Finite Frequency Seismic Tomography for Eastern Eurasia

Ting Yang,Yang Shen,Xiaoping Yang
2006-01-01
Abstract:: Seismic calibration for nuclear explosion monitoring requires accurate, high-resolution 3-D velocity models. Although seismic event locations have been improved recently using more accurate travel times measured by waveform cross-correlation, ray-based models and predictions are not strictly self-consistent with travel times measured by waveform cross-correlation, which are sensitive to 3-D velocity perturbations around the geometric ray paths. Finite-frequency seismic tomography (FFST), which utilizes the 3-D sensitivity kernels to determine the structure that best fits the observations from waveform cross- correlation, provides a self-consistent approach in model development and event location We apply the FFST method to determine the crustal and mantle velocity structure beneath eastern Eurasia. Taking advantage of the broadband feature of seismic records, we measure P-wave (primary wave) relative delay times by waveform cross-correlation in three frequency hands (0.03-0.1 Hz, 0.1-0.5 Hz, and 0.5-2.0 Hz). The measurements are inverted jointly to constrain velocity heterogeneities with different distances from the central geometric rays. The effect of strong variations in crustal structure beneath this region on travel time data is removed by conducting a frequency-dependent crustal correction. A comprehensive data set, including waveforms from the publicly accessible sources and other seismic networks in the region, has been collected for this study. Our preliminary model has similar patterns of high- and low- velocity but higher magnitude anomalies than ray-based tomographic models. The resolution checkerboard tests suggest that the resolution is better beneath eastern China, reflecting the better sampling in that part of the model in the current data set. The procedures for computing frequency-dependent travel time corrections from the finite-frequency model have been developed and tested in relocating ground truth (GT) events.
What problem does this paper attempt to address?