While there exist many procedures for automated arrival time picking of seismic phases, assessment of timing uncertainty and identification of such phases remains a tricky business. We routinely apply a semi-automated procedure that is tuned to a hand-picked and carefully evaluated data set. If you are interested in such procedure please check the following publications (they are available through my page on researchgate):

Diehl et al. 2009, GJI; Diehl et al. 2011, NMSOP; Kianimehr et al. 2018, JSeismol; Rezaeifar&Kissling 2018, GJI.

regards,

Edi Kissling

what do you mean by automatic? obtained by inspection of a seismograph?

One way is doing subspace detection, it works doing singular value decomposition of parent events and you can use the result for picking p and s waves.

Let me know if you wnat to see more things related to this.

C. L.

Unqualified comment: It depends upon how the numbers embody time and space for problem at hand, how they are separated and then what they are, continuous polychromatic wave forms or monochromatic band limited spectra. Controls depends upon the architecture (adequate propagation length especially is important) and distribution of the property of interest within the body of stimulated subject of interest.

For an isotropic and constant subject of interest with sufficient "depth" you can solve your problem by examining variability of (P- and S-wave) velocity for same frequency band, and for same velocity by variability of (P- and S-wave) frequency content.

There are other methods, too requiring detailed examination of signal morphology ( beyond scope of inquiry)

Over the years, data diversity and ambition to have an accurate picking making the development of new automatic techniques is an active field of research. But robust and accurate automatic algorithm that perform satisfactorily for all scenarios (type of source and receiver, noise level, application domain, among others) is not yet available.

We have recently developed a new adaptive algorithm to automatically pick the first arrivals in near-surface seismic data by combining three picking methods: multi-nested windows, higher order statistics, and Akaike information criterion (Article Development of an adaptive multi-method algorithm for automa...

I would recommend Anthony Lomax 's Filter Picker. It is very robust and has only 5 tuning parameters. It is widely used and therefore thoroughly tested. You can find the codes and related publications here:

http://alomax.free.fr/FilterPicker/

Colleagues, you are looking for an earthquake standard. This is the path to nowhere !!! It is necessary to study the trend of the seismic signal !!!. Every earthquake has its own individual trend !!! The problem of the earthquake prediction is finally solved by me !!! See my article, which is posted on my site. Sincerely, Vyacheslav Mikhailovich

You may use Fractal Dimension and LTA/STA methods for the automatic P & S waves detection & identification.

1) Article STA/LTA Fractal Dimension Algorithm of Detecting the P‐Wave Arrival

2) Article Detection of seismic refraction signals using a variance fra...

Hej all,

I would like to test all of your comments and will compare rusults.

Regards,

Colleagues, you are looking for an earthquake standard. This is the path to nowhere !!! It is necessary to study the trend of the seismic signal !!!. Every earthquake has its own individual trend !!! The problem of the earthquake prediction is finally solved by me !!! See my article, which is posted on my site. Sincerely, Vyacheslav Mikhailovich

Do not engage in this nonsense! I say to you for the thousandth time that it is necessary to monitor the change in time of the seismic signal and determine on this basis all the parameters of a future earthquake for a long time before its occurrence. Read the classics, gentlemen, my articles !! Sincerely, Diagnostic with 50 years of experience, Vyacheslav Mikhailovich

You can check the following paper:

Akram, Jubran, and David W. Eaton. "A review and appraisal of arrival-time picking methods for downhole microseismic dataArrival-time picking methods." Geophysics 81, no. 2 (2016): KS71-KS91.

You can use the LTA/STA method for the automatic P and S waves detection with data using SAC software.

Dear Colleagues,

Before, 4 months I had an intuitional moment about how we can do some of the types of Earthquakes prediction… I have mentioned it at Ilie Sandu question: How do You see the Earthquakes - Prediction, Forecast, or Random events? After had the luck to read the article of Vjacheslav Nagorny and saw profiles Sergey Shcherbina… From there, I got the next conclusion that the lot of earthquakes prediction will be resolved in the near future… Now I am referring only to  Vjacheslav Nagorny’s work… Yes, he made a very good work… The principle: -‘It is necessary to study the trend of the seismic signal !!!. Every earthquake has its own individual trend !!(....) I say to you for the thousandth time that it is necessary to monitor the change in time of the seismic signal and determine on this basis all the parameters of a future earthquake for a long time before its occurrence.’- is very good…But need something more such knowledge which has me why I know that is conception is good… In January  Vjacheslav, wrote to me at facebook account.’ Only an outsider can solve this problem. I took it and decided. Seismologists have stereotypes that they are afraid to give up. Vyacheslav’  the discussion can be read at my account of  Facebook https://www.facebook.com/laszloattila.horvath (Január 13.)

It’s went 2-4 months from the announcement one thing is interesting: the human individualism, egoism is bigger than the desire to test such a model of prediction which can help the human society. Nicola Tesla gave a lot for American society, we know well that how was appreciated his work…

Regards,

Laszlo 

Best method for automatic detect of P & S is LTA / STA method since which is examining within 5 seconds