SUMMARY We determine the 3‐D in situ shear‐wave velocities of shallow‐water marine sediments by extending the method of surface wave tomography to Scholte‐wave records acquired in shallow waters. Scholte waves are excited by air‐gun shots in the water column and recorded at the seafloor by ocean‐bottom seismometers as well as buried geophones. Our new method comprises three steps: We determine local phase‐slowness values from slowness‐frequency spectra calculated by a local wavefield transformation of common‐receiver gathers. Areal phase‐slowness maps for each frequency used as reference in the following step are obtained by interpolating the values derived from the local spectra. We infer slowness residuals to those reference slowness maps by a tomographic inversion of the phase traveltimes of fundamental Scholte‐wave mode. The phase‐slowness maps together with the residuals at different frequencies define a local dispersion curve at every location of the investigation area. From those dispersion curves we determine a model of the depth‐dependency of shear‐wave velocities for every location. We apply this method to a 1 km2 investigation area in the Baltic Sea (northern Germany). The phase‐slowness maps obtained in step show lateral variation of up to 150 per cent. The shear‐wave velocity models derived in the third step typically have very low values (60–80 m s−1) in the top four meters where fine muddy sands can be observed, and values exceeding 170 m s−1 for the silts and sands below that level. The upper edge of glacial till with shear‐wave velocities of 300–400 m s−1 is situated approximately 20 m below sea bottom. A sensitivity analysis reveals a maximum penetration depth of about 40 m below sea bottom, and that density may be an important parameter, best resolvable with multimode inversion.

Heterogeneous, non-stationary noise sources can cause traveltime errors in noise-based seismic monitoring. The effect worsens with increasing temporal resolution. This may lead to costly false alarms in response to safety concerns and limit our confidence in the results when these systems are used for quasi real-time monitoring of subsurface changes. We therefore develop a new method to quantify and correct these traveltime errors to more accurately monitor subsurface conditions at daily or even hourly timescales. This is based on the inversion of noise correlation asymmetries for the time-dependent distribution of noise sources. The source model is then used to simulate time-dependent ambient noise correlations. The comparison to correlations computed for homogeneous noise sources yields 1 Downloaded 04/23/17 to 132.239.1.231. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/ traveltime errors that translate into spurious changes of the subsurface. The application of our method to data acquired at Statoil's SWIM array, a permanent seismic installation at the Oseberg field, demonstrates that fluctuations in the noise source distribution may induce apparent velocity changes of 0.25 % within one day. Such biases thereby likely mask realistic subsurface variations expected on these timescales. These errors are systematic, dependent primarily on the noise source location and strength, and not on inter-station distance. Our method can then be used to correct for source-induced traveltime errors by subtracting these quantified biases in either data or model space. It can furthermore establish a minimum threshold for which we may reliably attribute traveltime changes to actual subsurface changes, should we not correct for these errors. In addition to the aforementioned real data scenario, we apply our method to a synthetic case for a daily passive monitoring overburden feasibility study. Both synthetics and field experiments validate the method's theory and application.2 Downloaded 04/23/17 to 132.239.1.231. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/ source distribution. Finally, instead of blindly performing waveform inversion with these biased traveltimes, we formulate how to remove these errors so that our velocities are unaffected by noise source non-stationarity. Preferably, this is done in the data space before proceeding to the tomography. We now discuss these steps in further detail.The SWIM array, shown in Figure 1, consists of 172 four-component receivers (MEMS accelerometer and hydrophone) linked through a single ocean-bottom cable. The configuration lies along the ocean floor, roughly 108 m below mean sea level, about 150 m south of the Oseberg C platform. Its purpose is to monitor cuttings injected into the overburden.Receiver spacing is ∼ 50 m for receivers on the outside boundaries of the array and reduces to ∼ 25 m for receivers in its inner portion. The spread of the array's outer right-side is around 1.74...

