A sticky situation: Melt transport via magma mush unlocking in analog experiments

How can magma be mobilized from a rigid state? Various geochemical and geophysical datasets suggest that crustal magmas spend most of their time close to the solidus as magma mushes, in which the crystal framework makes a magma mush rigid and potentially immobile in the absence of external disturbances. The unlocking of magma mush due to new magma injections is a ubiquitous process leading up to eruptions but is poorly understood on a micro-mechanics scale. We conduct scaled analog experiments (in a 14 × 6 × 20 cm tank) in a low Reynolds number regime. We utilize corn syrup to represent melt, 4 mm spheroidal plastic beads to represent crystals, and combine the two (∼60 vol% beads) to represent mush. Density differences of ∼100 kg/m3 drive the interaction between two initially distinct magma mush and melt-dominated analog layers. We use time-lapse photography and ML image segmentation to observe changes in the magma mush and melt-dominated layers over time. The maximum velocity of mush disaggregation is well-represented by an analytical model based on individual particle displacement at the liquid interface. However, this process of mush disaggregation is highly heterogeneous. The complex morphology of mush disaggregation is markedly different from a uniformly percolating porous media flow front. Our results have significant implications for interpretations of single-crystal geochemical records, such as trace element zonation patterns in zircons. Our results also highlight the key microphysical processes for melt transport through magma mushes and help to illuminate the complex nature of transcrustal magmatic systems.