Source processes of three aftershocks of the 1983 Goodnow, New York, earthquake: High-resolution images of small, symmetric ruptures

• Main Authors: Xie, Jiakang, Liu, Zuyuan, Herrmann, Robert B., Cranswick, Edward
• Format: Article in Journal/Newspaper
• Language: unknown
• Published: Seismological Society of America 1991
• Subjects: Northwest Territories
• Online Access: https://authors.library.caltech.edu/48065/; https://authors.library.caltech.edu/48065/1/818.full.pdf; https://resolver.caltech.edu/CaltechAUTHORS:20140806-102734498

Broadband, large dynamic range GEOS data from four aftershocks (M_L ∼ 2 to 3) of the 1983 Goodnow, New York, earthquake, recorded at hard-rock sites 2 to 7 km away from the epicentral area, are used to study rupture processes of three larger events, with an M_L = 1.6 event as the Green's function event. We analyze the spectra and spectral ratios of ground velocity at frequencies up to 100 Hz, and conclude that (1) there are resolvable P-wave f_(max) (51 and 57 Hz) at two sites; (2) there is abundant information on sources of larger events up to frequencies of 50 to 60 Hz; and (3) there is an unstable, nonlinear instrument resonance at about 90 Hz. We analyze the artifacts of low-pass filters in the deconvolved rupture process, including the limited time resolution and biases in the rise-time measurement. Some extensions of the empirical Green's function (EGF) method are proposed to reduce these artifacts and to precisely estimate the relative locations of events that are close both in time and in space. Applying the EGF method to three aftershocks, we find that these events have ruptures that are simple crack-like, characterized by small fault radii (∼ 70 to 120 m) and static stress drops that vary, depending on the size of the events, between about 5 to 16 bars. We also find that two of the events, with origin times differing by 0.60 sec, are separated by 190 ± 110 meters. Assuming a causal relationship between the two would result in a slow propagation velocity (0.05 ± 0.03 time of the local shear-wave velocity). We therefore interpret the corresponding ruptures as being distinct, the area between the rupturing patches having a characteristic length much smaller than those of the rupturing patches. Comparison of the results of this study with those obtained for the Goodnow main shock and microearthquakes in California and Hawaii suggests that the stress drops of the Goodnow aftershocks decrease considerably (by up to a factor of 25 or more) from those estimated for the main shock, even after the estimated large uncertainty in the latter estimates is considered. This decrease is similar to those reported for the Nahanni earthquakes (m_b = 5.0 to 6.5) in Northwest Territories and for the North Palm Springs earthquake (M_L = 5.9) sequence in California. The stress drops of the simple, crack like microearthquakes are significantly variable (by factors between 3 and 27) within the single source areas. The median of stress drops of the intraplate Goodnow aftershocks is lower (by factors of 2 to 4) than the medians calculated for the other, interplate microearthquake sequences.