Anatomy of an active submarine volcano

Abstract

Most of the magma erupted at mid-ocean ridges is stored in a mid-crustal melt lens that lies at the boundary between sheeted dikes and gabbros. Nevertheless, images of the magma pathways linking this melt lens to the overlying eruption site have remained elusive. Here, we have used seismic methods to image the thickest magma reservoir observed beneath any spreading center to date, which is principally attributed to the juxtaposition of the Juan de Fuca Ridge with the Cobb hotspot. Our results reveal a complex melt body beneath the summit caldera, which is ~14 km long, 3 km wide and up to 1 km thick. The estimated volume of the reservoir is 18-30 km 3 , more than two orders of magnitude greater than the erupted magma volumes of the 1998 and 2011 eruptions. Our images show a network of sub-horizontal to shallow dipping (<30°) features that we interpret as pathways facilitating melt transport from the magma reservoir to the eruption sites.

Seismic Signatures of Hydrothermal Pathways Along the East Pacific Rise Between 9°16′ and 9°56′N

“…Hereinafter, we will refer to the method as wave equation redatuming. In the past several years, the wave equation redatuming came into focus as the need to better understand deep marine subsurface (Arnulf et al, , ; Arnulf, Harding, Kent, et al, ; Arnulf, Harding, Singh, et al, ; Ghosal et al, ; Harding et al, ; Henig et al, ; Wang et al, ). There are several reasons behind the application of the wave equation redatuming on the data set collected along the EPR.…”

Distribution of melt along the East Pacific Rise from 9°30′ to 10°N from an amplitude variation with angle of incidence (AVA) technique

“…Full waveform inversion is computationally expensive and requires a good prior knowledge of the long wavelengths of the velocity model. Except for the recent work of Arnulf et al (2014), who examined the physical properties of the Axial Volcano magma body in 2-D, published applications of elastic full waveform inversion to oceanic spreading centre AML reflections have been limited thus far to 1-D analysis of data from point locations (Collier & Singh 1997Singh et al 1998Singh et al , 1999Canales et al 2006;Xu et al 2014). Most recently, a 1-D waveform inversion study was performed using data from our along-axis EPR 2008 survey at two contrasting locations: 9…”

Contrasting styles of deep-marine pyroclastic eruptions revealed from Axial Seamount push core records

“…Although areas of active hydrothermal venting and eruptive fissures intersecting the caldera currently occur at Axial Seamount ( Fig. 1; Hannington and Scott, 1988;Fornari and Embley, 1995;Arnulf et al, 2014), HMT lithofacies were not observed in strata younger than ∼870 years (Fig. 4).…”

“…These temperatures typically occur near the pillow lava/sheeted dike transition (Alt et al, 1986) or at shallower levels directly below hydrothermal fields (Hannington et al, 1998). Considering the depth range of Axial's active magma chamber (1.1-2.3 km; Arnulf et al, 2014) and likely steep geothermal gradients near the top of MOR melt lens, where hydrothermal cells react with the lower crust (Sleep, 1991;Alt, 1995), we estimate that the subsurface provenance depth for HMT particles would have been ∼600-100 m below sea floor (mbsf). The combined effects of hydrostatic and lithostatic pressure up to 29.9 MPa at these depths (using crustal bulk density of 2.4 kg/m 3 ; Gilbert et al, 2007) suggests that the depth of phreatomagmatic fragmentation is limited to the critical point of seawater (30 MPa; Bischoff and Rosenbauer, 1985;Portner et al, 2014).…”