Nori Nakata

I use ambient noise, especially traffic noise, to estimate the 2D near-surface S-velocity distribution. Near-surface velocities are useful for understanding structure, stiffness, porosity, and pore pressure for engineering/environmental purposes and static correction of active-source imaging. I extract Love waves propagating between each receiver pair from 12 h of traffic noise using seismic interferometry with power-normalized crosscorrelation. The receiver array contained three parallel lines, each of which had 100 transverse-component geophones. I apply double beamforming to the correlations at the parallel lines for improving the signal-to-noise ratio of the extracted Love waves to satisfy the stationary phase assumption for seismic interferometry. I use these Love waves for a dispersion analysis to estimate a 2D near-surface S-wave velocity model based on the multichannel analysis of surface waves. To improve the lateral resolution of the velocity model, I sort the extracted waves according to common midpoints (CMPs) and limited the maximum offset of receiver pairs. The dispersion analysis at each CMP is based on the assumption of layered media, and using all CMPs, I can estimate high-resolution 2D velocities down to 80 m depth. The velocity variations are similar to the location of strong reflectors obtained by a previous study. The main features of the velocity model are recovered even from 1 h of continuous traffic-noise data, which means that the proposed technique can be used for efficient 4D surveys.

Time reversal is a powerful tool used to image directly the location and mechanism of passive seismic sources. This technique assumes seismic velocities in the medium and propagates time-reversed observations of ground motion at each receiver location. Assuming an accurate velocity model and adequate array aperture, the waves will focus at the source location. Because we do not know the location and the origin time a priori, we need to scan the entire 4D image (3D in space and 1D in time) to localize the source, which makes time-reversal imaging computationally demanding. We have developed a new approach of time-reversal imaging that reduces the computational cost and the scanning dimensions from 4D to 3D (no time) and increases the spatial resolution of the source image. We first individually extrapolate wavefields at each receiver, and then we crosscorrelate these wavefields (the product in the frequency domain: geometric mean). This crosscorrelation creates another imaging condition, and focusing of the seismic wavefields occurs at the zero time lag of the correlation provided the velocity model is sufficiently accurate. Due to the analogy to the active-shot reverse time migration (RTM), we refer to this technique as the geometric-mean RTM or GmRTM. In addition to reducing the dimension from 4D to 3D compared with conventional time-reversal imaging, the crosscorrelation effectively suppresses the side lobes and yields a spatially high-resolution image of seismic sources. The GmRTM is robust for random and coherent noise because crosscorrelation enhances signal and suppresses noise. An added benefit is that, in contrast to conventional time-reversal imaging, GmRTM has the potential to be used to retrieve velocity information by analyzing time and/or space lags of crosscorrelation, which is similar to what is done in active-source imaging.

Body wave reconstruction from ambient seismic noise correlations is an important step toward improving volcano imaging and monitoring. Here we extract body and surface waves that propagate in Piton de la Fournaise volcano on La Réunion island using ambient noise cross correlation and array‐processing techniques. Ambient noise was continuously recorded at three dense arrays, each comprising 49 geophones. To identify and enhance the Green's function from the ambient noise correlation, we apply a double beamforming (DBF) technique between the array pairs. The DBF allows us to separate surface and body waves, direct and reflected waves, and multipathing waves. Based on their azimuths and slownesses, we successfully extract body waves between all the combinations of arrays, including the wave that propagates through the active magmatic system of the volcano. Additionally, we identify the effects of uneven noise source distribution and interpret the surface wave reflections.

