Himalayan sub-Moho earthquakes suggest crustal faults trigger eclogitized-drip tectonics

Along the 2000 km Himalayan arc, more than 100 sub-Moho earthquakes have been detected from their Sn/Lg amplitude ratios or S-minus-P delay times, concentrated most densely beneath a ~ 300 km segment in south Tibet where these earthquakes reach ~ 110 km depth. Possible explanations include Moho-penetrating faults and dripping eclogitized lower crust. We estimate geological strain-rates, temperatures, and timescales for these two processes from seismological, thermal, geological, and geodetic datasets. We use numerical modeling of viscous Rayleigh–Taylor dripping to show that eclogite viscosity must be $$\lesssim$$ 1–5 × 1021 Pa⋅s to allow drip formation within the available geological timescale (5–20 Ma). A deeply penetrating fault cannot by itself explain the 70–110-km seismicity because brittle failure in ultramafic mantle is implausible so far below the 70 km Moho at modeled temperatures and strain rates. Eclogitized lower crust is stronger at upper-mantle depths and enables brittle failure, but an isolated eclogite drip cannot explain the dominant dextral-slip focal mechanisms. We propose that eclogitization of mafic granulites in Indian lower crust occurs along lower-crustal shear zones associated with active faults and fluid intrusion, creating the density anomaly that drives Rayleigh–Taylor instability. As the eclogite drip grows, high strain within the drip creates brittle faulting at upper-mantle depths, albeit in what are crustal lithologies.