Axial Seamount is a hotspot volcano and the most magmatically-active feature on the Juan de Fuca mid-ocean ridge (Figure 1). The Ocean Observatories Initiative (OOI) cabled observatory at Axial provides a ground-breaking new opportunity to ask research questions that require a long-term stable ocean bottom seismometer (OBS) network and complementary geodetic and hydrothermal observations, with the ability to respond to results in near-real time, check instrument performance, and capture all data leading up to, and hopefully during, a seafloor eruption. This study, in collaboration with William Wilcock at University of Washington, uses the OOI cabled OBS network to study earthquake activity at Axial Seamount in order to understand how stress is partitioned spatially and temporally; the broad structure of the active fault system; the changes in physical properties of the volcano leading up to an eruption; and ultimately how seafloor spreading takes places.

Within the scope of this project we have developed and implemented high-precision, near real-time earthquake location algorithms that rapidly update and produce a catalog of fine-scale seismicity at Axial Seamount (Figure 2). The catalog images seismicity to within tens of meters resolution, bringing structural features into focus. We used this new catalog address several research questions related to the structure and dynamics of the Axial volcano, and the nature of the eruption itself. 

We demonstrated that Axial Seamount has a classic volcanic outward dipping ring fault caldera structure that was activated in a normal faulting direction during inflation (pre-eruption) and in a reverse fault direction during deflation (during eruption), consistent with deformation data (Wilcock et al, 2016; Figure 1).  This paper provided one of the highest resolution insights into earthquake activity through an eruption cycle, and thus volcano plumbing, not just on the seafloor, but terrestrially as well.  In addition, it provided the first ground-truthed evidence for a seismic signal associated with lava reaching the seafloor enabling us to provide unprecedented timing of the seismic, geodetic and eruptive processes.  

The ground-truthing of the eruptive signal (referred to as impulsive events) at Axial enabled us to also revisit the 2006 East Pacific Rise (EPR) eruption, where similar signals were observed. In Tan et al. (2016) we demonstrated not only the timing of the EPR eruption with great fidelity, but also that the EPR erupts in a fundamentally different way from Axial Seamount and volcanoes in general (Figure 3).  We demonstrate that, at least in the fast spreading case, mid-ocean ridges may behave more like tears in the crust than volcanoes, with extension stresses dominating the eruption rather than magma pressure. This study shows the unexpected utility of the data from the cabled array to answering questions in far away locations.  

By analyzing the size-frequency distribution of 60,000 earthquakes that locate along the ring fault of Axial Seamount, we found that periodic tidal loading (?20 kPa) causes stresses changes large enough to modulate b values in the earthquake distribution (Tan et al., 2019; Figure 4). We find that b-values are inversely correlated with tidal stress. The earthquake b value varies by ~0.09 per kPa change in Coulomb stress, indicating the potential use of b values to estimate small stress variations in the Earth?s crust. Our results provide a robust validation of the stress dependence of the earthquake b-values as previously observed in laboratory experiments. 

Using seismic noise interferometry, we observed an increase in seismic velocity in the southern part of Axial Seamount in three months before the volcano erupted in 2015, suggesting that compressional stresses dominated due to an inflating pressure source below the northern part of the caldera. A few weeks before the eruption, increased inflation rate of another pressure source below the southeastern part of the caldera resulted in the southern part of the edifice being subsequently dominated by extension. This resulted in a decrease in seismic velocity and an increase in seismicity rate that is concentrated above a region of the shallow magma reservoir previously inferred to have the highest melt content. These results suggest that the eruption was preceded by increased magma influx into a segmented shallow reservoir from which the erupted lavas were subsequently sourced. Our results highlight that fine scale segmentation of shallow reservoirs may be driven by variable influx of eruptible melt, but that interaction can occur during the same diking-eruptive sequence.

These studies emphasize the importance of the Axial site as a unique seafloor laboratory, and the importance of continued high-precision monitoring of earthquake activity as the volcano prepares for the next eruption.

Last Modified: 02/06/2019
Modified by: Felix Waldhauser