Earth's seismic velocity structure is heterogeneous at all scales, and mapping that heterogeneity provides insight into the processes that create it. At large scale lengths, seismic tomography is used to map Earth structure deterministically. At small scale lengths, structure can be imaged deterministically, but because it is impractical to image short-wavelength heterogeneity everywhere, we often resort to statistical methods to depict its variability. In this study, we develop random-field model representations of a 3-D P-wave velocity model at Long Beach, California, estimated from dense-array recordings of the ambient seismic wavefield. We focus on heterogeneity at the mesoscale, which is smaller than 10+ km scale of regional tomography but larger than the micro scale of borehole measurements. We explore four ellipsoidally anisotropic heterogeneity models, including von Kármán, Gaussian, self-affine and Kummer models, based on their autocorrelation functions. We find that the von Kármán model fits the imaged velocity model best among these options with a correlation length in the horizontal direction about five times greater than in the vertical direction, and with strong small-scale length variations. We validate our results by showing that our model accurately predicts the observed decay of scattered waves in the coda of a nearby earthquake, suggesting that quantitative measures of velocity variability will be useful for predicting high-frequency ground motion in earthquakes.

We retrieve P diving waves by applying seismic interferometry to ambient‐noise records observed at Long Beach, California, and invert travel times of these waves to estimate 3‐D P wave velocity structure. The ambient noise is recorded by a 2‐D dense and large network, which has about 2500 receivers with 100 m spacing. Compared to surface wave extraction, body wave extraction is a much greater challenge because ambient noise is typically dominated by surface wave energy. For each individual receiver pair, the cross‐correlation function obtained from ambient‐noise data does not show clear body waves. Although we can reconstruct body waves when we stack correlation functions over all receiver pairs, we need to extract body waves at each receiver pair separately for imaging spatial heterogeneity of subsurface structure. Therefore, we employ two filters after correlation to seek body waves between individual receiver pairs. The first filter is a selection filter based on the similarity between each correlation function and the stacked function. After selecting traces containing stronger body waves, we retain about two million correlation functions (35% of all correlation functions) and successfully preserve most of body wave energy in the retained traces. The second filter is a noise suppression filter to enhance coherent energy (body waves here) and suppress incoherent noise in each trace. After applying these filters, we can reconstruct clear body waves from each virtual source. As an application of using extracted body waves, we estimate 3‐D P wave velocities from these waves with travel time tomography. This study is the first body wave tomography result obtained from only ambient noise recorded at the ground surface. The velocity structure estimated from body waves has higher resolution than estimated from surface waves.

Passive seismic methods using earthquakes can be applied for extracting body waves and obtaining information of subsurface structure. In this study, we retrieve direct and reflected plane waves by applying seismic interferometry to the recorded ground motion from a cluster of regional earthquakes. We apply upgoing/downgoing P/S wavefield decomposition, time windowing, and multidimensional deconvolution to improve the quality of the extraction of reflected waves with seismic interferometry. The wavefield separation and seismic interferometry based on multidimensional deconvolution allow us to reconstruct PP, PS, SP and SS reflected waves without unwanted crosstalk between P and S waves. From earthquake data, we obtain PP, PS and SS reflected plane waves that reflect off the same reflector, and estimate P- and S-wave velocities.

Application of deconvolution interferometry to earthquake data recorded inside a building is a powerful technique for monitoring parameters of the building, such as velocities of traveling waves, frequencies of normal modes, and intrinsic attenuation. In this study, we apply interferometry to ambient‐vibration data, instead of using earthquake data, to monitor a building. The time continuity of ambient vibrations is useful for temporal monitoring. We show that, because multiple sources simultaneously excite vibrations inside the building, the deconvolved waveforms obtained from ambient vibrations are nonzero for both positive and negative times, unlike the purely causal waveforms obtained from earthquake data. We develop a string model to qualitatively interpret the deconvolved waveforms. Using the synthetic waveforms, we find the traveling waves obtained from ambient vibrations propagate with the correct velocity of the building, and the amplitude decay of the deconvolved waveforms depends on both intrinsic attenuation and ground coupling. The velocities estimated from ambient vibrations are more stable than those computed from earthquake data. Because the acceleration of the observed earthquake records varies depending on the strength of the earthquakes and the distance from the hypocenter, the velocities estimated from earthquake data vary because of the nonlinear response of the building. From ambient vibrations, we extract the wave velocity due to the linear response of the building.