Apply travel time ellipticity correction to inversion inputs

[geoscienceaustralia]

Apply travel time ellipticity correction to inversion inputs

[basaks]

@alexgorb how critical is this? Or are we ok for now?

We have the ellipticity correction working, so this should not be too difficult to implement, but the integration to the clustering code is yet to be done.

[alexgorb]

Ellipticity correction is critical.

[basaks]

We should also package the fortran code of ellipticity corection with passive-seismic for one step installation: #65.

[basaks]

This has been achieved in # 4537ea7.

However, @alexgorb , we have to pay some attention to ellipticity correction parameters used in the fortran code here. Specifically, the ecolat parameter has been defined as the co-latitude of the event location. However, and I have a TODO in the code, we need to make sure that the events that we are gathering arrivals from have latitudes defined in range (-90 to 90).

From the TODO:

# TODO: check co-latitude definition
# no `ecolat` bounds check in fortran ellipcorr subroutine
# no `origin.latitude` bounds check in obspy

Keeping this open for now until those are confirmed.

[basaks]

@alexgorb The above is important as if the events have their origins defined using a different range of latitudes, will result in wrong ellipticity corrections as the clustering code expects latitude defined in range [-90, 90] and converts the latitude to co-latitude.