321 ISSN 1069-3513, Izvestiya, Physics of the Solid Earth, 2017, Vol. 53, No. 3, pp. 321–331. © Pleiades Publishing, Ltd., 2017. Original Russian Text © A.A. Soloviev, A.I. Gorshkov, 2017, published in Fizika Zemli, 2017, No. 3, pp. 3–13. Modeling the Dynamics of the Block Structure and Seismicity of the Caucasus A. A. Soloviev a, b, * and A. I. Gorshkov a, b a Institute of Earthquake Prediction Theory and Mathematical Geophysics, Russian Academy of Sciences, Moscow, 117997 Russia b Geophysical Center, Russian Academy of Sciences, Moscow, 119296 Russia *e-mail: soloviev@mitp.ru Received October 13, 2016 Abstract⎯Based on the morphostructural zoning scheme of the Caucasus, the block structure ref lecting the real fault geometry and the block formation of the region is constructed. Several dozens of numerical exper- iments are conducted for simulating the dynamics of the block structure and the arising seismicity. The mod- eling relies on the following principles. It is assumed that the structure is composed of perfectly rigid blocks separated by infinitely thin fault planes. On the fault planes and on the blocks' bottoms, the blocks viscoelas- tically interact with each other and with the underlying medium. At each time instant, the translational dis- placements and rotations of the blocks are calculated from the condition of the quasi-static equilibrium of the entire block structure. The earthquakes occur in accordance with the dry friction model at the time instants when within a certain segment of the fault the stress-to-pressure ratio exceeds the given threshold. The mod- eling yields the synthetic catalog of the Caucasian earthquakes the spatial distribution of which reflects a set of characteristic features of the real seismicity. The similarity is observed in the magnitude–frequency dia- grams of the synthetic and real seismicity. The comparison of the positions of the epicenters of the strong syn- thetic earthquakes with the results of recognizing the highly seismically active areas in the Caucasus demon- strates the presence of such epicenters in a few highly active areas where, according to the observations, strong earthquakes have not occurred to date. DOI: 10.1134/S1069351317030120 1. INTRODUCTION Inasmuch as the duration of instrumental seismo- logical observations (about 100 years) is much shorter than the time of action of the tectonic factors causing seismicity, inferring an adequate idea of its time behav- ior from the observational data alone is to a certain extent challenging. One of the probable approaches in this situation consists in using the technique of the numerical modeling of the block structure dynamics and seismicity which yields the synthetic earthquake catalogs covering sufficiently long time intervals. In this paper we present the results of modeling the dynamics of the block structure of the Caucasus. The modeling is based on the morphostructual zoning map of the Caucasus which was constructed for solv- ing the problem of recognizing the strong-earthquake prone areas (Soloviev et al., 2013; 2016). Morpho- structural zoning (Alekseevskaya et al., 1977; Rants- man, 1979; Gvishiani et al., 1988; Gorshkov, 2010) relies on the concept of the lithosphere as a hierarchi- cal system of different-order blocks starting from the natural grains of rocks up to the planetary-scale litho- spheric plates (Sadovskii and Pisarenko, 1991). The morphostructural zoning maps (Weber et al., 1985; Cisternas et al., 1985; Rantsman et al., 1996) were pre- viously used for constructing the block model of the Western Alps (Vorobieva et al., 2000) and the Kachchh rift zone in northwestern India (Vorobieva et al., 2014). One of the key principles underlying the modeling conjectures that the seismically active region consists of a system of blocks separated by thin weakened fault zones which are less consolidated than the blocks, compared to which, based on the characteristic times of the seismotectonic process (a few thousand years), the blocks can be assumed to be perfectly rigid. The earthquakes originate in the fault zones. The technique that we used for modeling in the present paper was previously applied for studying the seismicity and geodynamics of the Vrancea region (Panza et al., 1997; Soloviev et al., 1999; 2000), Mid- dle East (Sobolev et al., 1996), Western Alps (Voro- bieva et al., 2000), Italy and its neighboring territories (Peresan et al., 2007), and the Tibet–Himalaya region (Ismail-Zadeh et al., 2007), as well as the Kachchh rift zone (Vorobieva et al., 2014). In all cases, with the appropriately selected model parameters, the proper- ties of the observed and model seismicity were similar.