PPHASEPICKER is a powerful tool for automatically picking P-phase onsets with high precision without requiring detection interval or threshold settings. The algorithm detects P-phase onset in single-component acceleration or broadband velocity records using the histogram method. It also computes signal-to-noise ratio (SNR).
PPHASEPICKER has been integrated into the “Automated Processing and Review Interface for Strong Motion Data (PRISM)” software (https://earthquake.usgs.gov/research/software/prism/) of the U.S. Geological Survey in order to identify pre-event time-window for systematic and automated processing of large numbers of strong-motion data.
An example MatLAB code is provided in zip file to show how to run PPHASEPICKER using three sample waveforms, one from strong earthquake and others from micro seismic events.
For your intended application, please note that filter corner frequencies may need to be updated manually within the code based on the frequency content of the input record.
Kalkan, E. (2016). “An Automatic P-phase Arrival Time Picker“, Bull. of Seis. Soc. of Am., 106(3): 971-986, doi: 10.1785/0120150111.
Dr. Erol Kalkan, P.E. (2019). An automated P-phase Arrival Time Picker with SNR output (https://www.mathworks.com/matlabcentral/fileexchange/57729-an-automated-p-phase-arrival-time-picker-with-snr-output), MATLAB Central File Exchange. Retrieved .
That's a good comment. Note that background noise always exists in the signal itself too so it is not possible to say P-phase onset is the cutoff between noise and seismic signal. It is a cut-off between noise and noise+seismic signal. Even if you filter out the signal based on FAS of the noise, you still remove part of the seismic signal within the same passband. Nonetheless, I will include your suggestion into the code as short window SNR output. Thanks...
Erol in part where you Compute SNR, you take:
but x is all signal (noise first and signal then). I think that is necessary select only seismic wave signal without noise for example using a window centered in ppicking time (that suppose is named Tm) and taking 2 second after this value and 1 second first this value (using a window of width 3s). So the signal window will be [Tm-1s;Tm+2s].
I just completed a new algorithm to pick both P- and S-phases. It is now in MatLAB FEX.
This algorithm no find the time value when p-phase end, and the s-phase pick. How find them? Thanks.
Thank you! I have included a link above for download of the paper.
Hello, would it be possible for you to email me your paper - “An Automatic P-phase Arrival Time Picker“, mentioned in the reference section to - "email@example.com". I would really like to read it.
Updated the reference
Added signal-to-noise ratio computation
Code is modified to prevent numerical errors due to very low amplitude input signals
Added BSSA paper describing the picker
Fixed "nbins" for arbitrary sampling rates; tested for inhomogeneous (arbitrarily sampled) weak acceleration records from China.
Modified for low sampled input waveforms (lower than 100 samples-per-second, sps) and third example (waveform with 80 sps) is included in demo file.
Tested for MatLAB 2013a, 2014a and 2016a.
A new micro seismic earthquake example is included.
Typo has been fixed.
Text is revised
Create scripts with code, output, and formatted text in a single executable document.