In sedimentary environments, horizontal fine layering can cause significant complications when analyzing and comparing seismic data with different frequencies and propagation directions. At the Whitchester test site, three boreholes penetrate upper Carboniferous cyclical sediments, involving interbedded carbonates, sandstones, and mudstones, with seismic velocities ranging from less than 3.0 km/s in the mudstones to over 4.5 km/s in the carbonates. Initial crosshole results from two boreholes approximately 200 m deep and 75 m apart showed a poor correlation with the borehole logs recorded in a third, intermediate borehole. Furthermore, the tomographic images were inconsistent with the expected geology. The objective of this paper is to reconcile the crosshole seismic data with the borehole log information. The borehole information must be up-scaled to match the resolution of the crosshole experiment. However, upscaling techniques based on Backus averaging lead to the prediction of significant anisotropy (of the order of 20% in places) with a vertical symmetry axis. This anisotropy is a result of a combination of fine layering and the intrinsic mineral anisotropy measured on the core samples, although it is the layer-induced anisotropy that dominates at the crosshole frequencies of 400 Hz. This prediction was tested by including anisotropy parameters into the analysis of the crosshole data, using anisotropic velocity tomography. In the central part of the final anisotropic velocity tomograms, where the ray coverage is adequate, these crosshole results are consistent with the anisotropy predictions. Once the anisotropy was properly accounted for, the tomographic images are consistent with the layered nature of the geology at the site, and they show the location and throw of a known fault in the section. We conclude that, at this site and at crosshole frequencies, seismic layer-induced anisotropy plays a significant role and must be accounted for when processing and interpreting crosshole seismic data.
