Hélène Carton, Lamont-Doherty Earth Observatory

Among all the properties of a magma lens one might want to investigate, its presence, its width and its vertical position in two-way time to the seafloor are the most readily obtained from multi-channel seismic reflection imaging (of which the first product is an “image” in either two-way time of wave propagation or depth if a reliable large-scale velocity field is available). Thickness of the magma lens (in the absence of a distinct bottom reflection which has not been convincingly imaged so far), the detailed velocity structure (P and S-wave velocities within the lens as well as immediately above and below), and precise estimation of the melt content represent more difficult information to extract, with detailed analyses typically carried out at discrete locations. Such quantitative information however brings powerful constraints that shed light on the interactions between the magmatic system and the hydrothermal system (e.g., high-velocity hydrothermal roof capping the melt lens inferred near 14°S on the Southern EPR), and can be incorporated in modeling studies. If the seismic acquisition is conducted as a 3D survey, proper focusing of out-of-plane energy (otherwise commonly the source of ambiguities in 2D sections) results in enhanced resolution of the resulting images, and the data density and continuity in two directions of space is such that both small-scale features can be detected and larger structures can be mapped and studied in fine detail. That turn can help select the best target locations to apply advanced seismic methods for detailed quantification of magma lens properties, such as 1D or 2D elastic full waveform inversion.

3D seismic data processing of the 2008 Langseth EPR data at the 9°50’N ISS is being carried out to generate a post-stack time migrated volume providing a view of the 3D distribution of magma lens reflectors (both on and off-axis), within an ~20km across-axis by ~27km along-axis area processed with a grid cell size of 6.25m across-axis by 18.75m along-axis. The axial melt lens is ~500-600m wide beneath the northern vent cluster, and ~1km wide beneath the southern vent cluster, with great geometrical complexity especially visible within, but not limited to, the fourth order ridge axis discontinuities. An event at the same two-way time to the seafloor and labeled “frozen top” of the axial magma chamber on the 1985 Conrad lines is not present in coincident Langseth data, which suggests it might have been an artefact related to the reverberatory nature of the seismic sound source. Also consistently imaged in the Langseth data volume are the layer 2A event (turning ray which once projected to zero-offset is generally interpreted to mark the extrusives/dykes contact), and the Moho (though not directly beneath the axial magma body). Some coherent sub-horizontal events are imaged within the extrusives at the ridge crest, while the subsurface continuation of faults observed at the seafloor is not imaged perhaps due to the lack of a strong enough acoustic impedance contrast. Off-axis melt lenses are detected as reflectors identical in character to the axial melt lens: frequency content, polarity, shape on stacked images with broad edge diffractions which if imaged alone could be misinterpreted as dipping reflectors within the crust. The OAMLs are imaged in a variety of positions, both vertically (from a shorter two-way time below seafloor than the axial melt lens, to sub-Moho) and in distance to the axis (nearest at ~1km, furthest at ~17km), and seem to be roughly equi-dimensional bodies. Like for the axial magma body, work beyond the production of a well-focused image is required to determine the thickness and melt content of the off-axis magma bodies.