Accession Number:

ADA525396

Title:

Cepstral F-Statistic Performance at Regional Distances

Descriptive Note:

Conference paper

Corporate Author:

Report Date:

2001-10-01

Pagination or Media Count:

10.0

Abstract:

We have developed a cepstral F-statistic method that attaches statistical significance to peaks in the cepstra of seismic data. These peaks often result from echoes such as depth phases and thus provide a means of identifying possible depth phase candidates. Detections from this method are stacked as a function of their pP-P and sP-P delay times predicted by IASPEI travel-time tables using a modified version of the network stacking method of Murphy et al. 1999. The method detects depth phases with signal-to-noise ratio SNR greater than 2, as long as the P wave SNR is greater than 5 to 8, providing a wide range of applicability. We have tested the method on limited datasets from the United States Geological Survey, the Prototype International Data Center, and the International Data Center, and have shown the method to be more reliable at automatically picking possible depth phases than current algorithms. We are now in the process of further testing the method using the extensive datasets at the Research and Development test bed at the Center for Monitoring Research. We have successfully applied the method to events with epicentral distances greater than 12 degrees and focal depths greater than 15 km. Our focus during the past year has been to examine the technique at near-regional distances for small-to-moderate sized events of varying depths. To accomplish this task, we have acquired a high-quality ground-truth dataset compiled by Ratchkovski and Hansen 2001 using the Alaska Earthquakes Information Center AEIC network. We have chosen a subset of the 14,000 events they relocated with magnitudes ranging from 3.5 to 5.1 Md, and we are in the process of applying the method to the seismic data recorded for these events at regional distances using arraysstations ATTU, BCAR, BMAR, KDAK, ILAR, and IMAR.

Subject Categories:

  • Statistics and Probability
  • Seismic Detection and Detectors

Distribution Statement:

APPROVED FOR PUBLIC RELEASE