Crustal Structure and Surface-wa Ve Dispersion

INTRODUCTION IN 1953, through cooperation of the Bernard Price Geophysical Institute of the University of the Witwatersrand and the University of Natal, a Columbia University-type long-period seismometer (To = 15 sec., Tg = 75 sec.) was installed in the seismological observatory of the University of Natal at Pietermaritzburg, Union of South Africa. This instrument was well situated for receiving surface waves from the shock in northern Algeria of September 9, 1954, and the aftershock next day, which was of such intensity that its seismogram supplemented that of the original shock for the larger phases. The dispersion of the Rayleigh waves from these seismograms can be measured with greater precision than has been practicable heretofore for continents, because the path is longer (7,890 km.) than any which has been available for a long-period vertical instrument, and is remarkably free from obvious anomalies such as major mountain ranges. The scarcity of suitable seismograms for the study of Raleigh-wave dispersion along continental paths is due to the fact that suitably placed long-period vertical seismographs have not been available until recently. DATA According to B.C.I.S. the epicenter of the main shock of September 9, 1954, H = 01:04:37, was 36° 17' N, 1° 28' E, and the Pasadena magnitude was 6%. The aftershock of September 10, 1954, H = 05:44:05, was placed at 36~6 N, 1~3 E, Praha magnitude 6.2. The Rayleigh-wave portion of both seismograms is reproduced in figure 1. Time is marked in figure 1 as minutes after the origin time, and the beginnings of the Rayleigh-wave train, the Lg train, and the Rg train are indicated. Also marked is a group of waves which is interpreted as an Rg phase reflected from the continental margin. The Rayleigh-wave train for the main shock shows clearly the typical Rayleigh-wave dispersion for continental paths which has been commonly studied in the past. With the arrival of the Lg phase, the apparent intensity of which is greatly reduced by the decrease in sensitivity of the seismograph for short periods, it becomes more difficult to read the ordinary Rayleigh waves, but they can bEHead by shifting from the main shock to the aftershock at a point about halfway between the arrival of Lg and that of Rg, entirely out to the beginning of Rg. The waves of intermediate frequency arriving between Lg and Rg are probably a part of the Lg train. Rg begins abruptly and with great amplitude, as may be seen very clearly on the seismogram of the aftershock. The inverse dispersion in this phase is clearly demonstrated for the first time on this seismogram. The point at 44 min. 45 sec. after the origin time just prior to the arrival of a new shorter-period reflection train is taken to represent an Airy phase corresponding to the minimum value of group velocity for Rayleigh-wave propagation across continents. The short-period waves (which likewise show a lengthening of period with time) just following this Airy phase are interpreted as a reflection from the continental bound~ry at some point such that the path of the reflected wave exceeds the great circle * Manuscript received for publication March 25, 1955.