In order to investigate the large errors which sometimes characterize fundamental-mode attenuation coefficient determinations, we have made many such determinations in the Basin and Range province and have plotted particle motion for the observed three-component seismograms. Rayleigh wave attenuation coefficient values determined for four two-station paths across the Basin and Range province fluctuate between about −2.0 × 10−3km−1 and +3.0 × 10−3km−1 at periods between 6 and 33 s. Partide motion plots indicate that many of those determinations are contaminated by arrivals from non-great circle paths and from waves generated at heterogeneities along the path, factors which lead to systematic errors in the measured attenuation coefficient values. Attenuation coefficient determinations for the path MNV to ELK, which is within 20° of being normal to the structural trends of the Sierra Nevada and Great Valley, are, however, relatively free from such contamination. Mean values for that path decrease rapidly with period from about 3.0 × 10−3 to about 0.7 × 10−3km−1 between 6 and 10 s and then decrease slowly to about 0.3 × 10−3km−1 at 33 s. Standard deviations range between 0.2 × 10−3 and 0.3 × 10−3km−1 for most of the period range, but increase to about 0.4 × 10−3 at periods between 6 and 8 s. These results indicate that careful screening, based upon criteria such as three-dimensional particle motion, are necessary to obtain surface wave attenuation data of sufficient quality to use for inverting for crustal anelasticity in complex regions.

First Page Preview

First page PDF preview
You do not currently have access to this article.