A melt channel beneath an oceanic lithosphere originated from mid-ocean ridge

It’s a well-known concept that the upper mantle is divided into lithosphere and asthenosphere by their rheological properties. Many geophysical researches have reported abrupt changes in several physical properties at a boundary called lithosphere-asthenosphere boundary (LAB), which is generally associated with a seismic low-velocity zone (LVZ). What causes the LVZ is a matter of debate, although the LVZ may play an important role in the plate tectonics. Mehouachi and Singh (2018) could provide clues for the nature and the formation of LVZ in their work.

The LVZ has been detected by studies using receiver function (RF), a data processing method to detect a seismic velocity change across a certain boundary. These studies have been mainly based on S-wave velocity variations in the LAB due to the RF’s insensitivity to P-wave. It leads to a debate over the role of melt because the S-wave velocity changes can be accounted for by a model that does not require melt (e.g. grain boundary sliding model). Mehouachi and Singh (2018) avoided this problem by using an ultra-deep seismic reflection survey over an oceanic lithosphere. They surveyed the St Paul Fracture Zone from 40- to 70-Myr-old oceanic lithosphere and acquired a 215-km-long seismic reflection profile. The seismic image showed two reflectors for each lithospheres, which were regarded as the top and the bottom of the LVZ. The distance between the reflectors was measured at ~18 km from the 40-Myr-old lithosphere and ~12 km from the 70-Myr-old lithosphere, respectively, suggesting thinning of the LVZ. They also found an 8.5 ± 4 % decrease in P-wave velocity at the LVZ which was ascribed to the presence of melt.

Furthermore, the authors calculated the amount of melt which might be in the LVZ by measured P-wave velocity drop. Then, they converted the amount of melt into an H2O concentration that could stabilize it. Interesting results were obtained. The H2O concentrations at certain temperatures were high at the LVZ of 40-Myr-old lithosphere compared to that of 70-Myr-old lithosphere, but if the H2O concentrations were integrated from the top to the bottom of the LVZ, the values stayed about the same-13,960±5,700 ppm km where it was 40-Myr-old and 14,440 ± 4,100 ppm km where it was 70-Myr-old. The results may imply that the amount of volatiles trapped in an early stage of formation of the LVZ is maintained without additional loss or gain, and the thinning over time is the main process to enrich the volatile concentrations in it.

Based on these results, they suggested that the volatiles in the LVZ came from corner flow of ridge axes. There have been discussions about the effects of horizontal flux from mid-ocean ridges. Recent studies indicated that 60 % of deep melt could possibly be channeled along LAB. Mehouachi and Singh (2018) divided the integrated H2O concentrations by the depth of the melting regime (~160 km; Keller et al. 2017) and obtained the average H2O concentration of the melt required to make the LVZ. According to the estimation, 90 ± 35 ppm of water was required, which is plausible to explain by 60 % of the volatile concentrations in the horizontal flux.

It’s surprising that the mid-ocean ridge was pointed as a major source area of the LVZ, because MORB magmas are known as volatile-deficient magmas compared to those from other tectonic settings. Moreover, the authors’ suggestions imply the unknown functions of the ridge in plate tectonics. Since melt in the LAB decreases its viscosity, the effects of slab pull and ridge push as power sources of plate movements would be rated higher than conventional thoughts. In addition, because the low pressure conditions like an oceanic LAB can stabilize carbonatite melt, the volatiles in the LVZ can be a source that accounts for carbonatite magmatism.