Origin of mafic magmas beneath northwestern Tibet: Constraints from 230Th-238U disequilibria

²³⁸U-²³⁰Th disequilibria and Sr, Nd, and Pb isotopic and chemical data for young (<120 ka) trachyandesites from the Ashikule Basin (AKB) in northwestern Tibet provide constraints on the origin of magmas produced within this region of continental collision. Compared to lavas from both continental and oceanic settings, the AKB samples show large excesses of ²³⁰Th with respect to ²³⁸U (up to (²³⁰Th)/(²³⁸U) = 1.36). Partial melting of garnet-bearing lithologies (garnet peridotite, garnet pyroxenite, or eclogite) could be responsible for these ²³⁰Th excesses and could plausibly occur either in the lithospheric mantle or in the lower crust. Small porosities (<0.4%) and slow melting rates (<10^-4 kg m^-3 yr^-1) are required in the case of a garnet peridotite residue, although larger porosities (up to 18%) and melting rates (>10^-3 kg m^-3 yr^-1) are permitted in the case of an eclogitic residue; hydrous metasomatic phases, if present, would lower these limits further. The source of the AKB magmas has probably been enriched in incompatible elements relative to bulk Earth since at least the middle Proterozoic (>1 Ga), likely via metasomatism by a relatively dry silicate melt, and, if the ultimate source of the lavas is mantle lithosphere, concentrations of incompatible trace elements in the source could be similar to those of xenoliths and massifs. The ²³⁰Th-²³⁸U disequilibria provide additional information to evaluate the methods previously proposed to explain melt generation beneath northern Tibet. The measured ²³⁰Th enrichments are uncharacteristic of melts generated by subduction but could potentially be produced during shear heating of the uppermost lithospheric mantle, by convective removal of the lower lithosphere and heating of the remaining lithospheric mantle, or by decompression during extension across a releasing bend of a strike-slip fault. The diversity of mechanisms that could be responsible for these relatively small-degree melts suggests that the cause of melt production may have varied over time and/or over space in northwestern Tibet and therefore that timing of volcanism may not be directly related to any single tectonic event.