@article{levy_bohnenstiehl_sprinkle_boettcher_wilcock_tolstoy_waldhauser_2018, title={Mechanics of fault reactivation before, during, and after the 2015 eruption of Axial Seamount}, volume={46}, ISSN={["1943-2682"]}, DOI={10.1130/g39978.1}, abstractNote={Ocean-bottom seismic and seafloor pressure data from the Ocean Observatories Initiative’s Cabled Array were used to study fault reactivation within Axial Seamount (offshore Oregon, USA). Microearthquakes that occurred during 2015–2016 were located on portions of an outward-dipping ring fault system that was reactivated in response to the inflation and deflation of the underlying magma chamber. Prior to an eruption in April 2015, focal mechanisms showed a pattern of normal slip consistent with the differential vertical uplift of the caldera floor relative to the rim. During the eruption, seismic activity remained localized along these outward-dipping structures; however, the slip direction was reversed as the caldera floor subsided. After the eruption, as the volcano reinflated and the caldera floor uplifted, these faults exhibited sparser seismicity with a more heterogeneous pattern of slip. Monitoring the evolution of ring fault behavior through time may have utility as a metric in future eruption forecasts. INTRODUCTION At active volcanoes, as magma is withdrawn from a chamber during an eruption, faults commonly form in response to the subsidence and collapse of the overlying material—developing as either inward-dipping (normal) or outwarddipping (reverse) structures, depending on the amount of subsidence, geometry of the magma chamber, and tectonic setting (Cole et al., 2005; Holohan et al., 2005; Acocella, 2007; Martí et al., 2008). These collapse structures often exhibit a circular to elliptical pattern in plan view and are commonly referred to as ring faults. Reactivation of ring faults has been documented in regional earthquakes studies (e.g., Nettles and Ekström, 1998; Shuler et al., 2013; Gudmundsson et al., 2016); however, the mechanical role of these structures during different phases of the volcanic cycle remains poorly understood. Axial Seamount is a basaltic shield volcano located at the intersection of the Cobb-Eickelberg hotspot and Juan de Fuca Ridge (offshore Oregon, USA; intermediate spreading rate of 55–60 mm/yr). The summit hosts a caldera at 1500 m below sea level (bsl) that is an elongate depression ~3 km wide and 8.5 km long, with walls up to ~150 m high that are buried by younger lavas to the south (Fig. 1). Multichannel seismic reflection studies have imaged a 3-km-wide by 14-km-long magma chamber offset slightly to the east of the caldera at a depth of 1.1−2.3 km beneath the caldera floor (Arnulf et al., 2014). Volume-predictable eruptive behavior has been proposed based on bottom-pressure recorder (BPR) studies that have tracked the inflation and deflation of Axial Seamount for nearly two decades, capturing diking events in 1998, 2011, and 2015 (Chadwick al., 2012; Nooner and Chadwick, 2016). Seismicity associated with the 1998 and 2011 eruptions was recorded by regional hydrophone arrays (Dziak and Fox, 1999) and by ocean bottom hydrophones (Dziak et al., 2012), respectively. Following the 1998 eruption, temporary arrays of 4–10 ocean bottom seismometers (OBSs) monitored local seismicity for 15 mo (Sohn et al., 2004). During the 1998 and 2011 eruptions, pointsource elastic deformation models indicate that the volume of the magma reservoir decreased by ~0.21 km3 and ~0.15 km3, as dikes propagated 55 km and 33 km, respectively, along the southern rift zone (Chadwick et al., 2012). The most recent eruption at Axial Seamount began on 24 April 2015 and was recorded by a seven-station network of three-component OBSs installed within the caldera as part of the Ocean Observatories Initiative (OOI; http://oceanobservatories.org; Kelley et al., 2014). A dike propagated northward from the eastern margin of the magma chamber, erupting a series of lava flows extending from the northeast caldera floor along the north rift zone up to ~14 km north of the caldera (Chadwick et al., 2016). Seafloor explosions associated with the emplacement of these flows indicate that the 2015 eruption persisted over a period of ~26 d (Wilcock et al., 2016), during which time the volume of the magma reservoir (modeled as a prolate spheroid) decreased by ~0.29 km3 (Nooner and Chadwick, 2016). During the time period immediately surrounding the eruption (January–September 2015), Wilcock et al. (2016) identified two steeply dipping, outward-facing planes of microearthquakes beneath the southern portion of the caldera. These structures were interpreted to represent portions of a ring fault system reactivated in response to the inflation and deflation of the magma chamber. In this study, we created an independent catalog of microearthquakes (median Mw ~1.0) for a longer time period between January 2015 and December 2016. These data confirmed the proposed fault geometry and allowed us to track changes in fault slip direction over a nearly 2 yr period. OBS data were used to generate a time series of composite focal mechanism solutions GEOLOGY, May 2018; v. 46; no. 5; p. 1–4 | GSA Data Repository item 2018138 | https://doi.org/10.1130/G39978.1 | Published online XX Month 2018 © 2018 eological Society of A erica. For permission to copy, contact editing@geosociety.org. A’}, number={5}, journal={GEOLOGY}, author={Levy, S. and Bohnenstiehl, D. R. and Sprinkle, P. and Boettcher, M. S. and Wilcock, W. S. D. and Tolstoy, M. and Waldhauser, F.}, year={2018}, month={May}, pages={447–450} }