Path Corrections for Regional Phase Discriminants

In order to utilize regional seismic phases for reliable source identification, it is essential to account for wave propagation effects that may obscure the subtle differences in source radiation for distinct source types. Recent work has established three main approaches to the problem of characterizing and correcting regional phases for propagation effects: empirical methods based on parametric dependence on path properties such as pathlength or crustal waveguide parameters, empirical interpolation methods such as cap-averaging and Kriging for stationdependent source region correction surfaces, and waveform modeling methods based on complete waveform synthesis for one-, two-, or three-dimensional crustal models. Empirically based methods remain the most viable general approach, as model based methods are still greatly limited by our wave propagation techniques and by lack of independent constraints on three-dimensional waveguide structure. Kriging and waveguide parameter regressions for frequency dependent regional phase amplitude ratio measurements for International Seismic Monitoring Stations, NIL and ZAL were conducted in the first year of this study. The data have a large amount of scatter, much of which is attributed to wave propagation effects; however, propagation corrections are limited by our understanding of source induced scatter and by effects of partial or total blockage of regional phases. We have addressed the issue of source induced scatter by analysis of a well-controlled data set of regional recordings of shallow earthquakes with reasonably well-determined focal mechanisms. By using events in very close proximity with varying depth and/or focal mechanisms recorded on common paths, we have isolated sourcerelated contributions to variance in regional seismic discriminant measurements involving P/S-type ratios (e.g., Pn/Lg, Pg/Lg, Pn/Sg, and Pg/Sg) for earthquakes in southern California and southern Nevada. Broadband waveforms for earthquakes near the Nevada Test Site (NTS) and in southern California recorded at stations of the Livermore NTS Network (LNN), the Caltech TERRAscope Network, and the Berkeley Digital Seismic Network (BDSN) are used to obtain the seismic discriminant measurements. Pn/Lg and Pg/Lg spectral amplitude ratios for various frequency bands between 0.5 and 10 Hz from vertical velocity waveforms are computed for 17 tightly located clusters of earthquakes. For each cluster, we examine the correlations between observed Pn/Lg and Pg/Lg ratios and estimated source depths or previously determined source mechanisms. Focal mechanism predictions do not match the observations well, likely due to refraction and directivity effects that overwhelm the pointsource radiation pattern. However, the data do show significant scatter that is clearly not attributable to varying path structure. Source depth effects show weak trends, with events shallower than 3 kilometers having enhanced Pg/Lg and Pn/Lg ratios that contribute to the overall scatter. Depth dependence is of great concern, as it undermines the empirical approaches to path correction. Variance estimates for logarithmic amplitude ratios normalized by mean amplitude ratios vary from 0.01 to 0.05 for different phase pairs, and there is strong dependence on propagation distance. These variance parameters appear to be suitable for use in spatial interpolations of observed P/S type ratios for a specific unidentified event path. With respect to blockage effects, we have undertaken a specific study of blockage associated with Lg phases traversing the northern margin of Tibet. Using observations at WMQ, we explore the spatial variation of Lg transmission across the Plateau, manifested in large corner frequency shifts and disappearance of the Lg signal. The corner frequency shift shows systematic evolution with transit length in the Plateau, indicating that there is progressive, not discrete, blockage of the wavefield, and this is likely due to the anomalous crustal properties of Northern Tibet. Correction using Report Documentation Page Form Approved