Rate of active shortening across the southern thrust front of the Greater Caucasus in western Georgia from kinematic modeling of folded river terraces above a listric thrust

Abstract

Although the Greater Caucasus (GC) Mountains accommodate a significant fraction of orogen-perpendicular Arabia-Eurasia convergence at their longitude, the locations and slip rates of the active structures absorbing this shortening are poorly known. Here we report the first late Quaternary shortening rate for an active thrust in the GC determined from dating and reconstructing a deformed landform. In the Rioni basin along the southwestern flank of the GC, terraces along the Enguri River have been folded by the south-vergent Tsaishi anticline and record Quaternary shortening along the southernmost exposed frontal structure. Terrace surfaces are well preserved and together with underlying Mesozoic-to-Neogene strata clearly record folding. We use high-resolution topographic data from differential GPS surveys and bedding attitudes across the fold to construct a kinematic model that describes both the terrace deformation and finite shortening across the anticline recorded by the folded strata. We calculate that a deformed terrace records 106.1 +38.7/-36.4 m of slip since abandonment. Optically stimulated luminescence dating of quartz in a loess cap on this terrace indicates it was abandoned before 94.1 ± 12.8 ka, implying a geologic shortening rate of 1.09 +0.62/-0.43 mm/yr since ∼100 ka. This work confirms that the frontal structure of the Rioni fold-thrust belt is Quaternary-active, but has a rate that falls short of the regional geodetic shortening rate of 3-5 mm/yr. We infer that the missing late Quaternary shortening is accommodated by additional structures within the GC to the north. Folding of late Quaternary terraces and a possible coseismic scarp along the frontal structure of the GC together imply that structures in the western GC may be capable of coseismic rupture within and beneath the adjacent basin, in addition to blind ruptures within the main range such as the 1991 Racha earthquake. Thus, the GC system appears to be broadly analogous to the Himalaya, where the foreland fold-thrust belt connects by a shallow seismogenic detachment to structures beneath the main range. The geometry of this detachment may influence earthquake rupture initiation and propagation, potentially playing a significant role in the seismic hazard of the GC.