The effect of centimeter-scale folding and crenulation on anisotropy homogenization of schists and phyllites from the NW-Tauern Window (Eastern Alps, Austria)

Abstract

The interpretation of seismic data in orogens is usually difficult to decipher as structural information is limited to surface and borehole data. Seismic interpretations very much depend on the elastic wave velocity model, which in the simplest case is a function of rock composition. Seismic velocities can also be anisotropic, i.e. depend on the wave propagation direction inside the rock. Seismic anisotropy can be subdivided into intrinsic (crystallographic preferred orientation (CPO)) and extrinsic (shape preferred orientation, compositional layering or fractures) anisotropy. Microstructures in thin section scale have an impact not only on millimeter-scale but also on larger anisotropies in the field such as meter- to kilometer-scale folds. Here we explore the effect of microstructure (mainly folding and crenulation) on the homogenization of seismic anisotropy from samples of millimeter to thin section scale. The investigated samples are phyllosilicate- and graphite-rich samples (Innsbruck quartzphyllite and Bündner schist) from the N-S running Brenner Base Tunnel Project (NW-Tauern Window). Phyllosilicate-rich sections with layers of different composition and structure were selected from drill core samples of the exploration tunnel. The CPO of phyllosilicates and graphite from 1.5 – 3.5 mm thick cylinders was measured using high energy X-ray diffraction at DESY (Hamburg, Germany) and the ESRF (Grenoble, France). Pole figure data was directly extracted using single peak fitting. The CPO of quartz was determined by using EBSD. Seismic velocities for each sample were computed using µXRF-based modal composition and single crystal stiffness tensors. We measured the smallest representative volume element which we consider to be undisturbed by microstructural effects. Therefore, we estimate an upper bound of expected intrinsic velocity anisotropies. Thin section-scale anisotropies were modeled from the upper bound anisotropy and the observed microstructure, i.e., small-scale folding. Computed velocities were compared to Vp-anisotropy measurements on the drill cores. The velocity anisotropy is primarily governed by the content and distribution of phyllosilicates and graphite. Given the crystal symmetry and the low single crystal elastic anisotropy, phases such as feldspar, quartz or calcite can be considered as irrelevant with respect to seismic anisotropies. The simulation of a crenulation cleavage has a stronger impact than centimeter-size folding: The crenulation cleavage reduces the anisotropy for example from 14 % to 12 %. Centimeter-size folding with observed interlimb angles of 140° in contrast is negligible. The effect of microstructures like centimeter-scale folds and crenulation has only a limited impact on anisotropies of foliated rocks during homogenization from millimeter to thin section-scale. We assume that during homogenization to a larger scale, the effect of folding with small interlimb angles or different fold axes within the homogenized volume will have a stronger influence on seismic anisotropy.