L.M. Warren and P.M. Shearer
Institute of Geophysics and Planetary Physics, Scripps Institution of Oceanography, University of California, San Diego
\def\tstar{t^*} \def\deltstar{\delta{\tstar}} We study variations in the attenuation structure of Earth. Attenuation anomalies reflect thermal and compositional variations, so mapping these anomalies may help reveal features such as rising plumes or subducting slabs. The strongest attenuation and largest anomalies occur in the top few hundred kilometers of the mantle. However, lateral variations may persist to deeper depths and we investigate how large the variations are and if they correspond with dynamical features suggested by velocity models. Attenuation along a given path can be measured from its spectral decay with frequency, so we compute the spectra of direct $P$ and $S$ wave arrivals recorded by the global seismic networks. Each spectrum is the product of source, receiver, and propagation response functions as well as local source- and receiver-side effects. We correct each spectrum for average source and attenuation models and the known instrument response. Since there are multiple receivers for each source and multiple sources for each receiver, we can approximate the source- and receiver-side terms by stacking the appropriate log spectra. The resulting source-specific response functions include any remaining source spectrum and near-source $Q$ structure; the receiver stacks include the site response and near- receiver $Q$ structure. Assuming the stacks represent attenuation and other effects in the upper mantle around the earthquakes and stations, the residual spectrum (after correcting for these effects) represents variations in attenuation along the rest of the ray path. After measuring $\deltstar$ from the slope of the residual log spectrum, we bin the $\deltstar$ values by turning depth and plot them at their turning points. The resulting maps show laterally coherent patterns over small length scales, but the patterns do not correlate for the different depth ranges. We also see no correlation between the $\deltstar$ values and similarly-binned travel time residuals. The average $\deltstar$ values increase sharply at the base of the mantle. While some of this increase may be caused by increased attenuation in D'', reflectivity synthetics suggest that interference from the $ScS$ arrival will also raise t* values.