Surface-wave Attenuation and Crustal Anelasticity in Central North America By

The southeastern Missouri earthquake of October 21, 1965 generated fundamentaland higher-mode Love and Rayleigh waves which were recorded at numerous North American stations. Love-wave amplitude radiation patterns were determined and found to be consistent with theoretical patterns predicted by a fault-plane solution previously inferred from Rayleigh-wave data. The radiation patterns were used to estimate the source spectrum and values for Love-wave attenuation coefficients for the mid-continent of North America by a least-squares iterative process. The source spectrum derived from Love-wave amplitudes exhibits a peak at periods between 5 and 9 sec and decreases to a lower DC level at longer periods, in agreement with the source spectrum determined previously for Rayleigh waves. The Love-wave attenuation coefficients decrease rapidly from about 0.0018 km -1 at a period of 4 sec to about 0.0001 km -1 at a period of 20 sec. At periods between 20 and 40 see the values seem to remain nearly constant. The crust in the mid-continent of North America is characterized by relatively low Q~ values, 75 to 300, in its upper portion. At depths between 15 and 20 km, Qp increases sharply and decreases again at greater depths. The decrease can be explained as being due to increasing temperature in a homogeneous material, but the sharp increase requires a change in the chemical constitution of the material at mid-crustal depths. I N T R O D U C T I O N Seismic surface-wave amplitude radiation patterns have been successfully .used to determine earthquake fault-plane solutions. Several authors (e.g. Wu and Ben-Menahem, 1965; Wu, 1968; Canitez and Toks6z, 1971; Mitchell, 1973) have compared observed radiation patterns with those computed according to the theory of Ben-Menahem and Harkrider (1964) to obtain a solution. That theory predicts that surface-wave amplitudes and initial phases generated by'earthquakes will vary azimuthally in a manner determined by the depth, strike, dip, and slip angle of the earthquake. The patterns are also model dependent and may vary with frequency and mode number. In a recent study, Mitchell (1973) used Rayleigh-wave amplitude radiation patterns to determine a fault-plane solution for the southeastern Missouri earthquake of October 21, 1965. Having made that determination, it was then possible to estimate values for the Rayleigh-wave attenuation coefficients and to estimate the shape of the source spectrum. Those estimates were made at periods between 4 and 50 sec. An interesting result of that study was that the shape and orientation of the short-period Rayleigh-wave radiation patterns were roughly the same as the area of perceptibility of the earthquake which was determined by Kisslinger and Nuttli (1965). This similarity, combined with lower values of Rayleigh-wave attenuation coefficients in eastern North America as compared with western North America supported the previous conclusions of Necioglu and Nuttli (1971) and Nuttli (1973) that the larger areas of perceptibility of eastern North American earthquakes result from the smaller values of Rayleigh-wave attenuation coefficients in eastern North America. 1057 1 0 5 8 B R I A N J. M I T C H E L L This paper presents the results of similar computations for Love waves generated by the October 21, 1965 southeastern Missouri earthquake. Love-wave amplitude radiation patterns are first computed and found to be consistent with those predicted by the favored fault-plane solution of a previous paper by Mitchell (1973). Using that earthquake as a known source, a least-squares fitting of the observed and theoretical patterns yields Love-wave attenuation coefficient values as well as an estimate of the source spectrum. The region over which the observations of the present study and those of the author's previous paper were made lies in the mid-continental region of North America, between the Rocky Mountains and the Atlantic coast. The relative uniformity of this region allows one to use the surface-wave attenuation coefficients to make a determination of the depth distribution of the quality factor of shear waves in the crust.