No abstract

SUMMAR YTo study crustal rock seismic anisotropy and its effect on seismic wave propagation, we measure the seismic velocity anisotropy of two amphibolites, one biotite gneiss and one biotite schist from the Hidaka metamorphic belt in central Hokkaido, Japan, under confining pressures up to 150 MPa. The rock microstructures show foliation and lineation characterized by lattice preferred orientation (LPO) of hornblende or biotite. P-and two S-wave velocities are measured along the direction perpendicular to the foliation plane and two directions in the foliation plane: perpendicular and parallel to the lineation. We assume orthorhombic symmetry based on the rock microstructures and obtain Tsvankin's anisotropic parameters (an extension of Thomsen's parameters for orthorhombic symmetry). P-and S-wave phase velocity surfaces are calculated from anisotropy parameters and compared with the measured velocities along particular directions and with the velocity contour maps calculated from the Voigt averages of singlecrystal elastic constants based on the orientation of measured LPO data. Qualitatively, the measured velocity anisotropy agrees with the velocity contour calculated from the LPO data, although large quantitative differences exist between them. All anisotropy patterns can be approximated as transverse isotropy or its modification, appearing as orthorhombic symmetry. Biotite schist (containing 30 per cent volume ratio biotite) shows strong S-wave anisotropy, and the phase velocity surfaces of P waves show a large deviation from ellipticity in the plane perpendicular to the foliation and parallel to the lineation. In the same plane, S waves show a singularity due to a large bulge of the SV velocity surface.

Nonhyperbolic moveout analysis plays an increasingly important role in velocity model building because it provides valuable information for anisotropic parameter estimation. However, lateral heterogeneity associated with stratigraphic lenses such as channels and reefs can significantly distort the moveout parameters, even when the structure is relatively simple. We analyze the influence of a low-velocity isotropic lens on nonhyperbolic moveout inversion for horizontally layered VTI (transversely isotropic with a vertical symmetry axis) models. Synthetic tests demonstrate that a lens can cause substantial, laterally varying errors in the normal-moveout velocity V nmo ð Þ and the anellipticity parameter g. The area influenced by the lens can be identified using the residual moveout after the nonhyperbolic moveout correction as well as the dependence of errors in V nmo and g on spreadlength. To remove such errors in V nmo and g, we propose a correction algorithm designed for a lens embedded in a horizontally layered overburden. This algorithm involves estimation of the incidence angle of the ray passing through the lens for each recorded trace. With the assumption that lens-related perturbation of the raypath is negligible, the lens-induced traveltime shifts are computed from the corresponding zero-offset time distortion (i.e., from "pull-up" or "push-down" anomalies). Synthetic tests demonstrate that this algorithm substantially reduces the errors in the effective and interval parameters V nmo and g. The corrected traces and reconstructed "background" values of V nmo and g are suitable for anisotropic time imaging and producing a high-quality stack.

The development of reliable systems for monitoring injected CO2 is essential in carbon capture and storage projects. We applied time‐lapse surface wave analysis to measure temporal variations of the shallow subsurface among 11 periods (0.2–21.6 days) of continuous seismic data acquired from 2014 to 2016 at the Aquistore CO2 storage site in Canada. We focused on monitoring environmental influences on shallow seismic velocity, which are unrelated to CO2 injection into deep reservoirs. A continuous, controlled seismic source system called Accurately Controlled Routinely Operated Signal System was used to enhance the temporal resolution and source repeatability. Observed phase velocities were clearly higher in winter than in warmer seasons. The seasonal variation could be reproduced by an increase in the shallow S wave velocity during winter associated with the greater extent of freezing of partially saturated rock. We also observed gradual increases or decreases in phase velocities as the seasons changed, which could be related to gradual freezing or melting of ice. Our monitoring system thus could be effective for monitoring temporal variations of the shallow subsurface associated with the degree of freezing. Furthermore, the high temporal stability of our monitoring approach in warm seasons may make it possible to immediately identify CO2 leakage in the shallow subsurface.

Moveout analysis of wide-azimuth reflection data is usually performed for smoothly varying velocity fields. However, velocity lenses in the overburden can cause significant laterally-varying errors in the moveout parameters and distortions in data interpretation. The main goal of this paper is to develop a correction for the influence of an elongated velocity lens on the NMO ellipse and nonhyperbolic reflection moveout. We show that an analytic expression for the NMO ellipse in stratified media with lateral velocity variation, introduced in our previous paper, adequately represents lens-induced distortions. The correction for lateral heterogeneity is controlled by the second derivatives of the interval vertical traveltime. This analytic correction provides a quick estimate of the contribution of velocity lenses and substantially mitigates lens-induced distortions in the effective and interval NMO ellipses. To remove the influence of velocity lenses on nonhyperbolic moveout inversion of wide-azimuth data, we propose a prestack correction algorithm that involves computation of the lensinduced traveltime distortion for each recorded trace. The method is designed for a horizontally-layered overburden but can handle laterally heterogeneous target layers with dipping or curved interfaces. Synthetic tests confirm that this algorithm accurately removes lens-related traveltime shifts and significantly reduces errors in the azimuthally varying moveout parameters. The developed methods should increase the robustness of seismic processing of wide-azimuth surveys, especially those acquired for fracture-characterization purposes.

No abstract

Monitoring time‐lapse changes in subsurface physical properties of the near‐surface critical zone is increasingly important with respect to climate change, environmental conservation/remediation, geohazard mitigation, and geotechnical engineering activities. Innovative controlled‐source cross‐well seismic monitoring surveys combined with full waveform inversion analysis enable us to map small and highly localized changes by repeatedly scanning the subsurface between borehole sensors at depth. In the Kanto Basin, Japan, we successfully monitor the dynamic transient fluid‐flow effects of the subsurface injection of microbubble water, which is of interest for soil contamination remediation and preventing earthquake liquefaction. The fluid migration is detected by observing P‐wave velocity changes (∼1%) within a very thin (∼1 m) sediment layer at a depth of ∼25 m. The injected microbubble water of the differential physical properties (temperature) is observed to follow geological and hydrologic preferential fluid‐flow paths rather than diffusing equally in all directions away from the injection well.

Onshore hydrocarbon exploration in the back-arc region of Japan suffers from seismic imaging challenges. Seismic data are typically acquired in a difficult environment: along a crooked line over a rough surface topography underlain by severe near-surface weathering layers. The area is geologically complex, and prestack depth processing is desirable. However, the data quality is suboptimal at near offsets, and it hinders the migration velocity analysis required for depth processing. Thus, a geologic interpretation needs to rely on time-processing results. Nevertheless, some data sets are rich in low-frequency components and contain clear refracted and wide-angle reflected waves, both of which are favorable conditions for the application of waveform tomography. We have used one such data set, and we determined the applicability of waveform tomography to estimate the P-wave velocity distribution. To mitigate difficulties in the data, we carefully optimized waveform tomography strategies and diligently validated the results. We evaluated the effects of the crooked line by conducting waveform tomography in 2D and 2.5D. We determined that for this data set, the effect of the crooked line was not substantial enough to require the 2.5D implementation, but it was large enough to require careful trace editing. We also minimized the effects of large variations in amplitudes due to near-surface effects and source and receiver characteristics, primarily by fitting phase components of the data during waveform tomography, but also by adjusting the amplitudes in a surface-consistent manner to stabilize the source estimation process. The final velocity model delineated known shallow sedimentary structures, but it also revealed deep largethrust structures critical to the structural understanding of the area. The result confirmed the capability of waveform tomography to yield a reliable velocity model, and it also opened up the possibility of applications of depth processing in the region.