Transverse Minimisation Method

If the polarisation of the shear energy is known then the energy on the transverse component is indicative of splitting (Silver and Chan, 1998; Vinnik et al., 1989). The method works by searching for the splitting parameters that when applied to the data removes the most energy from the transverse component. In principle the method is very similar to the eigenvalue method of Silver and Chan 1991, except the polarisation of the shear energy us defined by the user, rather than being searched for by principal component analysis. By including the polarisation as a priori information the measurement is more robust to noise, however, the method is vulnerable to error if the polarisation is specified incorrectly.

Continuing with the Real data example, for which we obtained the following plot using the eigenvalue method.

To do the transverse minimisation method we use the TransM class with the polarisation specified. In the case of the SKS wave the polarisation is equal to the backazimuth direction, this can be calculated using obspy.

# use obspy to get backazimuth
from obspy import geodetics
dist, az, baz = geodetics.base.gps2dist_azimuth(evlat,evlon,stlat,stlon)

# make the measurement
m = sw.TransM(realdata, pol=baz, lags=(2,))
m.plot()

Notice that the transverse minimisation method returns a more focussed result.

Now checkout the Rotation Correlation Method.