Experimentally based water budgets for dehydrating slabs and consequences for arc magma generation

Abstract

Phase diagrams of hydrous mid-ocean ridge (MOR) basalts to 330 km depth and of hydrous peridotites to 250 km depth are compiled for conditions characteristic for subduction zones. A synthesis of our experimentally determined phase relations of chlorite, lawsonite, epidote-zoisite, amphibole, paragonite, chloritoid, talc, and phengite in basalts and of phase relations from the literature of serpentine, talc, chlorite, amphibole, and phase A in ultramafics permits calculation of H 2 O contents in hydrous phase assemblages that occur in natural compositions. This yields the information necessary to calculate water budgets for descending slabs. Starting from low-grade blueschist conditions (10-20 km depth) with H 2 O contents between 5 and 6 wt% for hydrated oceanic crust, complete dehydration is achieved between 70 and >300 km depth as a function of individual slab geotherms. Hydrous phases which decompose at depth below volcanic arcs are lawsonite, zoisite, chloritoid, and talc (š phengite) in mafic compositions and chlorite and serpentine in peridotite. Approximately 15-35% of the initially subducted H 2 O are released below volcanic arcs. The contribution of amphibole dehydration to the water budget is small (5-20%) and occurs at relatively shallow depth (65-90 km). In any predicted thermal structure, dehydration is a combination of a stepwise and a continuous process through many different reactions which occur simultaneously in the different portions of the descending slab. Such a dehydration characteristic is incompatible with 'single phase dehydration models' which focus fluid flow through a unique major dehydration event in order to explain volcanic fronts. As a consequence of continuously progressing dehydration, water ascending from the slab will be generally available to depth of ca. 150-200 km. The fluid rising from the subducting lithosphere will cause partial melting in the hot portion of the mantle wedge. We propose that the volcanic front simply forms above the mantle wedge isotherm where the extent of melting is sufficient to allow for the mechanical extraction of parental arc magmas. Thermal models show that such an isotherm (ca. 1300ºC) locates below volcanic fronts, slab surface depths below such an isotherm are compatible with the observed depths of the slab surface below volcanic fronts.

Thermal structure due to solid-state flow in the mantle wedge beneath arcs

“…However, the solidus temperatures for fluid-saturated metasediment and metabasalt are very similar [e.g., Johnson and Plank, 1999;Schmidt and Poli, 1998;Nichols et al, 1994]. Thus, geochemical data support the inference that partial melting of subducted sediment and/or basalt is common in present-day subduction zones.…”

“…Clearly, a large variety of thermal models with temperature-dependent viscosity satisfy the constraints on arc lower crust and shallow mantle temperature. Panels B-F also show experimental, fluid-saturated solidii for sediment [Johnson and Plank, 1999;Nichols et al, 1994] and basalt [Lambert and Wyllie, 1972;Schmidt and Poli, 1998]. Panels C-E compare an isoviscous model with = 10 21 Pa s to models with different values of and Q/(RT ).…”

“…60 km below surface, where top of subducting plate ~ 110 km Kelemen et al, 2002, Figure 5 (Johnson & Plank, 1999) fluid-saturated basalt solidii (dashed: Lambert & Wyllie, 1972 solid: Schmidt & Poli, 1998) fluid-saturated sediment solidus (Nichols et al, 1994) fluid-saturated sediment solidus (Johnson & Plank, 1999) fluid-saturated basalt solidus (Lambert & Wyllie, 1972) fluid-saturated sediment solidus (Nichols et al, 1994) fluid-saturated sediment solidus (Johnson & Plank, 1999) fluid-saturated basalt solidus (Lambert & Wyllie, 1972) fluid-saturated sediment solidus (Nichols et al, 1994) fluid-saturated basalt solidus (Schmidt & Poli, 1998) fluid-saturated sediment solidus (Nichols et al, 1994) fluid-saturated sediment solidus (Johnson & Plank, 1999) fluid-saturated basalt solidus (Lambert & Wyllie, 1972) fluid-saturated sediment solidus (Nichols et al, 1994) SW Oregon NE Japan …”

Origin and evolution of Cenozoic magmatism of Sardinia (Italy). A combined isotopic (Sr–Nd–Pb–O–Hf–Os) and petrological view

“…Osmium is qualitatively considered to be a strongly compatible element in peridotitic assemblages, whereas Re is moderately incompatible (e.g., Shirey and Walker, 1998;Meisel et al, 2001;Hauri, 2002 We propose that the earliest (~43-23 Ma) subduction-related igneous phases in the western Mediterranean are not directly related to the metamorphic dehydration processes of the subducting Alpine Tethys plate, but rather to the pushing effects of this plate over the ancient lower crust of the upper plate. Assuming a convergence rate between the upper (i.e., Europe) and lower plates (i.e., western branch of the Alpine Tethys) as slow as 1-3 cm/yr (e.g., Gueguen et al, 1998;Faccenna et al, 2001;Carminati et al, 2010Carminati et al, , 2012) and a 45° angle of the subducting plate ), the front of the Alpine Tethys slab would have reached the depth at which the amphibole is expected to breakdown (~80-120 km; Schmidt and Poli, 1998;Niida and Green, 1999;Syracuse and Abers, 2006) between 3.8 Myr (3 cm/yr, 80 km depth) and 17 Myr (1 cm/yr, 120 km depth) later.…”

A community benchmark for subduction zone modeling

“…This thermomechanical approach should yield a reasonable prediction of the temperatures in the slab, which then can be linked to the metamorphic processes leading to slab dehydration, arc volcanism and seismogenesis. High accuracy of the thermal models is needed since thermodynamic, field-based and experimental predictions of the relevant metamorphic phase changes predict relatively narrow pressure-temperature ranges for the key phase changes (Schmidt and Poli, 1998;Connolly and Kerrick, 2002;Hacker et al, 2003;Forneris and Holloway, 2004;Abers et al, 2006).…”