Dynamic Two-Phase Flow Modeling of Melt Segregation in Continental Crust: Batholith Emplacement Versus Crustal Convection, With Implications for Magmatism in Thickened Plateaus

Magmatic processes in the continental crust such as crustal convection, melt ascent, magma emplacement, and batholith formation are not well understood. We solve the conservation equations for mass, momentum, and energy for two-phase flow of melt and solid in 2D, for a thick continental crust heated from below by one or several heat pulses. A simplified binary melting model is incorporated. We systematically vary (a) the retention number, characterizing melt mobility, (b) the intensity of heat pulses applied at the bottom, and (c) the density of the solidified evolved rock. Two characteristic modes are identified: (a) in the “batholith emplacement mode,” segregation is sufficiently strong allowing melts to separate from the convective flow. This melt freezes to form buoyant SiO₂-rich layers. (b) In the “convective recycling mode,” melts are formed in the lower crust, rise together with the hot rock with little segregation, freeze at shallow depth but are partly recycled back to the lower crust where they remelt. Phase-change-driven convection dominates. Mode (a) is favored by high heat input, multiple heat pulses, high melt mobility, and low density of the evolved rock. Mode (b) is favored by less intense heating, less melt mobility, and denser evolved rocks. A scaling law is derived based on the thermal, melt, and compositional Rayleigh numbers and the retention number. The Altiplano-Puna low-velocity zone (LVZ) could represent the batholith emplacement mode with buoyant and voluminous magmas causing intense volcanism. The Tibetan LVZ is not associated with intense volcanism and might represent the convective recycling mode.