Answers are grounded in this paper and cite their sources.
Grabens unload the crust and can steer ascending dikes away from their axis, but what controls whether a dike erupts on the graben floor, at the bounding scarps or beyond them remains debated. I used a two-dimensional, plane-strain model of magma-driven dike propagation, coupling linear elastic fracture mechanics with viscous lubrication flow, to simulate basaltic, basaltic-andesitic and andesitic dikes beneath a 0.5-km-deep graben inspired by the Hat Creek Graben, California. Dikes were injected singly, sequentially or simultaneously in four boundary-condition and reference-state configurations, with and without regional extension. The reference state dominated the results. In a gravity-loaded crust, all dikes erupted on the graben floor. In a lithostatic crust, the graben unloading deflected all dikes outwards, to vents 8–18 km from the axis. Regional extension progressively reduced the deflection, and 10 MPa, comparable to the unloading pressure, returned all vents to the graben floor. For dikes fed at a constant supply rate, higher viscosity, fracture toughness and supply rate reduced the deflection. For dikes fed by a reservoir at fixed pressure, viscosity mainly controlled the propagation time. Buoyancy had little effect. Viscosity therefore does not explain the scarp-aligned basaltic andesites of the Hat Creek Graben, which more likely reflect off-axis sources, a lower deviatoric stress during their eruption, or capture by faults. Neighbouring dikes repelled each other by a few hundred metres in a crust with large deviatoric stress, but merged at depth in a lithostatic crust. Vent location beneath grabens is primarily a record of the regional stress state.
No takes yet. Share an insight, caveat, or question.
Khaled Rahimi (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: