Episodic magma flow in narrow channels through the lower crust in Iceland: geodetic evidence

Magma flow through the ~10-35 km thick lower crust in Iceland is highly episodic. Half a century of crustal deformation studies in Iceland has provided indirect information on a number of magma channels in the lower crust. Geodetic modelling of volume change associated with magma accumulation in the upper crust, and extrusion rates for eruptions directly fed from the lower crust or the mantle, reveal magma flow through lower crust at rates of 0.5-50 m³/s or less (except at the Hekla and Bárðarbunga volcanoes where it is has been higher in recent eruptions). Considering the viscosity of basaltic magma flowing in such channels (~ 10-100 Pa s), driving pressure related to density difference between magma and host rock along the channel length (of about 300 kg/m³) and buoyancy in an underlying magma lens counterbalanced by viscous drag, suggests the cross-sectional area of these channels are <30 m², corresponding to cross-sectional area of a circle with radius of about 3 m or less. Such channels have provided: (i) continuous magma flow over years to a decade to a level of about 3-6 km depth in magma domains (near the boundary between the lower and upper crust, or close to the brittle-ductile boundary), feeding repeated eruptions such as in the Svartsengi volcanic system on the Reykjanes Peninsula since 2023, in the 1975-1984 Krafla rifting episode, and caused inflation of Askja volcano since 2021 without an eruption, (ii) episodic flow directly feeding eruptions without significant accumulation of magma in the shallow crust such as at Fagradalsfjall 2021, 2022 and 2023 eruptions on the Reykjanes Peninsula and the 2010 flank eruption at Eyjafjallajökull on Fimmvörðuháls, and (iii) episodic flow to intrusions or magma domains at Eyjafjallajökull volcano in 1994, 1999 and 2009, Öræfajökull volcano in 2017-2019, and into the Upptyppingar lower crustal intrusion in 2007-2008.

The cross-sectional area of lower crustal magma channels is very minor compared to that of dikes that have formed in the upper crust in the last 50 years in Iceland, that according to geodetic modelling have allowed magma flow rates of up to >5000 m³/s. We suggest that the lack of stored tectonic stress in the ductile lower crust limits the possibilities of formation of extensive dikes there.