In situ observations of velocity changes in response to tidal deformation from analysis of the high‐frequency ambient wavefield

Abstract

We report systematic seismic velocity variations in response to tidal deformation. Measurements are made on correlation functions of the ambient seismic wavefield at 2–8 Hz recorded by a dense array at the site of the Piñon Flat Observatory, Southern California. The key observation is the dependence of the response on the component of wave motion and coda lapse time τ. Measurements on the vertical correlation component indicate reduced wave speeds during periods of volumetric compression, whereas data from horizontal components show the opposite behavior, compatible with previous observations. These effects are amplified by the directional sensitivities of the different surface wave types constituting the early coda of vertical and horizontal correlation components to the anisotropic behavior of the compliant layer. The decrease of the velocity (volumetric) strain sensitivity Sθ with τ indicates that this response is constrained to shallow depths. The observed velocity dependence on strain implies nonlinear behavior, but conclusions regarding elasticity are more ambiguous. The anisotropic response is possibly associated with inelastic dilatancy of the unconsolidated, low‐velocity material above the granitic basement. However, equal polarity of vertical component velocity changes and deformation in the vertical direction indicate that a nonlinear Poisson effect is similarly compatible with the observed response pattern. Peak relative velocity changes at small τ are 0.03%, which translates into an absolute velocity strain sensitivity of Sθ≈5 × 103 and a stress sensitivity of 0.5 MPa−1. The potentially evolving velocity strain sensitivity of crustal and fault zone materials can be studied with the method introduced here.

Noise-based monitoring and imaging of aseismic transient deformation induced by the 2006 Basel reservoir stimulation

“…Solution quality can further be improved by considering correlations constructed from all three components if available (Schaff, 2012;Hillers et al, 2014), which might also reveal anisotropic responses (Hillers et al, 2015b). Fluctuations in relative velocity change measurements are attenuated using weighted inversion schemes .…”

“…Practitioners homogenize the frequency content of the ambient seismic wavefield and attenuate the influence of source-related transients using a range of frequency-and time-domain normalizations (e.g., Bensen et al, 2007). Our preprocessing chain has been applied previously in a variety of imaging and monitoring contexts (Poli et al, 2012;Boué et al, 2013;Hillers et al, 2014Hillers et al, , 2015b. The daily seismograms were initially high-pass filtered and then split into six 4-hlong segments.…”

“…For the stretching method, this means that an optimal dilation of the test waveform (i.e., matching the velocity change) cannot account for the residual dc (i.e., induced by changes in the scattering structure). The lapse time τ range for which scattered coda waves have sufficient energy can be estimated from the decay behavior of the average correlation coda envelope, as assessed on a logarithmic scale (Hillers et al, 2015b). Arrivals with energy above an asymptotic background level are associated with strong multiple scattered phases, in contrast to the relatively incoherent noise that dominates the waveforms in the tail.…”

“…Here, we focus on monitoring methods to track systematic changes in the crosscorrelation coda wavefield that are governed by variations in the elastic or scattering properties of the heterogeneous medium. In seismological contexts, these principles have been applied to resolve relative velocity changes between Oð0.01%Þ and Oð1%Þ associated with volcanic activity (Brenguier et al, 2008b;Obermann et al, 2013a), rapid (Wegler and Sens-Schönfelder, 2007;Brenguier et al, 2008a;Wegler et al, 2009;Hobiger et al, 2012;Froment et al, 2013) and slow (Rivet et al, 2011) slip on earthquake faults, water content in the shallow crust (Sens-Schönfelder and Wegler, 2006a;Meier et al, 2010;Froment et al, 2013;Hillers et al, 2014), thermal processes (Sens-Schönfelder and Richter et al, 2014;Hillers et al, 2015a), and tidal-induced deformation (Hillers et al, 2015b). Scattered wavefield sensitivity allows imaging of static and transient variations in the properties of structural and mechanical media (Larose et al, 2010;Rossetto et al, 2011;Obermann et al, 2013aObermann et al, , 2014Planès et al, 2015).…”

The time dependence of rock healing as a universal relaxation process, a tutorial

“…As a result, friction laws that are state-and rate-dependent have been applied to describe friction and slip on faults (Dieterich 1978(Dieterich , 1979Ruina 1983;Heslot et al 1994;Scholtz 1997). The imprint of deformation on the mechanical properties of rocks implies that the mechanics of rocks is, in general, nonlinear (Guyer & Johnson 2009;Rivière et al 2013Rivière et al , 2015, and observations show that the seismic velocity and/or attenuation changes systematically with earth tides (Reasenberg & Aki 1974;Yamamura et al 2003;Hillers et al 2015). Because of the dominance of shear motion in strongly scattering elastic media (Aki & Chouet 1975;Weaver 1982;Snieder 2002), the measured change in arrival times of seismic waves corresponds for such media mostly to a change in shear velocity.…”