A combined phase-matched-filtering and narrow-bandpass-filtering method and a surface-wave Love-to-Rayleigh spectral-amplitude-ratio method are used to analyze the 1982 New Brunswick earthquake sequence. The analysis was carried out on the long-period fundamental-mode surface-wave signals (10 to 50 sec) for the main shock and its four largest aftershocks recorded at three GDSN stations (SCP, LON, and ALQ) in North America.
Estimates of source mechanism parameters (focal depth, dip, rake, and strike) obtained for the 9 January 1982 main shock and the largest aftershock on 11 January 1982 are consistent with Wetmiller et al.'s (1984) results for these events using local and regional body-wave observations. Source mechanism parameters were also obtained for the 9 January 1982, the 31 March 1982, and the 16 June 1982 aftershocks which do not have independent estimates. Since these three events have low signal-to-noise ratios (S/N) at some frequencies, the source parameter estimates are not as well constrained as those of the main shock and the 11 January aftershock. The inferred principal stress orientations are in all cases consistent with the prevailing ENE maximum compressive stress characteristic of most areas in eastern North America.
Among the principal aftershocks, the most reliable source parameter estimates were for the 11 January aftershock because of its consistently high S/N seismograms. This aftershock consisted of thrusting on a fault plane conjugate to the main shock fault plane with dip of 40° to 55° to the northeast; rake of 60° to 80°; strike of 337° to 342°; and a depth of 6 km. This result also supports Wetmiller et al.'s (1984) findings for this event. The 9 January aftershock has a source mechanism similar to that of the main shock but a shallower depth of 6 km (versus 7 km for the main shock). The 31 March aftershock is associated with the upper portion of the main shock rupture, with a much shallower depth of approximately 3 km. The result for the 16 June aftershock is not as dependable; but it appears to have a deeper depth of 8 km or more, which suggests extension of the rupture zone to greater depth.
The results of this study indicate that surface-wave data from only a few three-component far-field stations can be used to infer source parameters (including depth) for events as small as about mb 4.5. To the extent that eastern North America is typical of other continental interiors in terms of crustal structure and attenuation, similar capabilities are anticipated for such areas.