Characteristics of tectonomagmatic earthquake swarms at the Southwest Indian Ridge between 16°E and 25°E

Abstract

SUMMARY The ultraslow spreading Southwest Indian Ridge (SWIR) is a prominent end‐member of the global mid‐ocean ridge system. It spreads with a full‐rate of 14–16 mm y−1 and shows several segments of various obliquities. The western SWIR consists of the Oblique and Orthogonal Supersegments lying at an epicentral distance of ∼21° to the VNA2 seismic array operated by the German Neumayer station in East Antarctica. The array monitors backazimuth, apparent velocity and signal‐to‐noise ratio of arriving waves and provides a data set of seismicity from the western SWIR over several years. Compared to the global seismological network, its detection threshold for earthquakes occurring at the western SWIR is more than 0.5 mb lower enabling a more comprehensive study of mid‐ocean ridge processes than the teleseismic earthquake catalogues. We identified a total number of 743 earthquakes occurring at the western part of the SWIR and calculated the body‐wave magnitudes (mb) from P‐wave amplitude picks on the VNA2 broad‐band sensor obtaining a magnitude range from mb 3.18 to mb 5.34. In the years of 2001, 2004, 2005 and 2008, significantly increased event rates indicated four earthquake swarms with up to 164 events lasting for several days. All swarms had strong events registered in the International Seismological Centre catalogue. The relocalization of these events confirmed that all swarms occurred in the same region on the Orthogonal Supersegment. We analysed event and moment release rate histories, b‐values and aftershock decay rates (Modified Omori Law) finding that the swarms of 2001, 2004 and 2005 have similarities in the temporal distribution of seismic moment and event numbers. The swarm of 2008 is smaller with high magnitude events at the swarm's onset which represent shear failure on normal faults. The application of the Modified Omori Law and the b‐value show that the earthquakes of the swarms do not follow the classical main shock–aftershock pattern of purely tectonic earthquake sequences. At the Orthogonal Supersegment, a continuous positive magnetic anomaly along the rift axis, a negative mantle Bouguer anomaly, basalts at the seafloor and potentially volcanic edifices indicate robust magmatic crustal accretion. The high‐resolution bathymetry shows ubiquitous rift‐parallel ridges with steep flanks towards the rift axis indicating high‐angle normal faults. The high‐magnitude earthquakes detected teleseismically during the swarms are generated there. We interpret that the swarms are caused by magmatic accretion episodes at a suggested volcanic centre of the Orthogonal Supersegment and that possible magma injection activates the steeply dipping fault planes.

Teleseismic earthquake swarms at ultraslow spreading ridges: indicator for dyke intrusions?

“…The comparison with the regionally observed swarms at the Neumayer array (Läderach et al this issue) showed, that the swarm duration defined from teleseismic earthquakes largely agreed with the swarm duration as defined by the more numerous regionally recorded events. Only one swarm (2001 December) had a preponderance of large events at its beginning and therefore a significantly shorter duration (1.1 days vs. 7.1 days) in the teleseismic record.…”

“…Section swir_orth displayed repeated seismic swarms concentrated at about 18°–19°E (site 7). The Neumayer seismic array in Antarctica is situated in a favourable location to record earthquakes from this portion of the ridge resulting in a detection threshold of about magnitude 3.2 (Läderach et al this issue). In that study, we show that all seismic swarms detected by the array between 2001 and 2009 are also seen teleseismically.…”

“…However, a meaningful analysis of the time dependence of the event rate can only be performed for a significant number of seismic events larger than the magnitude of completeness for the catalogue. This severely restricted the number of events in the present case leaving only two swarms for analysis (see description of applied method in Läderach et al this issue). However, the criterion of McNutt (1996) proved a helpful tool to identify potential mainshock–aftershock sequences.…”

“…At site 7, the swarm of 12/2001 showed a mainshock, but it occurred very late in the sequence. This swarm was analysed by Läderach et al (this issue) including many smaller regionally recorded events, but the event rate decay could not be approximated by a power‐law. Hence, all but three of 27 swarms analysed here are typical swarms with increased seismicity over longer periods without a dominant mainshock and its aftershocks.…”

“…Dziak et al 1995). Repeated swarm activity at the western SWIR was detected by the Neumayer seismic array in Antarctica between 2001 and 2009 (Läderach et al this issue). This observation further showed that earthquake swarms at ultraslow spreading ridges might be more common than expected from the reduced melt supply.…”