Massive carbon storage in the upper mantle via diapirism of subducted continental sediments

Slab material transfer processes at convergent margins are crucial for the long-term geochemical evolution of Earth's interior and exterior. In contrast to well-studied oceanic subduction, the fate of recycled crustal components in the mantle during continental subduction remains poorly constrained. As Earth's largest active orogen, the Himalaya–Tibet Orogen is an ideal site for investigating these processes. Cenozoic ultrapotassic-potassic rocks are widespread in the Himalaya–Tibet Orogen and can provide unique insights into crust–mantle interaction in a well-documented continental collision. Although there is an increasingly clear consensus on the relevance of Indian continental subduction to the formation of the post-collisional K-rich rocks in the Tibetan Plateau, considerable uncertainties remain about the sources and processes that produced these rocks. In this review, we integrate previously reported data with new Sr-Nd-Pb-Hf-Mg-Zn isotopic and mineral and whole-rock chemical data of post-collisional K-rich rocks in the Tibetan Plateau to demonstrate that these rocks originate from a mantle source containing recycled continental material; we further identify a change from carbonate-dominated to silicate-dominated components as Indian continental lithosphere subducted northward. This magma geochemistry not only supports a mélange melting model, but also indicates that sediments detached from the downgoing Indian continental lithosphere and diapirically migrated into the overlying lithospheric mantle. In the context of temporal-spatial evolution of regional magmatism, as well as geochemical and geophysical data, we further emphasize the importance of mélange diapirism as a slab-to-mantle material transport process, and illustrate how subduction of continental lithosphere likely resulted in massive carbon storage in the upper mantle of the Tethyan Realm.