Can high‐temperature, high‐heat flux hydrothermal vent fields be explained by thermal convection in the lower crust along fast‐spreading Mid‐Ocean Ridges?

Abstract

To cite this version:Fabrice Fontaine, M. Rabinowicz, M. Cannat. Can high-temperature, high-heat flux hydrothermal vent fields be explained by thermal convection in the lower crust along fast-spreading Mid-Ocean Ridges?. Geochemistry, Geophysics, Geosystems, AGU and the Geochemical Society, 2017Society, , 18 (5), pp.1907Society, -1925 Abstract We present numerical models to explore possible couplings along the axis of fast-spreading ridges, between hydrothermal convection in the upper crust and magmatic flow in the lower crust. In an end-member category of models corresponding to effective viscosities l M lower than 10 13 Pa.s in a meltrich lower crustal along-axis corridor and permeability k not exceeding 10 216 m 2 in the upper crust, the hot, melt-rich, gabbroic lower crust convects as a viscous fluid, with convection rolls parallel to the ridge axis. In these models, we show that the magmatic-hydrothermal interface settles at realistic depths for fast ridges, i.e., 1-2 km below seafloor. Convection cells in both horizons are strongly coupled and kilometerwide hydrothermal upflows/plumes, spaced by 8-10 km, arise on top of the magmatic upflows. Such magmatic-hydrothermal convective couplings may explain the distribution of vent fields along the East (EPR) and South-East Pacific Rise (SEPR). The lower crustal plumes deliver melt locally at the top of the magmatic horizon possibly explaining the observed distribution of melt-rich regions/pockets in the axial melt lenses of EPR and SEPR. Crystallization of this melt provides the necessary latent heat to sustain permanent 100 MW vents fields. Our models also contribute to current discussions on how the lower crust forms at fast ridges: they provide a possible mechanism for focused transport of melt-rich crystal mushes from moho level to the axial melt lens where they further crystallize, feed eruptions, and are transported both along and off-axis to produce the lower crust.

Crustal Magmatic System Beneath the East Pacific Rise (8°20′ to 10°10′N): Implications for Tectonomagmatic Segmentation and Crustal Melt Transport at Fast‐Spreading Ridges

“…Alternatively, deeper processes such as formation of melt channels in the lower crust, the presence of local sites of dike intrusion from the mantle, or along‐axis convection within the melt‐rich mush zone of the lower crust may also give rise to segmentation of the shallow melt reservoir (Fontaine et al, ; Kelemen et al, ; Natland & Dick, ; Sinton et al, ). While a closely coupling of hydrothermal and magmatic systems is expected our observations of the SAML geometry supports the hypothesis that patterns of melt transport in the lower crust, rather than shallow upper crustal processes, play the dominant role in magma reservoir segmentation.…”

“…The contribution of mantle versus lithosphere processes to the different scales of segmentation continues to be debated and strong feedbacks are expected. Recent models based on observations from fast‐spreading ridges and ophiolite analogs include (1) those that primarily involve mantle melting and transport (e.g., diapiric melt ascent (Le Mée et al, ), and small upper mantle melt heterogeneities, Gomez & Briais, ); (2) those that invoke mantle/lithosphere interactions (e.g., skewness in mantle flow and influence on plate boundary reorganziation, Toomey et al, ; VanderBeek et al, ), melt focusing due to lithosphere thickness gradients (Gregg et al, ; Hebert & Montési, ), influence of plate boundary migration on melt production, and melt focusing due to lithosphere gradients (Carbotte et al, ; Katz et al, ); and (3) those that pertain to finer‐scale segmentation and involve crustal processes (e.g., hydrothermal circulation cells within the upper crust, Fontaine et al, ; Tolstoy et al, , and convection within crustal magma reservoirs, Fontaine et al, ).…”

Fault-controlled deep hydrothermal flow in a back-arc tectonic setting, SE Tyrrhenian Sea

“…For modeling purposes, we have considered only two main units: the pre-Messinian acoustic basement and the Plio-Quaternary sedimentary unit (PQSU). Previous studies on individual hydrothermal systems have shown that permeability variations by one order of magnitude between lithological units represent a good approximation in numerical modelling 18,19,62,83 . In our preferred model we assumed permeability values of 1 × 10 −15 m 2 , 1 × 10 −16 and 1 × 10 −17 m 2 , for faults, sediments and basement, respectively.…”

“…High-resolution multibeam bathymetry constrains spatially hydrothermal activity 6,16 , while vent chemistry helps to determine the source of hydrothermal fluids as well as fluid crustal residence time related to rock permeability 17 . However, the lack of direct information, such as drilling, makes numerical modelling an important tool in understanding fluid/rock interactions and hydrothermal circulation within an active tectonic setting 18,19 . Modelling of deep fluid circulation constrained by geophysical data has already been proposed, as for instance that based on Self Potential (SP) and/or electromagnetic data proposed for subaerial volcanoes 20,21 , also part of arc/backarc systems 22–24 .…”

Heat Transfer From Convecting Magma Reservoirs to Hydrothermal Fluid Flow Systems Constrained by Coupled Numerical Modeling

“…Theissen‐Krah et al (2011, 2016) resolve viscous flow of the crust and hydrothermal circulation by Darcy flow, but the model does not resolve convection within the magma reservoir. Fontaine et al (2017) investigate interactions between magmatic and hydrothermal convection, but magma flow is simplified by applying a dimensionless form of the stream function derived from the Navier‐Stokes equations without temperature‐dependent magma properties. Also, both models do not resolve phase separation for Darcy flow.…”