6.2.8.1.7. eqcorrscan.utils.mag_calc.relative_magnitude

eqcorrscan.utils.mag_calc.relative_magnitude(st1, st2, event1, event2, noise_window=(-20, -1), signal_window=(-0.5, 20), min_snr=5.0, min_cc=0.7, use_s_picks=False, correlations=None, shift=0.2, return_correlations=False, correct_mag_bias=True)[source]

Compute the relative magnitudes between two events.

See eqcorrscan.utils.mag_calc.relative_amplitude() for information on how relative amplitudes are calculated. To compute relative magnitudes from relative amplitudes this function can weight the amplitude ratios by the cross-correlation of the two events. The relation used is similar to Schaff and Richards (2014), equation 4 and is:

\[\Delta m = \log{\frac{std(tr2)}{std(tr1)}} + \log{ \frac{(1+\frac{1}{snr_x^2})}{1+\frac{1}{snr_y^2}}\times CC}\]

If you decide to use this function you should definitely read the paper to understand what you can use this for and cite the paper!

Parameters:

  • st1 (obspy.core.stream.Stream) – Stream for event1

  • st2 (obspy.core.stream.Stream) – Stream for event2

  • event1 (obspy.core.event.Event) – Event with picks (nothing else is needed)

  • event2 (obspy.core.event.Event) – Event with picks (nothing else is needed)

  • noise_window (tuple of float) – Start and end of noise window in seconds relative to pick

  • signal_window (tuple of float) – Start and end of signal window in seconds relative to pick

  • min_snr (float) – Minimum signal-to-noise ratio allowed to make a measurement

  • min_cc (float) – Minimum inter-event correlation (between -1 and 1) allowed to make a measurement.

  • use_s_picks (bool) – Whether to allow relative amplitude estimates to be made from S-picks. Note that noise and signal windows are relative to pick-times, so using an S-pick might result in a noise window including P-energy.

  • correlations (dict) – Pre-computed dictionary of correlations keyed by seed-id. If None (default) then correlations will be computed for the provided data in the signal_window.

  • shift (float) – Shift length for correlations in seconds - maximum correlation within a window between +/- shift of the P-pick will be used to weight the magnitude.

  • return_correlations (bool) – If true will also return maximum correlations as a dictionary.

  • correct_mag_bias (bool) – Whether to correct for the magnitude-bias introduced by cc<1 and the presence of noise (i.e., SNR << ∞). Without bias-correction, the relative magnitudes are simple L2-norm-ratio relative magnitudes.

Return type:

dict

Returns:

Dictionary of relative magnitudes keyed by seed-id