S U M M A R YThe estimation of the Green's function between two points on the Earth's surface by the crosscorrelation of seismic noise time-series became a widely used method in seismology. In general, very long time-series (months to years) as well as massive normalization and/or data selection are necessary to obtain useful cross-correlation functions. One task of this study is to evaluate the influence of different established normalization methods on the obtained cross-correlation functions. Furthermore, we evaluate two waveform preserving time domain normalizations as well as a new fully automated data selection approach. The cross-correlation functions analysed in this study are obtained from 12 months of seismic noise recorded in 2004 at five seismic stations in the United States with station distances on a continental scale. For practical reasons, the cross-correlation functions of such long time-series are calculated by stacking the cross-correlation functions obtained from shorter time windows. We use this stacking process for the implementation of the waveform preserving time domain normalizations. The time window length is in general an important parameter of the cross-correlation processing, as it influences the normalization and data selection. Therefore, we evaluate the cross-correlation functions obtained with 47 different time window lengths between one hr and 24 hr. The time domain normalizations intend to suppress the influence of transient signals like earthquake waves as well as long-term (e.g. seasonal) amplitude variations. We compare the proposed waveform preserving time domain normalizations with the established running absolute mean normalization and the one-bit normalization. We demonstrate that a waveform preserving time domain normalization can replace a non-linear time domain normalization, if a time window length similar to the duration of the typically occurring transient signals is used. Next to the time domain normalizations also the spectral whitening in the frequency domain is evaluated. Spectral whitening is a powerful normalization to improve the emergence of broadband signals in seismic noise cross-correlations. Nevertheless, we observe spectral whitening to depend strongly on the time window length. An unwanted amplification of a persistent microseism signal is observed on the continental scale with time windows shorter than 12 hr. Our approach of automated data selection is based on a statistical time-series classification and reliably excludes time windows with transient signals occurring contemporaneously at both sites (e.g. earthquake waves). This data selection approach is capable to replace a nonlinear time domain normalization, but no improvement of the waveform symmetry or the signal-to-noise ratio of the cross-correlation functions is observed in general.

Models of in situ shear-wave velocities of shallow-water marine sediments are of importance for geotechnical applications, sediment characterization, and seismic exploration studies. Here pseudo-2D shear-wave velocity models are inferred from the lateral variation of Scholte-wave dispersion at five different geological sites in the Baltic Sea (northern Germany). To explore Scholte-wave dispersion and the lateral variability of shear-wave velocities, Scholte waves were excited by air gun shots in the water layer and recorded by stationary ocean-bottom-seismometers or buried geophones. We analyze the recorded seismograms in a common-receiver-gather using offset-windowed, multichannel dispersion analysis.The observed local slowness-frequency spectra for the different study sites vary significantly with respect to excitation amplitudes and phase slownesses of different modes, as well as the excited frequency range. The excitation amplitudes are influenced by the local shear-wave velocity structure, absorption, length of Scholte-wave travel path, and the elevation of the source above the sea floor.The inverted shear-wave velocities range from 50 m/s to 600 m/s. Directly at the sea bottom, shear-wave velocities of 50 m/s for fine muddy sand and 300 m/s for glacial till were inferred. The maximum vertical gradient was 680% (mean 250 m/s) within a depth range of 40 m, and horizontally 633% (mean 350 m/s) within 300 m distance. The layer boundaries in the inverted shear-wave velocity models are in good agreement with high-frequency, zero-offset compressional-wave reflections. However, it was not possible to acquire the fundamental Scholte mode above very soft, unconsolidated sediment with shear-wave velocities smaller 50 m/s. The analysis of synthetic data shows that this is due to the elevation of the source and the receiver response function.

By demonstrating offshore ambient-noise surface-wave tomography (ANSWT) at reservoir scale, we add a method to the commercially usable geophysical methods. Analysis of ambient-noise records at 126 locations above a hydrocarbon reservoir offshore Norway proves that the marine environment provides good conditions for 3D estimation of shear-wave velocities at frequencies above 0.1 Hz. The presented results are used to discuss potential application areas.