Physical mechanisms of seismic-wave attenuation

Abstract

The theoretical and experimental evidence concerning mechanisms likely to be responsible for the attenuation of seismic waves are reviewed. Intergranular thermoelastic relaxation, atomic diffusion, and dislocation mechanisms cannot be ruled out as significant causes of seismic attenuation, but the most effective mechanisms seem to be associated with partial melting, grain‐boundary relaxation, and a poorly understood mechanism called ‘high‐temperature, internal‐friction background’ which obeys an equation of the form Q^(‐1) = (A/ƒ) exp (—H*/RT). Here Q^(−1) is a dimensionless measure of anelasticity, f is the frequency, T is the absolute temperature, and A, H*, and A, H*,and RR are constants for a material of uniform composition and grain size. Many of the mechanisms considered here have a strongly frequency‐dependent Q^(−1). However, in a material as complex as a rock, there is unlikely to be a single discrete relaxation time or activation energy, or a single ‘typical’ grain size. The observed Q^(−1) is likely to result from a superposition of several mechanisms and involve a spectrum of parameters leading to weaker frequency dependence than is predicted from a single simple mechanism. The seismic data cannot at present resolve the question of whether or not Q^(−1) has an intrinsic frequency dependence. The extent and the limitations of the available seismic data are discussed. Experimental data on the attenuation of oxides are summarized, the measurements at high temperature and low frequency being emphasized.