The cases where a solution will not be attempted are as follows:

- Multiple phases at two stations, but no azimuths. This is a non-unique case, even though four different arrivals are present.
- Less than three phases from three different stations and no azimuths.
- A single phase at one station with an azimuth.

Note that if phases are weighted out due to large distance or a bad fit during the first iteration, there might not be a location even if more than 3 stations are available.

Weighting:

A number of different weights may be used to calculate the solution.

- 1.
- User specified weights: These are calculated using the HYPO71 style weight number 0 to 4, read with each phase, where 0 corresponds to w1=1.0, 1 to w1=0.75, 2 to w1=0.5, 3 to w1=0.25 and 4 to w1=0. Uncertain time is 9 meaning that absolute time is not used, see also use of S-P times on previous page.
- 2.
- Distance weighting: This is given by the formula w2=(xfar-delta)(xfar-xnear) where delta is the distance (km) of the event from the station and xnear and xfar are read from the station file, STATION0.HYP.
- 3.
- Bisquare weighting: This scheme, described by Anderson [1982] calculates residual weights, see details in HYP manual.
- 4
- Azimuth weighting: Azimuth residuals are divided by test(52), which is the error in azimuth that corresponds to a one-second error in arrival time. For example, if test(52)=5 (default), a phase residual of 5 degrees will become a residual of 1 (5/test(52)) in the parameter corrections and rms calculation.

All the above weights are multiplied together to calculate the weight used in the inversion. If the user-specified weight, w1, is changed by (2) or (3) above, changed to zero by the consistency check, or set to -1 because the phase is not recognized, an asterisk will appear after the final weight in the residual printout.

Peter Voss : Fri Nov 12 10:33:10 UTC 2021