Seismic surface waves and the crustal structure of the Pacific region*

summary The favorable location of Berkeley on the eastern margin of the Pacific makes possible a comparative study of surface waves coming directly from the epicenter to that station over paths that are purely Pacific or purely continental. Records of 378 earthquakes dating from November, 1910, to May, 1934, have been used in this study. Speeds and period of the initial impulses of Love waves have been measured and associated with wave-velocity. These waves show normal dispersion, the long waves having the greater speeds. Speeds over oceanic paths are higher than over continental paths, the difference diminishing with an increase in the wave-length. For short or long waves, they are about the same over all Pacific paths, but waves having intermediate periods (30 or 40 seconds) cross under the Aleutian deep faster than under the Polynesian Pacific. The data indicate for the crustal thicknesses under western North America and the Pacific the following approximate values: granite 20 kilometers, gabbro 40 kilometers, under western North America; basalt 25 kilometers, dunite 20 kilometers, under the Aleutian deep; basalt 30 kilometers, dunite 25 kilometers under Polynesia; and intermediate values under the remainder of the Pacific. A third discontinuity under the continent at depths somewhat greater than 60 kilometers is indicated. If the sub-Pacific is assumed to be single-layered, thicknesses of 35 to 45 kilometers form the best fit to the data. The thickness of the crust underlying the Pacific Islands is probably about 10 kilometers greater than that underlying the deeps of the North Pacific. Movements associated with Rayleigh waves apparently have their closest approach to theoretical conditions over central Pacific paths. The dominant periods in the coda are 8 to 9, 10, 13, and 16 seconds. The longer periods are dominant at the greater distances. The increase in period with distance is a discontinuous, step-like function. The structure underlying the Aleutian deep apparently is opaque to 13-second periods in the Rayleigh wave, and the central Pacific and possibly the Atlantic structure seems to be unfavorable in the transmission of waves having this period. Vibrations in the natural period or overtones thereof set up in the focal region being eventually transmitted as sympathetic vibrations to the region of the station seems to be a logical explanation for the dominant groupings in the coda. Should any section of the path be out of sympathy with a given period, this period would probably become subordinate if not lost. The natural period of a portion of the Arctic region, including Alaska, seems to be different from other parts of the world. The natural period of the San Francisco Bay region, or its dominant overtone, is observed to be about eight seconds.