Calibration studies at TXAR (Lajitas, Texas) used a modified version of the correlation method described by Cansi et al. (1993) in order to estimate azimuth and horizontal phase velocity of 144 events for which USGS mb values were available. Modifications to the correlation method include the Fourier interpolation of the data by a factor of 8 to obtain a virtual sample rate of 320/sec, use of an L1 norm (least absolute deviation) to obtain estimates of the azimuth and phase velocity, and a moving window display to indicate those portions of the waveform that show strongest correlation across the array.
Corrected phase velocities normally associated with Pn (less than 8.6 km/s) are generally seen for events at epicentral distance as far as 2,000 km. Upper mantle refracted first arrivals (P) with corrected phase velocities greater than 8.6 km/s are generally observed for epicentral distances beyond about 1,600 km. Phase identification is essential in order to select a suitable magnitude scale.
Based on the 144 well-located events by USGS and using the Denny et al. (1987) formula, the most reliable magnitude estimates are as follows:
For horizontal phase velocity less than 8.6 km/sec:
mb(D) = log A + 2.4 (log D) - 3.95 + C, with C = +0.3
For horizontal phase velocity greater than 8.6 km/sec:
mb(D) = log A + 2.4 (log D) - 3.95 + C, with C = −0.50
The M-discontinuity beneath TXAR was determined to the first order to strike along an azimuth of 109 degrees (NW-SE) and dip 11 degrees to the northeast. This result is consistent with the tectonic setting for the area.