Variations in earthquake rupture properties along the Gofar transform fault, East Pacific Rise

On a global scale, seismicity on oceanic transform faults that link mid-ocean ridge segments is thermally controlled(1,2). However, temperature cannot be the only control because the largest earthquakes on oceanic transform faults rupture only a small fraction of the area that thermal models predict to be capable of rupture(3-5). Instead, most slip occurs without producing large earthquakes(3,4,6). When large earthquakes do occur, they often repeat quasiperiodically(7,8). Moreover, oceanic transform faults produce an order of magnitude more foreshocks than continental strike-slip faults(7,9). Here we analyse a swarm of about 20,000 foreshocks, recorded on an array of ocean-bottom seismometers, which occurred before a magnitude 6.0 earthquake on the Gofar transform fault, East Pacific Rise. We find that the week-long foreshock sequence was confined to a 10-km-long region that subsequently acted as a barrier to rupture during the mainshock. The foreshock zone is associated with a high porosity and undergoes a 3% decrease in average shear-wave speed during the week preceding the mainshock. We conclude that the material properties of fault segments capable of rupturing in large earthquakes differ from those of barrier regions, possibly as a result of enhanced fluid circulation within the latter. We suggest that along-strike variations in fault zone material properties can help explain the abundance of foreshocks and the relative lack of large earthquakes that occur on mid-ocean ridge transform faults.