PSN-L Email List Message

Subject: Re: Phase picking
From: Larry Cochrane lcochrane@..............
Date: Tue, 31 May 2005 23:30:57 -0700

Hi Bob,

WinQuake can be used to learn how to pick the P and S waves using known event 
information. To do this, the following things need to be correct:

The data needs to be accurately timestamped and the location of the sensor must be 
known to a few meters. Nowadays, using GPS this shouldn't be a problem.

The reporting agency needs to report the location, depth and event time correctly.

The sensor to event distance must be under ~11000km do to the shadow effect of the 
earths core.

If these things are correct WinQuake, using the travel time tables, can place the P 
and S markers very accurately on the seismogram. I'm constantly amazed at how 
accurate the travel-time tables can be. The waves can take over 10 minutes to get to 
the station and the calculated P wave arrival time is less then two or three seconds 
off from the actual arrival time. That's less then a 0.5% error.

With a good signal to noise seismogram, the P wave should be easily picked no matter 
if the data is in acceleration, velocity or displacement. It's the point where you 
see the first out of noise signal from the event. Normally a vertical sensor will 
display the P wave more clearly then a horizontal sensor.

Picking the S wave is much harder. A horizontal senor will display the S wave better 
so using this type of sensor makes it easier to pick the start of this wave.

The depth of the event also makes a difference. With one seismogram it's hard to 
determine this parameter. One clue that the event is under 33 km is the lack of large 
surface waves.

Larry Cochrane
Redwood City, PSN

Bobhelenmcclure@....... wrote:
> Hi all,
>   I live on Long Island, New York. Local events are very rare here, so what I 
> detect are mostly teleseisms.  I consider myself pretty competent at the 
> recording and signal processing of such events to obtain accurate broadband 
> waveform reproduction and timing. My event waveforms correspond well with those of 
> nearby LD network station PAL.
>   I do not consider myself competent at P and S phase picking, and need 
> advice. What is the criterion for judging phase arrival time? Is it the first 
> velocity peak, emergent velocity, first displacement peak, emergent displacement, 
> or what? Most of the S phases I record have fairly long period, and the P minus 
> S time estimate depends greatly on the S phase time pick.
> Cheers,
> Bob (PSN station rem)


Public Seismic Network Mailing List (PSN-L)

[ Top ] [ Back ] [ Home Page ]