Department Logo for Earth, Planetary, and Space Sciences

Automatic detection and picking of direct P, S and fault zone head & trapped waves


Jan. 21, 2015, noon - 1 p.m.
Geology 1707

Presented By:
Zachary Ross
USC

See Event on Google. Subscribe to Calendar

We develop a set of algorithms for automatic identification and picking of seismic P and S body waves and fault zone head and trapped phases. The P body wave arrivals are picked using a kurtosis-based algorithm, while S body wave picking is done using polarization analysis and related filters to remove P-wave energy. Then, STA/LTA and kurtosis detectors are used in tandem to lock on a well-defined phase arrival. We search automatically in P waveforms for fault zone head waves generated by earthquakes on faults that separate different lithologies, and in S waveforms for trapped waves resulting from constructive interference in low velocity fault zone layers. The head wave detection and picking is performed by a multi-stage algorithm that accounts for basic characteristics (motion polarities, time difference, sharpness and amplitudes) of head and direct P phases. The algorithm for automatic identification of trapped waves calculates three features representative of trapped waves in time windows following the direct S wave in seismograms of near-fault stations. Statistical analysis of these features is used to identify candidate trapped waves in given stations. The methods are tested on synthetic seismograms and are shown to work well in data recorded along the San Andreas fault near Parkfield, the Hayward fault, the Landers rupture zone, the San Jacinto fault and data from Taiwan.