Late Paleozoic closure of the Ob-Zaisan Ocean along the Irtysh shear zone (NW China): Implications for arc amalgamation and oroclinal bending in the Central Asian orogenic belt

The largest accretionary orogen in the world, the Central Asian orogenic belt, has evolved through the assembly of various oceanic and continental blocks. Understanding the processes associated with the development of this orogenic belt relies on precise recognition of the boundaries between various terranes. One such major suture zone, which records the collisional history of the Siberian marginal arc system (Chinese Altai) with intra-oceanic arc systems (East/ West Junggar), is the Irtysh shear zone in NW China. The spatial continuity and the tectonic nature of this shear zone are still poorly under-stood, but its development has supposedly made a significant impact on the architecture of the western Central Asian orogenic belt and the formation of the Kazakhstan orocline. Here, we provide new insight into the evolution of this shear zone based on detrital zircon ages, Hf isotope composition, structural data and 40Ar/39Ar age constraints on the timing of deformation. Our results show a major discrepancy in detrital zircon populations and Hf model ages across the southern Chinese Altai and the northern East/ West Junggar, thus allowing us to map the exact location of the tectonic boundary. Detrital zircon data constrain the initial closure of the Ob-Zaisan Ocean to the late Carboniferous (<323 Ma), and new structural and 40Ar/39Ar geochronological data shed light on the subsequent collisional processes. We propose that the collisional zone between the Chinese Altai and the East/West Junggar was initially subjected to crustal thickening at ca. 323-295 Ma, followed by orogen-parallel extension (ca. 295 Ma) and transpressional deformation (ca. 286-253 Ma). The closure of the Ob-Zaisan Ocean in NW China postdates the initial phase of oroclinal bending in the western Central Asian orogenic belt, thus indi-cat-ing that oroclinal bending initiated during subduction. Based on our new constraints and other available geological data, we suggest that the early stage of oroclinal bending was likely driven by slab rollback.