PSN-L Email List Message

Subject: Re: Data filtering within WinQuake/WinSDR
From: chrisatupw@.......
Date: Sat, 28 Sep 2013 08:32:14 -0400 (EDT)

From: Thomas Dick 
To: PSN list 
Sent: Fri, 27 Sep 2013 20:02
Subject: Data filtering within WinQuake/WinSDR

I need some help. I've trying to solve this for over a year now.  If=20
look at Bob's (Locust Valley) and my post ( Evansville) for the Acari
Peru   quake on 9/25/13 -- in the data operations under Display/Modify
-- Bob does 3 manipulations and I do five. I understand the first which
is done within WinSDR, but what is wrong with what I am doing within

Additional comments on this area might help me too. My units detect
frequencies between .008 Hz and up to 1 Hz on teleseismic quakes and
higher up on regional events. I set WinSDr at Low Pass at 3 Hz --
should / could this be lower --- I assume this means I am cutting off
frequencies above 3 Hz with this filter, Strong ocean and Great Lake
storms require me to cut off sometimes at or above . 04 Hz. The words
Low pass cutoff and High pass cutoff as applied to  my location
situation seem confusing

   I know some of the purists want data that is unfiltered. For me I=20
want to see the finished product.

Hi Tom,

    I save the raw data from 5 Hz to ~30 seconds and then analyse it.
     I find that it helps to refer to frequency above 1 Hz and to=20
periods below this.
     0.008 Hz =3D 125 seconds ; 0.04 Hz =3D 25 seconds ; did you mean these=
please ?
     Microseisms here are about 0.2 Hz to 0.14 Hz - 5 to 7 seconds=20
period. I
occasionally also see ocean waves at about 0.1 Hz t 0.08 Hz - 10 to 12=20
Have you analysed the signals that you get from your Great Lakes ?
     Teleseismic quakes are usually from ~2 Hz to 30 seconds. Near=20
quakes, which you are likely to see, require a 5 Hz upper frequency,=20
but this limit
will still cut out the environmental noise, which may get serious above=20
~10/15 Hz.
     If you are using Larry's electronics, it amplifies from 10 Hz to 60=20

    A Low Pass filter passes frequencies BELOW the chosen value.
    A High Pass filter passes frequencies ABOVE the chosen value.
     A Band Pass filter only passes signals BETWEEN the upper and lower=20
     A Band Cut filter removes signals BETWEEN the upper and lower=20
chosen values.

     I put a LP filter at ~15 seconds into the raw data and use this to=20
find the long period
surface waves. I note the start times and then put a HP filter into the=20
data at about 2.5
/ 3 seconds, which avoids the microseisms and hopefully allows me to=20
identify the
relevant P and S waves. I don't see ANY Local quakes.
     Alternatively, you can use the STA/LTA filter to identify the P and=20
S waves directly
and only record short traces. But this can miss low amplitude P and S=20

     If you tell us what control buttons you are choosing, in what=20
sequence and the
periods / frequencies you have chosen, we may be able to help you more.=20
I presume
that you are using digital filtering ? Remember the limitations=20
involved. Do you see any
impulse interference / noise ?


     Chris Chapman=20

Public Seismic Network Mailing List (PSNLIST)

To leave this list email PSNLIST-REQUEST@.............. with 
the body of the message (first line only): unsubscribe
See for more information.

[ Top ] [ Back ] [ Home Page ]