Open AccessObservations of earthquake migration along active fault zones [Richter, 1958; Mogi, 1968] and related theoretical concepts [Elsasser, 1969] have laid the foundation for studying the problem of slow deformation waves in the lithosphere. Despite the fact that this problem has been under study for several decades and discussed in numerous publications, convincing evidence for the existence of deformation waves is still lacking. One of the causes is that comprehensive field studies to register such waves by special tools and equipment, which require sufficient organizational and technical resources, have not been conducted yet.
The authors attempted at finding a solution to this problem by physical simulation of a major shear zone in an elastic-viscous-plastic model of the lithosphere. The experiment setup is shown in Figure 1 (A). The model material and boundary conditions were specified in accordance with the similarity criteria (described in detail in [Sherman, 1984; Sherman et al., 1991; Bornyakov et al., 2014]). The montmorillonite clay-and-water paste was placed evenly on two stamps of the installation and subject to deformation as the active stamp (1) moved relative to the passive stamp (2) at a constant speed. The upper model surface was covered with fine sand in order to get high-contrast photos. Photos of an emerging shear zone were taken every second by a Basler acA2000-50gm digital camera. Figure 1 (B) shows an optical image of a fragment of the shear zone. The photos were processed by the digital image correlation method described in [Sutton et al., 2009]. This method estimates the distribution of components of displacement vectors and strain tensors on the model surface and their evolution over time [Panteleev et al., 2014, 2015].
Strain fields and displacements recorded in the optical images of the model surface were estimated in a rectangular box (220.00×72.17 mm) shown by a dot-and-dash line in Fig. 1, A. To ensure a sufficient level of detail in the analyses of the strain fields in each optical image, the selected area was covered with a uniform mesh (3.43×3.43 mm). In the zoomed-up images, the mesh was 32×32 pixels (a pixel of 0.107×0.107 mm). For each pair of optical images, we calculated cross-correlation functions of the intensity of pixels between pairs of the same size cells (Fig. 2). Directions and magnitudes of displacements of the cells were determined from displaced maximums of cross-correlation functions