Water Migration in the Subduction Mantle Wedge: A Two‐Phase Flow Approach

Abstract

Subduction zones are the main entry points of water into Earth's mantle and play an important role in the global water cycle. The progressive release of water by metamorphic dehydration induces important physical‐chemical processes, including subduction zone earthquakes. Yet, how water migrates in subduction zones is not well understood. We investigate this problem by explicitly modeling two‐phase flow processes, in which fluids migrate through a compacting and decompacting solid matrix. Our results show that water migration is strongly affected by subduction dynamics, which exhibits three characteristic stages in our models: (1) an early stage of subduction initiation; (2) an intermediate stage of gravity‐driven steepening of the slab; and (3) a late stage of quasi steady state subduction. Two main water pathways are found in the models: trenchward and arcward. They form in the first two stages and become steady in the third stage. Depending on the depth of water release from the subducting slab, water migration focuses in different pathways: a shallow release depth (e.g., 40 km) leads the water mainly through the trenchward pathway, a deep release depth (e.g., 120 km) promotes an arcward pathway and a long horizontal migration distance (~300 km) from the trench, and an intermediate release depth (e.g., 80 km) leads water to both pathways. We compare our models with seismic studies from southeast Japan (Saita et al., 2015, https://doi.org/10.1002/2015GL063084) and the west Hellenic subduction zone (Halpaap et al., 2018, https://doi.org/10.1002/2017JB015154) and provide geodynamical explanations for these seismic observations in natural subduction environments.

Correlation of Quaternary Volcano Clusters With Partial Melting of Mantle Wedge, Northeast Japan: A Numerical Model Study

“…In this study, we express the water bound in hydrous minerals and aqueous fluid in pores as the volume fractions of solid and fluid phases, respectively. The effects of dynamic and compaction pressures on behavior of the aqueous fluid were evaluated (e.g., Wada & Behn, 2015; Wang et al, 2019; Wilson et al, 2014). For the sake of simplicity, we neglect the dynamic and compaction pressures and calculate the behaviors of water and aqueous fluid using where ω and ϕ are the mass and volume fractions of water and aqueous fluid (·), respectively, and are the solid and aqueous fluid velocities (m s −1 ), respectively, Γ f ⟶ s and Γ s ⟶ f are the rates of mass transfer from aqueous fluid to water by hydration and from water to aqueous fluid by dehydration, respectively (kg m −3 s −1 ), ρ s and ρ f are the solid and aqueous fluid densities (kg m −3 ), respectively, and κ ω and κ ϕ are the artificial diffusivities of water and aqueous fluid (m 2 s −1 ), respectively.…”

“…In this study, we express the water bound in hydrous minerals and aqueous fluid in pores as the volume fractions of solid and fluid phases, respectively. The effects of dynamic and compaction pressures on behavior of the aqueous fluid were evaluated (e.g., Wada & Behn, 2015;Wang et al, 2019;Wilson et al, 2014). For the sake of simplicity, we neglect the dynamic and compaction pressures and calculate the behaviors of water and aqueous fluid using…”

Numerical modeling of subduction: State of the art and future directions

“…• preferential exhumation of mafic high-pressure rocks from warm subduction settings that promote greater fluid productivity, mobility, and buoyancy (van Keken et al, 2018); • non-stationary (i.e., warmer, transient) temperature field in relatively young subduction zones during exhumation of some HP rocks (e.g., Willner et al, 2004;Federico et al, 2007;Krebs et al, 2008;Blanco-Quintero et al, 2011; Peacock, 2020 and references therein); • upward advection of heat by circulation of rocks in the accretionary prism and the subduction channel (Gerya et al, 2002;Gerya and Stöckert, 2006;Menant et al, 2019Menant et al, , 2020Peacock, 2020); • upward heat advection by updip percolation of slab-derived fluids within the forearc (e.g., Wang et al, 2019b); and • neglecting of additional heat sources, such as shear heating (e.g., Wada and Wang, 2009;van Keken et al, 2018;Kohn et al, 2018;Abers et al, 2020) and/or latent heating from metamorphic reactions (especially from mantle serpentinization in the upper plate, Gerya et al, 2002;Gerya and Stöckert, 2006;Kerswell et al, 2021) by simplified subduction models. Whereas some of these reasons (e.g., shear heating, Wada and Wang, 2009;Kohn et al, 2018;van Keken et al, 2018;Abers et al, 2020) have already been extensively investigated, others need systematic consideration in future numerical modeling studies.…”

“…The addition of water lowers the melting temperature of rocks (i.e., "flux melting"), which can trigger melting even in a relatively cool environment on the top and above a subducting slab. Therefore, slab devolatilization and related melting processes are often implemented in numerical models of subduction using simplified kinematic transport approaches (e.g., Gerya et al, 2002Gerya et al, , 2008aArcay et al, 2005;Faccenda et al, 2009bFaccenda et al, , 2012Quinquis and Buiter, 2014;Menant et al, 2019Menant et al, , 2020 and water diffusion models (Richard et al, 2006;Richard and Bercovici, 2009) as well as more advanced fluid and melt percolation models (e.g., Iwamori, 1998Iwamori, , 2000Iwamori, , 2007Cagnioncle et al, 2007;Hebert et al, 2009;Dymkova and Gerya, 2013;Iwamori and Nakakuki, 2013;Wilson et al, 2014;Tian et al, 2019b;Wang et al, 2019b;Cerpa et al, 2019). Numerical modeling of fluid and melt transport in subduction zones is thus technically diverse and spans a number of interesting scientific directions including:…”

“…• Quantification of depth-dependent fluxes of H 2 O from subducting slabs worldwide and their contribution to the global water cycle in the mantle (van Keken et al, 2011). • Pathways of fluids and melts in subduction zones and their effects for volcanic arc development (e.g., Iwamori, 1998Iwamori, , 2000Iwamori, , 2007Cagnioncle et al, 2007;Hebert et al, 2009;Iwamori and Nakakuki, 2013;Wilson et al, 2014;Tian et al, 2019b;Wang et al, 2019b;Cerpa et al, 2019). • The influence of mantle hydration on subduction channel formation, the viscosity structure above slabs, the degree of rheological coupling between plates and subduction dynamics (e.g., Gerya et al, 2002;Arcay et al, 2005Arcay et al, , 2006Gorczyk et al, 2007a;Gerya et al, 2008a;Hebert et al, 2009;Gerya and Meilick, 2011;Baitsch-Ghirardello et al, 2014a;Quinquis and Buiter, 2014).…”