Quantitative characterization of seismic tremors in the northern Cascadia margin
The episodic tremor-and-slip (ETS) events refer to the concurrence of westward crustal movements, as evident by GPS observations, and tremor-like seismic activity in the northern Cascadia margin. The regular occurrence of ETS events in the region has been remarkable. In addition to the 14-month peri...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | English en |
| Published: |
2010
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/1828/2178 |
| Summary: | The episodic tremor-and-slip (ETS) events refer to the concurrence of westward crustal movements, as evident by GPS observations, and tremor-like seismic activity in the northern Cascadia margin. The regular occurrence of ETS events in the region has been remarkable. In addition to the 14-month period, secondary tremor activities, most of them lasting less than one or two days, are also found with no corresponding GPS signatures. However, the identification of tremor activity is mainly based on visual examination of regional/Iocal seismic records. In this study, we attempt to develop an algorithm that can quantitatively characterize the level of tremors from a collection of seismic waveform data. For each hour of waveform at a given station, the process begins with the calculation of moving average and scintillation index with various time lengths. The scintillation index, essentially the "normalized variance of intensity of the signal", is adapted from the studies of pulses in radio waves and is an efficient tool to identify the pulse-like characteristics of tremor signals. Values of the indices are fed into a series of logic gates that use a combination of both parameters to determine if sufficient tremor activity exists. To demonstrate the effectiveness of our algorithm, seismic waveform data are collected for the known February/March 2003 ETS event. Our analysis gives results consistent with the work done manually. Implementation of our algorithm is straightforward and free from human intervention. Thus, it is potentially possible to automate the tremor monitoring process that may give early warning of the exact arrival time of ETS events. |
|---|