S21B-2681
Investigating Sources of Uncertainty in Surface Wave Ellipticity Measurements across the USArray

Tuesday, 15 December 2015
Poster Hall (Moscone South)
Julie Marie Schnurr, Tolulope M Olugboji and Vedran Lekic, University of Maryland College Park, College Park, MD, United States
Abstract:
Raleigh wave ellipticity, usually reported as vertical-to-horizontal amplitude ratio ZH, can be used to constrain earth structure. In the 0.02-0.2 Hz range, ZH ratios are typically measured from ambient noise and are used to map variations in crustal structure, especially shallow-layer density (e.g. Lin et al. 2012). At lower frequency (<0.02 Hz), ZH ratios offer unique constraints on lithospheric density, but measurements at a single station show large variability, which has yet to be fully understood. Past investigators have attributed this variability to primarily result from small-scale variations in wavespeed (Ferreira & Woodhouse, 2007), raising the prospect of using these observations as a constraint on small-lengthscale structure.

We measure ZH ratio using large magnitude earthquakes recorded at US Array stations. We make measurements on 1-D and 3-D synthetic waveforms, as well as earthquake data, comparing our results to model-based predictions. To analyze the large data set from the USArray, we develop an algorithm based on cluster analysis, in which similar waveforms are grouped together, based on similarity, to enable more efficient user control. ZH ratio is then calculated as the ratio of the envelope of the vertical and horizontal components of narrow band-pass filtered waveforms. Cross correlation is used to align the waveforms and ZH ratios are selected in time windows in which correlation is high and ZH is stable. The mean and observational standard deviation of the ZH ratio is measured within the selected time windows.

We explore the frequency and spatial distribution of ZH measurement uncertainty as well as difference between the measurements and model-based predictions. In order to identify which parameters control the stability of measurements, we quantify the effects of frequency, earthquake depth, epicentral distance, and variations with back azimuth on the stability and accuracy of ZH measurements. Finally, we identify regions where ZH ratios bear evidence of sharp lateral variations in structure.