Thermal Structure Of Subduction Zones Research Articles

Spatio-temporal variability in slab temperature within dynamic 3-D subduction models

SUMMARY Spatio-temporal variability in arc geochemistry and the conditions recorded by exhumed rocks suggest subduction zone thermal structure evolves in time and along-strike. Although much effort has been dedicated to studying subduction zone thermal structure, we lack an understanding of spatio-temporal temperature variability during time-dependent subduction. We model 3-D, dynamic subduction and examine the time evolution of the along-strike temperature difference of the slab’s upper surface (‘slab-top’) at the centre relative to the edge of the subduction zone. We examine this slab-top temperature variability for subduction systems of different widths and with different plate mobilities (i.e. fixed versus free subducting and overriding plates). In all of our models, the main control on slab-top temperature is convergence rate; either by simply controlling the rate of slab sinking or via the effect it has on the decoupling depth (DD). In the early stages of subduction, more rapid convergence at the plate centre produces a cooler slab relative to warmer slab edges. For mature subduction, this flips; a shallower DD at the slab centre produces warmer temperatures with respect to the edge. Importantly, our maximum along-strike temperature changes are reduced (≤50 °C) relative to previous kinematically driven modelling studies, due to a reduced role for slab-top heating via toroidal flow. Our dynamic subduction models, therefore, point towards a strong time dependence in the sense of along-strike temperature variation, but with relatively low absolute values in geometrically simple subduction zones.

Anomalous thermal structures of subduction zones revealed by thermal properties of clinochlore at high temperature and pressure

Clinochlore is a major hydrous mineral in subduction zones, and its thermophysical properties at high temperature and pressure are critical to the thermal structures of subduction zones. Here, we used the pulse heating method to measure thermal diffusivity and thermal conductivity of clinochlore at 0.5−4.0 GPa and 298−1373 K. Our results indicate that upon heating, thermal diffusivity and thermal conductivity decrease from ∼9.6 × 10−7 m2 s−1 to 4.3 × 10−7 m2 s−1 and from ∼3.5 W m−1 K−1 to 1.9 W m−1 K−1, respectively, before dehydration, but this trend is reversed after dehydration. In general, the pressure derivatives for the thermal transport properties also decrease with temperature before dehydration. Lattice heat transfer is the dominant mechanism before dehydration, but fluid is involved after dehydration. Using our experimental data, we simulated the temperature distribution of subducting slabs containing clinochlore at volume fractions of 0%, 10%, 20%, 50%, and 100%. Our simulations showed that the heat insulation effect caused by the presence of clinochlore could result in an increase in temperature by 30−60 K for the upper part of the subducting slab.

On unusual conditions for the exhumation of subducted oceanic crustal rocks: How to make rocks hotter than models

The thermal structure of subduction zones controls many important processes such as metamorphic devolatilization, arc magmatism, and seismicity. However, the thermal state of subducted oceanic slab predicted by subduction zone thermal models down to ∼80 km depth is systematically cooler than inferred from exhumed metamorphic slab rocks, raising questions about the current understanding of subduction dynamics portrayed by these models. Upon synthesizing previous petrological studies and estimating slab ages of ancient subduction zones from metamorphic terranes, we think that exhumation of subducted oceanic metamorphic rocks is extremely rare and likely requires unusual processes that are not an integral component of normal subduction dynamics. We construct simple numerical scenarios to illustrate how the thermal regime under some unusual conditions at the beginning and ending stages of a subduction margin may deviate from the normal subduction process. At the beginning stage of subduction, a shallow maximum decoupling depth (MDD) between the slab and mantle wedge enables viscous mantle wedge flow to reach a shallow depth, bringing heat to make slab rocks warmer than in a mature subduction zone. At the ending stage of subduction, if subduction cessation is in the form of slab stalling and if the stalled slab is not immediately exhumed, the slab rocks can be warmed up by the heat from the warm mantle wedge and underlying asthenosphere. In either situation, the slab rocks can be much warmer than normal before they are exhumed by other dynamic processes. Observed exhumed rocks are not expected to represent the normal subduction process and should not be directly compared with thermal models that are designed for the normal process. To extract important information about general geodynamic processes from these rocks, we must first understand the processes these rocks went through.

Thermal Structure of the Paleo‐Continental Subduction Zone: Insights From Quantitatively Constrained Prograde P–T Paths of Exhumed LT/UHP Eclogites in the Dabie Orogen

