We describe the most comprehensive and detailed high resolution survey of an axial volcanic ridge (AVR) ever conducted, at 45°N on the Mid-Atlantic Ridge. We use 3 m resolution sidescan sonar, deep-towed magnetic field measurements, video observations from eleven ROV dives, and two very-high-resolution bathymetry and magnetic surveys. The most recently active AVR has high topographic relief, high acoustic backscatter, high crustal magnetization and little faulting. It is sharp-crested, 25 × 4 km in extent and 500 m high, and is covered by approximately 8000 volcanic “hummocks” whose detailed nature is revealed for the first time. Each is an individual volcano ≤ 450 m in diameter and ≤ 200 m high, ranging from steep-sided (45°) cones to low domes. Many have suffered significant flank collapse. Hummocks tend to align in rows parallel to the AVR axis, parallel to its NE-trending spurs or, on its lower flanks, sub-normal to the AVR trend. These latter are spaced 1–2 km apart and comprise 1–2 km-long rows of single volcanoes. We infer that their emplacement is controlled by down-flank magma transport, possibly via lava tubes. The AVR contains only one large flat-topped seamount. The flanking median valley floor consists of either older hummocky volcanic terrain or flat-lying, mostly sediment-covered lavas. These typically have low-relief lobate surfaces, inflation and collapse structures, and occasional lava tubes and tumuli. The AVR displays open fissures, mostly along its crest. There is direct evidence for only a few small faults on the AVR, though steep, outward-facing slopes draped by elongate pillows may be small normal faults covered by lava. The surrounding median valley floor is heavily fissured. Normal faults cut it and an older AVR, the latter displaying significant outward facing faults. High crustal magnetization, an approximate proxy for crustal age within the Brunhes, is confined to the active AVR. Magnetic palaeointensity measurements are consistent with ages up to ~ 12 ka for several samples from the active AVR and ≥ 12 ka for one sample from the median valley floor. This is much less than the predicted spreading age, implying distribution of melt off-axis or episodic AVR growth.
Read full abstract