A seafloor spreading event captured by in-situ seismo-geodesy

Over geological times, the growth of the ocean floor involves magmatic and tectonic extension at mid-ocean ridges. Because seismo-geodetic monitoring of these submarine plate boundaries remains challenging, little is known on how these systems operate on yearly timescales. Here we report the first in-situ observation of a rifting event at a mid-ocean ridge segment, that combines hydroacoustic, direct-path ranging and bottom pressure measurements, with repeated seafloor mapping.

The event started on April 26, 2024 at the axis of the Southeast Indian Ridge near 37˚S, two months after instruments had been deployed across the ridge axis and nearby Amsterdam transform fault. The event began as a rapidly migrating swarm of extensional seismicity along the axial valley. It caused 4 m of subsidence of the valley floor, and over a meter of horizontal extension across the valley. We interpret this as the deflation of a magma reservoir feeding propagating dykes and inducing aseismic slip on a valley-bounding fault. The dyke eventually led to the outpouring of ~150 million m³ of lava at the seafloor, while triggering seismic activity on the abutting transform faults.

Using 2-D elastic dislocation models, we randomly sampled 10 million combinations of sill, dyke and fault geometries to assess how well they could account for the observed displacements. Out of these, about 2200 yielded a satisfactory root mean squared (RMS) misfit (< 20 cm), which have all in common: (i) a sill at least 3500 m deep compacting by 10-20 m; (ii) a dyke rooted at the sill and extending to sub-seafloor depths of tens of meters with a metric opening; (iii) a metric slip on an axial-valley bounding fault down to a few km; and (iv) a total horizontal extension of 2 to 4 m, distributed between the dyke and the fault. Most models favour the dyke taking up more extension than the fault. The measured and modelled horizontal displacements are equivalent to 31 to 63 years of spreading at the average rate of 6.3 cm/yr inferred from space geodesy. They are considerably larger than the centimetric offsets caused by the swarm of Mw≈5 earthquakes and must therefore have accrued aseismically during the early stages of the spreading event.

This unique set of observations provides a detailed chronology of a seafloor spreading event, and, along with modelling, suggests that aseismic slip plays a major role during such events, thereby explaining the well-documented earthquake deficit on normal faults at mid-ocean ridges.