Ultraslow Spreading Processes - A Microseismicity Study of the Knipovich Ridge

Abstract

Along the global mid-ocean ridge system, new seafloor is constantly formed as tectonic plates drift apart. When spreading rate is reduced to less than 20 mm/yr the spreading dynamics change drastically and thereby the entire appearance of these ultraslow spreading ridges differs from faster spreading ridges. Melt is unevenly distributed such that volcanic centers receive more melt than the ridge on average does. Amagmatic segments in between are the melt-poor counterpart. The process of melt focusing is suggested to guide melt along the lithosphere – asthenosphere boundary from amagmatic segments towards volcanic centers. Until now, the processes acting at ultraslow spreading ridges are not completely understood. Key questions are the scale of melt focusing and how melt is extracted at the volcanic centers, the role of detachment faults and the extent of rock alteration. With a microseismicity study on the scale of an entire segment, spanning from one volcanic center to another, these questions could be addressed. The unique microseismicity study was conducted at the Knipovich Ridge, that is a very oblique, ultraslow spreading ridge and part of the Arctic Ridge System. The ocean bottom seismometer network of in total 30 stations was deployed for around one year along 160 km of the rift axis. It covered the Logachev volcanic center, which is the major volcanic center of the Knipovich Ridge, and a second volcanic center south of it. For the recorded data I used automatic earthquake detection and picking of P- and S-phases with a subsequent manual pick check. In this way I extracted in total 14401 earthquakes from the recorded data in the study area. The earthquakes in this comprehensive earthquake catalog were located with different algorithms. 8435 earthquakes with a maximum depth error and Smajor of 5 km and a RMS of 0.4 s were classified as reliably located and used for further interpretations. I determined fault plane solutions for 44 events. Furthermore, I used the earthquakes for a local earthquake tomography of the Logachev area and the entire area covered by the network. From this extensive, unique dataset I found a varying segment-scale pattern of seismicity. The maximum depth of seismicity marks an undulating boundary of the mechanical lithosphere. It is shallowest at the Logachev volcanic center and deepens away from it for 70 km. Thus, I find new evidence that melt focusing may act over distances of 70 km to accumulate the above average melt volumes at volcanic centers. It can therefore be considered an essential mechanism shaping ultraslow spreading ridges. The different spreading styles, magmatic and amagmatic, caused by the uneven melt supply, can be distinguished by their characteristic microseismicity patterns. Magmatic sections are, besides shallower maximum depth of faulting, characterized by shallow seismicity, that is absent in suggested amagmatic sections. At these amagmatic sections, I infer from yield strength envelopes rock alteration reaching 9 km below the seafloor. The differing magmatic section hosts a partial melt area with its top at around 10 km below the sea level underneath the Logachev volcanic center. This is indicated by high Vp/Vs-ratios and low S-velocities from the local earthquake tomography. Elevated temperatures around the partial melt area lead to more ductile behavior resulting in a gap in the seismicity. Directly above the partial melt area high earthquake swarm activity maps the ascent path of melt to the surface. Lateral feeding as observed at orthogonal spreading segments is prevented by the obliquity of the Knipovich Ridge. Hence, I conclude that melt is redistributed and extracted depending on the crustal structure of the ridge. The obliquity of the Knipovich Ridge also affects the transform motion that seems to be hosted on small fault planes producing too small seismicity to be recorded by the KNIPAS network. Additionally, seismicity patterns of active detachment faults are not observed. This stands in contrast to the Southwest Indian Ridge, where detachment faults are a common feature. Consequently, I find that detachment faults are not ubiquitous at ultraslow spreading ridges. With the first ocean bottom seismometer survey on ultraslow spreading ridges, that covers entire spreading segments, this study yields an important contribution in understandig spreading processes and dynamics. It gives insights how melt is distributed along the melt poorest endmembers of the mid-ocean ridge system and how the interplay of magmatic and tectonic activity shapes the lithospheric structure.