pre_filt for HH* waveform remove response

Users Mailing List Archive
1

Hi

I’m trying to remove instrument response from channels HHZ,HHN,HHE, for events magnitude:2-4.5, and station distance to each event less than 5 degrees.

The waveforms come through FDSN -IRIS.

Q1 – what would you recommend to use for pre_filt = [?, ?, ?, ?] and water level parameters in :

remove_response(inventory=inventory , pre_filt=pre_filt, water_level=60, plot= False, zero_mean= True)

should I change any other default parameters?

Q2- would you recommend to get these waveforms from a different web service provider?

Thank you

Payman

2

Appropriate settings of pre_filt and/or water_level depend mainly on

- the used instruments (their instrument response)
- the frequency range of interest (for your magnitude range, i'm
   guessing something like 0.2-20 Hz)

You might want to visualize your instruments response, first of all. See:

https://docs.obspy.org/packages/autogen/obspy.core.inventory.inventory.Inventory.plot_response.html

If the channels are named correctly and you only have HH streams, you
should be looking at broadband data (corner frequency e.g. 120s). In
that case it's trivial, just use pre_filt setting well outside of your
desired target frequency range.

E.g.
- pre_filt=[0.02, 0.05, 50, 70]
- water_level=None

But if you have short period instruments as well, there's always a
trade-off because you might be interested in signal content in frequency
ranges where sensitivity of your instrument is going down already.

Please note the option to plot all steps during instrument response removal:

http://docs.obspy.org/packages/autogen/obspy.core.trace.Trace.remove_response.html?highlight=plot

Where the waveforms come from shouldn't matter as long as you get them.
FDSNWS is usually the best choice (nowadays preferred distribution
method by most data centers).

best,
T