Arrival angle anomalies of Rayleigh waves observed at a broadband array: a systematic study based on earthquake data, full waveform simulations and noise correlations

Abstract

S U M M A R YDeviation of seismic surface waves from the great-circle between source and receiver is illustrated by the anomalies in the arrival angle, that is the difference between the observed backazimuth of the incident waves and the great-circle. Such arrival angle anomalies have been known for decades, but observations remain scattered. We present a systematic study of arrival angle anomalies of fundamental mode Rayleigh waves (20-100 s period interval) from 289 earthquakes and recorded by a broadband network LAPNET, located in northern Finland. These observations are compared with those of full waveform synthetic seismograms for the same events, calculated in a 3-D Earth and also compared with those of seismograms obtained by ambient noise correlation. The arrival angle anomalies for individual events are complex, and have significant variations with period. On average, the mean absolute deviation decreases from ∼9 • at 20 s period to ∼3 • at 100 s period. The synthetic seismograms show the same evolution, albeit with somewhat smaller deviations. While the arrival angle anomalies are fairly well simulated at long periods, the deviations at short periods are very poorly modelled, demonstrating the importance of the continuous improvement of global crustal models. At 20-30 s period, both event data and numerical simulations have strong multipathing, and relative amplitude changes between different waves will induced differences in deviations between very closely located events. The source mechanism has only limited influence on the deviations, demonstrating that they are directly linked to propagation effects, including near-field effects in the source area. This observation is confirmed by the comparison with seismic noise correlation records, that is where the surface waves correspond to those emitted by a point source at the surface, as the two types of observations are remarkably similar in the cases where earthquakes are located close to seismic stations. This agreement additionally confirms that the noise correlations capture the complex surface wave propagation.

Seismic anisotropy of Precambrian lithosphere: Insights from Rayleigh wave tomography of the eastern Superior Craton

“…Bias may arise from angular deviations of the surface wave propagation paths from their predicted great-circle arc [Pedersen, 2006], with deviations of ≤20 ∘ possible [e.g., Cotte et al, 2000;Foster et al, 2014]. At short periods (∼20 s) great-circle arc deviations may be as high as 15 ∘ [Pedersen et al, 2015]. Foster et al [2014] showed that the standard deviation of the distribution of great-circle arc deviations decreases from 0.4 at short periods (∼20 s) to less than 0.1 at longer periods (∼100 s).…”

Using surface waves recorded by a large mesh of three-element arrays to detect and locate disparate seismic sources

“…The legends are similar to those in Figs 4 and 7. & Ekström 2002;Foster et al 2014;Pedersen et al 2015;Doran & Laske 2017). These complexities can result in deviations from the great circles, causing arrival angle anomalies.…”

Generalized interferometry – I: theory for interstation correlations

“…14(b), several earthquakes may be included. Their position is not restricted to the great-circle path that may be a poor approximation to the actual wave propagation path in 3-D heterogeneous media (Pedersen 2006;Pedersen et al 2015). The use of earthquakes located off the great-circle, anticipated by Passier & Snieder (1995), has the potential to significantly increase the amount of exploitable information.…”