Investigating the Capability to Extract Impulse Response Functions From Ambient Seismic Noise Using a Mine Collapse Event

Abstract

Using recordings of a mine collapse event (Mw 4.2) in South Korea in January 2015, we demonstrated that the phase and amplitude information of impulse response functions (IRFs) can be effectively retrieved using seismic interferometry. This event is equivalent to a single downward force at shallow depth. Using quantitative metrics, we compared three different seismic interferometry techniques—deconvolution, coherency, and cross correlation—to extract the IRFs between two distant stations with ambient seismic noise data. The azimuthal dependency of the source distribution of the ambient noise was also evaluated. We found that deconvolution is the best method for extracting IRFs from ambient seismic noise within the period band of 2–10 s. The coherency method is also effective if appropriate spectral normalization or whitening schemes are applied during the data processing.

Generation of Infrasound Waves at Sites of Underground Mine Collapses

“…Based on the amplitudes of the short‐period waves, the local seismic magnitudes of the 2015 and 2016 collapses were estimated to be 2.5 and 3.5, respectively, whereas the moment magnitudes were estimated to be much larger, 4.2 and 4.4, respectively, because of enhanced long‐period energy. The focal mechanism solutions indicated that the source type of both events was close to crack closure, that is, an implosion source (Kwak et al, ). The magnitude discrepancy, and in particular the relative enhancement of surface waves, can be interpreted as being caused by shallow seismic sources; for example, landslides that have a large difference between surface and local magnitudes are distinct from tectonic earthquakes (Weichert et al, ).…”

Optimal Stacking of Noise Cross-Correlation Functions

“…Cross-correlations of ambient seismic noise have been widely used to image the Earth's elastic (L. Feng and Ritzwoller 2019;Shapiro et al 2005; S.-M. Wu et al 2021;X. Yang andGao 2018, 2020) and anelastic structure (Prieto et al 2009), model ground motions (Marine A Denolle et al 2013Denolle et al , 2014Denolle et al , 2018Kwak et al 2017;Viens and Marine A Denolle 2019;Viens et al 2017), and monitor transient velocity changes in the shallow subsurface (F. Clements and Marine A Denolle 2018;Donaldson et al 2019;K.-F. Feng et al 2021;Olivier et al 2019; Q.-Y. Wang et al 2017;Z.…”

Improving the Retrieval of Offshore‐Onshore Correlation Functions With Machine Learning

“…Empirical studies showed that seismic interferometry by deconvolution with no preprocessing can be used to retrieve both the amplitude and phase information of CCFs (Viens et al., 2017). Deconvolution functions (DFs) have been used to simulate the long‐period ground motions from moderate (Denolle et al., 2013; Prieto & Beroza, 2008; Sheng et al., 2017; Viens et al., 2014; Viens, Koketsu, et al., 2016) and large (Denolle et al., 2014, 2018; Viens, Miyake, & Koketsu, 2016) crustal earthquakes as well as mine collapse events (Kwak et al., 2017). However, the retrieval of reliable amplitudes is still debated as it strongly depends on the location and characteristics of ambient seismic field sources (Stehly & Boué, 2017; Stehly et al., 2006; Tsai, 2011).…”