Analysis of Upper and Lower Mantle Structure Using Shear Waves

Abstract

This thesis addresses the fundamental problem of determining the radial and lateral structure of the earth's interior using body wave observations. My approach is a cautious one, involving detailed analysis of a substantial data set in which I attempt to isolate the contributions from radial and lateral structure in both the upper and lower mantles. This is an elusive task, as I am concerned with the 2 or 3% fluctuations about standard earth models, which produce rather subtle effects on teleseismic signals. However, this is the level of precision to which the three dimensional structure of the earth must be determined if we are to map the dynamical and compositional configuration of the earth. Studies similar to those described here cannot be conducted on a detailed global basis, given the intrinsic limitations due to station and earthquake distribution, so I have repeatedly emphasized the qualitative implications of my results, as they probably provide a representative sampling of the subtle, but significant heterogeneity of the earth. The topics addressed in the following work appear unrelated at first glance, ranging from determination of the detailed shear velocity structure at the base the mantle to variations in attenuation and velocity structure of the upper mantle. However, the data set used throughout is largely the same, and this in itself indicates the need to consider all of the complexity discussed herein for future progress to be made in mapping the three dimensional structure of the earth with a high level of confidence. The first chapter of the thesis presents results of a waveform analysis of transversely polarized SH signals that propagate through the lowermost mantle. The waveforms of these phases show clear interference patterns due to interaction with a discontinuous shear velocity increase about 280 km above the core-mantle boundary. This discontinuity, which has not been detected previously, is manifested in signals sampling three widely separated portions of the lower mantle, and hence is a good candidate for a global radial earth structure. Very detailed inspection of the signals reveals evidence for lateral variations in the depth of the discontinuity, which provides a procedure by which to map the structure of the D region in detail. Analysis of the relative amplitudes of SH and ScSH signals reveals that the velocity gradient above the core boundary is consistent with the smoothly varying gradients in most gross earth models, but evidence for local high velocity gradients at the base of the mantle is detected in ScSV signals. Chapter II presents a travel time analysis of the same data set used in Chapter I, which was motivated by the observation of large amplitude and travel time anomalies in the S and ScS data. An emphasis is placed on isolating the portions of the S and ScS paths which are anomalous. A strong empirical case is made for the existance of localized regions with scale lengths of 1000 to 2000 km and 2% velocity anomalies within the lower mantle at depths from 1000 to 2700 km. The long period signals traversing these regions show as much as a factor of 2 amplitude enhancement or diminution. This result demonstrates that both amplitude and travel time anomalies are induced by lateral structure in the portion of the mantle assumed to be homogeneous in most seismological studies. The third chapter is an analysis of the influence of upper mantle variations in attenuation, velocity structure and receiver structure on the S and ScS signals analyzed in the lower mantle studies. These variations contaminate and complicate the interpretation of any data set used to study deeper earth structure. Along with evidence for very strong and abrupt variations in upper mantle properties, results are presented which indicate the inadequacy of assumptions that are frequently made about the nature of long period body-wave receiver functions.