Improved Surface Wave Detection and Measurement Using Phase-Matched Filtering with a Global One-Degree Dispersion Model
Abstract
The primary goal of this project is to improve the capability to identify and detect surface waves for the purpose of earthquake/explosion discrimination. We are developing improved, higher resolution earth and dispersion models. The models consist of approximately 600 distinct crust and upper mantle structures, with surface layering and/or ocean depths that vary on a one-degree grid. There are a total of 64,800 earth models and dispersion curves, but the tomographic inversion is performed only for the 600 distinct crust and upper mantle models, with the shallow structure constrained by other information. The data set used in the inversion now consists of approximately 548,000 phase and group velocity dispersion measurements obtained from a variety of sources. Surface sediments are defined using the global sediment maps of Laske and Masters (1997), and ocean bathymetry is defined using the Etopo5 topographic data set. Automatic identification of surface waves at the International Data Centre is currently performed by narrow-band filtering the data at several frequencies, and then comparing the arrival times with a regionalized dispersion model. We have implemented and tested a new procedure in which we first phasematch filter the data and then apply narrow-band filters to the compressed waveform and use a detection test similar to the current test. This allows us to take advantage of the improved signal-to-noise ratio of the phase-match filtered waveforms, while retaining the robustness of narrow-band filtering for frequency dependent signal identification. After phase-matched filtering, the predicted arrival time is zero at all frequencies, so we test to see if the arrivals are within a time window similar to that used in the existing test. To test the procedure, we processed the same data set using five-degree and one-degree models with and without phase-matched filtering.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2001
- Accession Number
- ADA529612
Entities
People
- David A. Adams
- G. E. Baker
- Jeffry L. Stevens
Organizations
- Leidos