Improving ambient noise cross-correlations in the noisy ocean bottom environment of the Juan de Fuca plate

Amphibious Shear Wave Structure Beneath the Alaska‐Aleutian Subduction Zone From Ambient Noise Tomography

“…Continuous daily records from those stations are corrected according to the clock correction list. After the clock correction, for continuous seismic records extracted from OBS stations, tilt and compliance noise are removed from vertical component using transfer functions constructed from the horizontal component and pressure gauge (e.g., Bell et al, 2014;Janiszewski et al, 2019;Tian & Ritzwoller, 2017). Then I apply routine two station ambient noise interferometry process (noise correlation) to both land and OBS data by following the procedures described by Bensen et al (2007) and Ritzwoller and Feng (2019).…”

Overview of Pre- and Post-Processing of Ambient-Noise Correlations

“…Zheng et al (2011) found that they needed to use the opposite correlation time lag from the direction to Aso volcano. Tian and Ritzwoller (2017) identified persistent signals in their cross-correlations between OBS stations that they argued were related to local persistent microseismic sources in shallow waters of the Juan de Fuca plate.…”

“…These so called "denoising" techniques have been shown to improve the SNR of earthquake data recorded on the seafloor and to reduce distortions (e.g., Ball et al, 2014;Bell et al, 2015;Dolenc et al, 2007). Additionally, Bowden et al (2016) and Tian and Ritzwoller (2017) have shown that these methods also improve vertical (ZZ) component cross-correlations of ambient noise by improving the SNR of the first-overtone and the fundamental mode Rayleigh waves, particularly for shallower water stations, and, importantly, by extending the measurements to longer periods. Tian and Ritzwoller (2017) also showed that the methods produce records that provide more information about the origin of the coherently propagating ambient noise that is the basis for ambient noise tomography.…”

“…Additionally, Bowden et al (2016) and Tian and Ritzwoller (2017) have shown that these methods also improve vertical (ZZ) component cross-correlations of ambient noise by improving the SNR of the first-overtone and the fundamental mode Rayleigh waves, particularly for shallower water stations, and, importantly, by extending the measurements to longer periods. Tian and Ritzwoller (2017) also showed that the methods produce records that provide more information about the origin of the coherently propagating ambient noise that is the basis for ambient noise tomography. Figure 5.12 presents example record sections of vertical component ambient noise cross-correlations in two pass-bands observed using Cascadia Initiative data from the Juan de Fuca Plate.…”

“…This figure is adapted from Fig. 8 of Tian and Ritzwoller (2017). record section is in shallow water, station J49A at a depth of 123 m. The reduction of tilt and compliance noise greatly improves the SNR of the re-covered Rayleigh waves, particularly at periods above 20 s that are difficult to observe without de-noising.…”