Melt distribution along the axis of ultraslow spreading mid-ocean ridges

Ultraslow spreading mid-ocean ridges (<15 mm/y full spreading rate) differ from faster spreading ridges by their uneven melt distribution. Crustal thickness varies along axis from zero to more than 8 km at volcanic centers. These volcanic centers receive more melt than the regional average and may remain spatially located for millions of years. The segmentation pattern and active volcanism at ultraslow spreading ridges therefore differs from faster spreading ridges where elongate axial volcanic ridges typically erupt magma. \nUsing networks of ocean bottom seismometers with an along-axis extent of about 60 km at three differing ridge segments, we could show that the maximum depth of brittle faulting, equivalent approximately to temperatures of 600-700°C, varies drastically along axis. Ridge sections that lack an igneous crust exhibit a thick lithosphere as evidenced by the deepest mid-ocean ridge earthquakes observed so far at more than 30 km depth. Beneath areas of basalt exposure, in particular beneath pronounced volcanic centers, the axial lithosphere may be more than 15 km thinner allowing for melt flow at the base of the lithosphere towards the volcanoes, a process that has been postulated to explain the anomalous melt distribution at the slowest spreading ridges. \nSpreading events at ultraslow spreading ridges are unusual as we found from two spreading episodes at 85°E Gakkel Ridge and Segment 8 volcano on the Southwest Indian Ridge. These eruptions were preceded or accompanied by large (M>5) and long-lasting earthquakes swarms and active magmatism lasted over 3-16 years. A massive hydrothermal event plume and sounds from deep submarine explosive volcanism were observed at Gakkel Ridge. At the Segment 8 volcano, we imaged a melt reservoir extending to about 8 km depth below the volcano that potentially fed a sill intrusion recorded by an ocean bottom seismometers about 30 km away at a neighboring subordinate volcanic center. \nTo better understand the segmentation and melt transport at ultraslow spreading rigdes, we recently conducted a segment-scale seismicity survey of Knipovich Ridge in the Norwegian-Greenland Sea, instrumenting the ridge along 160 km of its axis with 28 ocean bottom seismometers for a period of a year, the currently largest mid-ocean ridge microseismicity experiment.