Copyright © 2003 Elsevier B.V. All rights reserved.
D″ shear velocity heterogeneity, anisotropy and discontinuity structure beneath the Caribbean and Central America
Accepted 11 July 2003. ;
Abstract
The D″ region in the lowermost mantle beneath the Caribbean and Central America is investigated using shear waves from South American earthquakes recorded by seismic stations in North America. We present a large-scale, composite study of volumetric shear velocity heterogeneity, anisotropy, and the possible presence of a D″ discontinuity in the region. Our data set includes: 328 S(Sdiff)-SKS differential travel times, 300 ScS-S differential travel times, 125 S(Sdiff) and 120 ScS shear wave splitting measurements, and 297 seismograms inspected for Scd, the seismic phase refracted from a high-velocity D″ layer. Broadband digital data are augmented by high-quality digitized analog WWSSN data, providing extensive path coverage in our study area. In all, data from 61 events are utilized. In some cases, a given seismogram can be used for velocity heterogeneity, anisotropy, and discontinuity analyses. Significant mid-mantle structure, possibly associated with the ancient subducted Farallon slab, affects shear wave travel times and must be corrected for to prevent erroneous mapping of D″ shear velocity. All differential times are corrected for contributions from aspherical mantle structure above D″ using a high-resolution tomography model. Travel time analyses demonstrate the presence of pervasive high velocities in D″, with the highest velocities localized to a region beneath Central America, approximately 500–700 km in lateral dimension. Short wavelength variability overprints this general high-velocity background. Corrections are also made for lithospheric anisotropy beneath the receivers. Shear wave splitting analyses of the corrected waveforms reveal D″ anisotropy throughout the study area, with a general correlation with heterogeneity strength. Evidence for Scd arrivals is pervasive across the study area, consistent with earlier work, but there are a few localized regions (100–200 km) lacking clear Scd arrivals, which indicates heterogeneity in the thickness or velocity gradients of the high-velocity layer. While small-scale geographic patterns of heterogeneity, anisotropy, and discontinuity are present, the details appear complex, and require higher resolution array analyses to fully characterize the structure. Explanations for the high-shear wave speeds, anisotropy, and reflector associated with D″ beneath the Caribbean and Central America must be applicable over a lateral scale of roughly 1500 km2, the dimension over which we observe coherent wavefield behavior in the region. A slab graveyard appears viable in this regard.
Author Keywords: Shear velocity; Anisotropy; Core–mantle boundary; D″; Heterogeneity
Article Outline
- 1. Introduction
- 2. South American earthquake data set
- 3. D″ heterogeneity inferred from differential travel times
- 3.1. Differential travel time residuals
- 3.2. Inferred D″ heterogeneity
- 3.3. Mid-mantle contamination of ScS-S times
- 4. D″ anisotropy inferred from shear wave splitting
- 4.1. Shear wave splitting measurements
- 4.2. Inferring D″ anisotropy
- 4.3. Correlation between heterogeneity and anisotropy
- 5. Shear wave reflections off of high velocities in D″
- 6. Discussion
- 7. Conclusions
- Acknowledgements
- References
Corresponding author. Tel.: +1-480-965-7653; fax: +1-480-965-8102.
1 Tel.: +1-831-459-3164.






E-mail Article
Add to my Quick Links

Cited By in Scopus (29)

280 km above the core-mantle boundary. The new data set consists of long-period tangential component recordings at WWSSN stations in Africa, the Middle East, and Europe for 11 intermediate and deep focus Indonesian earthquakes. In the distance range 70–82° the waveforms show an arrival between SH and ScSH with systematic moveout. From 89 to 94° there is a strong distortion of the SH waveforms, indicating the arrival of several phases closely spaced in time. The relative time shifts of similar complexity in the corresponding sSH phases requires a deep mantle origin. The depth dependence and moveout of the interference effects are well-predicted for both SH and sSH phases by a model with a lower mantle discontinuity. Alternative explanations of the interference as resulting from receiver reverberations, SKS contamination, multiple source complexity, or near source multipathing are ruled out by systematic tests. While it is apparent that lateral variations in the lower mantle velocity structure prevent any single model from fitting all of the data, synthetic waveform modeling (using generalized ray theory and reflectivity) shows that the data can be well-fit by a model with a discontinuity similar in size and depth to that proposed for the previously investigated regions (Lay and Helmberger), but with a negative velocity gradient within the D″ layer.




