The Mongol-Okhotsk Ocean closed when the Amuria block, normally considered to have been part of the North China block since the early Mesozoic, and the southern margin of Siberia collided in Late Jurassic to Early Cretaceous times. The resulting suture runs WSW-ENE and is reasonably well defined to the east of longitude 100°E. Because no evidence exists for any westward prolongation of the Mongol-Okhotsk Ocean suture toward the Tarim block, the cryptic termination of the suture is an enigma, compounded by the fact that a tomographically identified slab in the lower 1000 km of the mantle, interpreted as a remnant of Mongol-Okhotsk oceanic lithosphere, has a clear N-S trend, at almost right angles to the surface suture. No sensible explanation can be constructed for a rotation of some 90° of this slab. There is a solution, however, to both these enigmas if we consider that the Triassic Mongol-Okhotsk Ocean existed east of an initially meridian-parallel, but later progressively more sinuous, late Paleozoic Pangea margin. This margin consisted of Siberia, Amuria, and the China continental elements. The Mongol-Okhotsk Ocean was subducting westward during the early Mesozoic and likely older times underneath this margin. This would readily explain the tomographic N-S slab orientation at depths of 2000 km and greater. Paleomagnetic inclination differences between the global apparent polar wander path in Siberian coordinates and results from the North China block show a gradually diminishing trend with time, as these cratons approached each other during the Jurassic. During this time, the paleomagnetic data of the North China block show that it underwent a slight northward motion, but with a considerable counterclockwise rotation of ∼90°. At the same time, the Mongol-Okhotsk Ocean–bordering margin of Eurasia (between Siberia and Tarim) moved southward by ∼30° and rotated 45° clockwise. These continental scissoring movements caused doubly vergent subduction of the Mongol-Okhotsk Ocean. Paleomagnetic data suggest final closure of the Mongol-Okhotsk Ocean in latest Jurassic–earliest Cretaceous time. Arc-related rocks above the subduction zone follow the outline around the core of the Tuva-Mongol belt in the eastern Altaids between Amuria and Siberia, and they form a tightening, westward-convex Tuva-Mongol orocline. This large-scale oroclinal bending of the crust above a disappearing ocean is reminiscent of similarly tightening oroclines in Kazakhstan and Variscan Europe, which closed earlier by subduction in the late Paleozoic.

Published paleomagnetic data from well-dated sedimentary rocks and lavas from the Lhasa terrane have been reevaluated in a statistically consistent framework to assess the latitude history of southern Tibet from ca. 110 Ma to the present. The resulting apparent polar wander path shows that the margin of the Lhasa terrane has remained at lat ~20° ± 4°N from ca. 110 to at least 50 Ma and has drifted northward to its present latitude of 29°N since the early Eocene. This latitude history provides a paleomagnetically determined collision age between the Tibetan Himalaya and the southern margin of Asia that is ca. 49.5 ± 4.5 Ma, if not a few millions of years earlier after considering reasonable estimates for shortening within the suture zone. This collision occurred at lat ~21° ± 4°N, or perhaps ~2° lower if an average-size forearc is considered. These paleomagnetic data indicate that at most, only 1100 ± 560 km of post–50 Ma India-Asia convergence was partitioned into Asian lithosphere. The lower bound of these paleomagnetic estimates is consistent with the magnitude of upper crustal shortening and thickening within Asia calculated from structural geologic studies. Thus, a substantial amount of the shortening within, and therefore surface uplift of, the Tibetan Plateau predates the Tibetan Himalaya–Lhasa collision. These conclusions suggest that the Tibetan Plateau is similar to the Altiplano of the Andes, in that most of the plateau developed at subtropical latitudes above an oceanic sub-duction zone in the absence of a continent-continent collision. A direct implication of these findings is that 1700 ± 560 km or more post–50 Ma India-Asia convergence was partitioned into the lower plate of the orogenic system (i.e., units of Indian affinity). Recent paleomagnetic and plate tectonic analyses suggested significant extension of Greater India lithosphere after breakup from Gondwana but prior to collision with the southern margin of Asia. Cretaceous extension within Greater India was inferred to open an oceanic Greater India Basin, which would have maintained a deep tropical water mass along the southern edge of greater Asia throughout most of the Paleogene. We suggest ways in which future climate models can incorporate this paleogeography to more accurately explore how Paleogene atmospheric processes interact with or are modified by the juxtaposition of a tropical ocean basin and the high uniform topography of the Tibetan Plateau.