Seismological constraints on source properties of the mb=4.0, 1985 Ardsley, New York, earthquake: A characteristic rupture?

Abstract

On October 19, 1985, a magnitude 4.0 earthquake occurred in southern Westchester County, New York, approximately 30 km north of central Manhattan. Barstow et al. (1988) concluded that this earthquake consisted of two sub events, with the larger sub event occurring 0.6 s after the first one. We present analysis of data for the main shock as well as aftershocks and foreshocks which were recorded on the New York State Seismic Network. Because this data set includes a range of source magnitudes from less than 2 to 4 recorded for virtually identical source‐receiver pairs, an empirical Green's function technique can be used to study the source characteristics of the larger events. The corner frequency of the main shock and an attenuation parameter t* are resolved from the initial portion of the seismogram at station PAL in Palisades, New York, 8 km northwest of the epicenter. Spectra of small foreshocks and aftershocks can then be corrected for attenuation using t*. Following an iterative procedure, it is then possible to make reliable determinations of seismic corner frequency for these events. Several conclusions are obtained: (1) The spectrum of the first subevent in the main shock can be well modelled by a Brune spectrum from 1 to 10 Hz, with a corner frequency of 7 ± 1 Hz. At higher frequencies, the source spectrum contains sharp peaks which do not fit an omega‐square model and may reflect complexities in the rupture process. (2) The average stress drop of the sequence is ≈ 205 bars. The stress drop of the main event is 540 bars, but no stress drop dependency on moment is resolved over the magnitude range 1.8–4.0. (3) Attenuation parameters t* are obtained for the 8 km northwestward path from Ardsley to Palisades and for a 52‐km westward path to a station in the Hudson Highlands, and their similarity suggests an increase of Q with depth. (4) Assuming that the two main shock sub events did not rerupture the same area but were contiguous, a maximum rupture dimension is 440–680 m. The rupture diameter of the largest aftershock, mb=3, is determined to be approximately 320 meters. The combined rupture for the sequence inferred from spectral methods is consistent with the size of the rupture inferred from the area surrounded by the early aftershocks (700 m diameter). The rupture dimension of the Ardsley sequence may be characteristic of the Dobbs Ferry fault and other seismogenic faults in the Manhattan Prong, which are segmented with characteristic dimensions of 0.5–2 km (Seeber and Dawers, 1989). Assuming that the range of stress drops in the Ardsley sequence is also characteristic of the Manhattan Prong, this range of dimensions corresponds to a range of characteristic magnitudes between mb of 3.3 and 5.5. We note that this magnitude range encompasses the five largest known earthquakes in the New York City seismic zone: the 1985 Ardsley main shock and events in 1737, 1783, 1884, and 1895.