AbstractThe discrepancy of geothermal gradients predicted by numerical modeling and recorded by peak metamorphic conditions of exhumed high/ultrahigh pressure (HP/UHP) metamorphic rocks triggers many uncertainties on the thermal structure of subduction zones, which is especially significant for continental subduction zones. Well‐constrained prograde P‒temperature (T) paths of HP/UHP rocks reflect the change in P–T conditions during continental subduction and are insusceptible to later processes, making them robust recorders of the thermal structure of subduction zones. To make a reliable estimation of the thermal structure of a continental subduction zone, the prograde P–T paths of three low (L)T‒UHP eclogites in the Dabie orogen were robustly constrained using a multiple thermobarometry method. The results demonstrate that these eclogites underwent three stages of metamorphic evolution during continental subduction from epidote amphibolite‐facies at ∼450°C–470°C/11–13 kbar, through amphibole eclogite‐facies at 585°C–600°C/17–19 kbar, to peak UHP eclogite‐facies at 600°C–630°C/27–29 kbar, indicating a low and continuously changing geothermal gradient from ∼12°C/km at lower crust depth (40 km) to 10°C/km at depth of 60 km and to ∼6°C/km at sub‐arc depth (100 km). The consistency of this result and the average thermal gradient recovered by peak metamorphic conditions of exhumed HP/UHP metamorphic rocks of continental origin suggests that the continental subduction zones may have a common thermal structure characterized by strongly concave upward geothermal gradient. A comparison of the results of this study and analytical models shows that shear heating may play a crucial role in the thermal evolution of continental subduction zones.

Supplemental Material of Thermal properties of clinochlore under high-temperature and -pressure conditions and implications for anomalous thermal structures of subduction zones.

Supplemental Material of Thermal properties of clinochlore under high-temperature and -pressure conditions and implications for anomalous thermal structures of subduction zones.

Redox species and oxygen fugacity of slab-derived fluids: Implications for mantle oxidation and deep carbon-sulfur cycling

The generation and migration of slab-derived fluids modulate subduction zone seismicity, arc magmatism, and deep volatile cycling. However, the redox species and oxygen fugacity (fO2) (hereafter expressed as log units relative to the fayalite–magnetite–quartz buffer, △FMQ) of slab-derived fluids are highly debated. Here we conducted phase equilibria modeling on altered oceanic crust (AOC) and serpentinites along typical subduction geotherms in the C-S-bearing system over a pressure range of 0.5–6 GPa. With the averaged compositions of AOC and serpentinite, our calculated results show that oxidized carbon-sulfur species dominate slab-derived fluids during slab subduction. As a result, slab-derived fluids are highly oxidized and at or above the typical △FMQ values of arc magmas at forearc to subarc depths. The predicted oxidized carbon and sulfur species are compatible with natural observations in fluid inclusions from many oceanic HP metamorphic rocks. More importantly, it is revealed that, the redox state of slab-derived fluids is primarily controlled by the redox budget (RB) of the slab prior to subduction. Subduction-zone thermal structure, however, only exerts a minor influence on the slab-derived fluid fO2, which is supported by the similar fO2 ranges in arc lavas from cold and hot subduction zones. Our models further show that, if an open system is assumed, most of carbon (>70%) and sulfur (>50%) in cold subducted AOC and serpentinite would be lost at subarc depths. Small amounts of carbon and sulfur could be transported into the deeper mantle via closed-system subduction and open-system cold subduction, supplying the source materials for volatile-rich intraplate magmas and superdeep diamonds.

Nitrogen enrichments in sheeted dikes and gabbros from DSDP/ODP/IODP Hole 504B and 1256D: Insights into nitrogen recycling in Central America and global subduction zones

The fate of subducted crustal nitrogen (N) in modern-style subduction zones remains hotly debated from nearly complete return to the surface via arc volcanoes to net ingassing into the deep mantle beyond the sub-arc depth. One of the major obstacles for this controversy is the lack of constraint on the input N flux from altered oceanic crust (AOC). While notable N enrichment by low-temperature hydrothermal alteration has been widely reported in the upper volcanic section of AOC, N behavior during high-temperature alteration of the underlying intrusive oceanic crust (i.e., sheeted dikes and gabbros) remains poorly understood. Here we examined the bulk-rock N concentrations and isotopic compositions of the AOC (with a main focus on sheeted dikes and gabbros, yet including basalts from sections that have not been examined before) from ODP/IODP Hole 1256D and DSDP/ODP Hole 504B, which are two reference sites for subducting material into the warm Central America (CA) subduction zone. The results show that the intrusive sections have comparable N concentrations and isotopic compositions to their overlying upper volcanic rocks at 1256D (sheeted dikes: 16.4±7.2 ppm and −1.3±1.5‰; gabbros: 11.3±5.6 ppm and +1.3±1.0‰; basalts: 11.8±3.8 ppm and +0.9±1.7‰) and 504B (sheeted dikes: 7.6±2.1 ppm and +2.1±1.7‰; basalts: 8.0±3.8 ppm and +2.4±2.9‰). These data can be readily explained by mixing between inherited mantle N and secondary N mostly derived from seawater/sediments with a minor contribution from abiotic N2 reduction. The secondary N in sheeted dikes and gabbros was mainly incorporated at moderate- to high-temperature (>250 °C) alteration stages and likely resides in chlorite, secondary plagioclase and amphibole. These new data clearly indicate that the intrusive section of AOC is a non-negligible N reservoir. Using these data, together with previously published N concentration data of basalts from Holes 1256D and 504B and of subducting sediments from Hole 1039B offboard the CA trench, we obtained a total N input flux of 3.1±0.6 × 109 mol‧yr−1 to 4.0±0.8 × 109 mol‧yr−1 into the CA trench. Integrating recently published N output flux of 0.58 × 109 to 1.4±0.6 × 109 mol‧yr−1 at the CA arc, an N recycling efficiency of 15±3–45±5% was yielded for the CA margin, indicating that up to 55±5–85±3% of subducted slab N is possibly transported beyond the sub-arc depth in this warm subduction zone. Based on the new finding of comparable N enrichment between basalts and intrusive oceanic crust, our modeling also gave a revised N recycling efficiency of 12±10–17±12% for the cold Izu-Bonin-Mariana (IBM) subduction zone, which is consistent with previous estimate of 4–17%. While these new estimates do not provide a conclusive constraint on the effect of subduction-zone thermal structure on N recycling, they consistently suggest that a large fraction of slab N (>50%) can survive the sub-arc filter and be delivered to the deeper mantle in global subduction zones.

