸ 1969 by the American Geophysical Union. krishnan. This least-squares method has been summarized by Otsuka [1966a, b]. A subroutine has been added to the least-squares program that calculates the standard errors in velocity and azimuth by using the formulas given by Kelly [1964]. An alternative, but less precise, means of velocity and azimuth determination is a correlation method, which has been described by Bittill and Whiteway [1965]. Niazi [1966] considered the effect of a dipping M discontinuity on measurements of the parameter dT/dA and azimuth. By using the theory outlined in Niazi's paper, a computer program has been written that will work out the effect on apparent velocity dA/dT and azimuth of any combination of dipping interfaces for a seismic event whose azimuth and distance are accurately known. The time residuals for each seismometer of the array are also calculated. The Novaya Zemlya nuclear explosion of October 27, 1966, at a distance of 106.0 ø, was well recorded by the array, and a number of later phases as well as P were visible. The high quality of the record of P, together with accurate knowledge of the expected apparent velocity of diffracted P, makes this explosion ideal for determining a crustal structure beneath the array. Discussion of possible crustal structures 2034 LONGITUDINAL WAVES FROM NOVAYA ZEMLYA EXPLOSION 2035 Am., 43, 4'9-61, 1953. Jeffreys, H., The Earth, pp. 103-105, Cambridge University Press, New York, 1962a. Jeffreys, t/., Travel times for Pacific explosions, Geophys. J., 7, 212-219, 196...