Thermal Conductivity and Thermal Diffusivity of Talc at High Temperature and Pressure With Implications for the Thermal Structure of Subduction Zones

AbstractThe thermal conductivity () and thermal diffusivity () of talc have been measured over a range of temperature (298–1,373 K) and pressure (0.5–3.0 GPa) conditions using the transient plane‐source method. The results show that both the thermal conductivity and thermal diffusivity are dependent upon the prevailing temperature and pressure conditions to a certain extent. As the temperature and pressure increase, the thermal diffusivity monotonically decreases, while the thermal conductivity initially decreases between 298 and 973 K and then increases from 973 to 1,173 K. At low temperatures, phonon scattering is the dominant mechanism for heat transfer; at higher temperatures, photon radiation and dehydration become more prevalent. At temperatures greater than 1,173 K, the thermal conductivity decreases significantly due to aqueous liquid accumulation. Talc may be the cause of the high geothermal gradient in the hot subduction zone.

The thermal structure of subduction zones predicted by plate cooling models with variable thermal properties

SUMMARY Previous modelling studies have investigated the effects of experimentally constrained thermal properties (i.e. thermal conductivity, specific heat and the thermal expansion coefficient) on the thermal structure of subduction zones. However, these studies have not carefully considered whether the assumed thermal structure of the slab before subduction is consistent with geophysical observations. This study investigates the effects of thermal properties on the thermal structure of the Tohoku subduction zone, northeast Japan, by using the slab temperature at the trench determined from plate cooling models. Three types of thermal properties were tested: constant, temperature-dependent and temperature- and lithology-dependent types. For each case, the parameters for the plate cooling models were inferred based on the observed surface heat flow and seafloor depth using Bayes’ theorem. It was found that the predicted temperature and location of phase boundaries in the slab, which are possibly related to intermediate-depth earthquakes, are similar for the three cases. This suggests that, in the Tohoku subduction zone, constant thermal properties can be used in modelling to examine phenomena related to slab dehydration. The depth uncertainties for isotherms in the oceanic plate and slab increase with temperature, and are about ±10 and ±20 km for the 600 and 1200 °C isotherms, respectively. When this uncertainty is considered, the location of the serpentinite-out boundary matches that of the lower plane of double seismic zone, suggesting that dehydration may be important in triggering intermediate-depth seismicity. However, the large uncertainty makes it difficult to discuss in detail the origins of intraplate earthquakes, the lithosphere–asthenosphere boundary, and the lower boundary of the slab in terms of temperature.

High‐P metamorphism of garnet–epidote–amphibole schists from the Yuli Belt, Eastern Taiwan: Evidence related to warm subduction

AbstractHigh‐P (HP) rocks exhumed from ancient subduction zones have been controversially argued to record much warmer geotherms than numerical modelling results. The Yuli belt in Eastern Taiwan was regarded as a mélange consisting of HP rock‐bearing tectonic blocks within metasediments (or in situ schists), witnessing the subduction of the South China Sea Plate beneath the Philippine Sea Plate and the subsequent collision during the period of 15–5 Ma. Two garnet–epidote–amphibole schist samples selected from the Tamayen block in the Yuli belt are investigated for their metamorphic evolution. The rocks comprise garnet and paragonite porphyroblasts in a foliated matrix consisting of amphibole, epidote, chlorite, quartz, albite and rutile. Large paragonite flakes crosscut clearly the main foliation, and in places develop narrow coronae of muscovite, albite or chlorite. Fine‐grained pargasite (Amp1) is the dominant variety of amphibole that defines the foliation, whereas a few coarser grains (Amp2) penetrate the foliation and both are further rimmed by a narrow light‐coloured amphibole (Amp3). Petrographic observation and phase equilibria modelling using thermocalc suggest three stages of metamorphic evolution for these rocks. The pre‐peak prograde to the peak‐stage evolution can be revealed from garnet zoning to occur in the mineral assemblages containing lawsonite and glaucophane which cannot be observed petrographically due to the post‐peak decompressional evolution. The peak conditions are constrained to be ~22 kbar/545℃ using the isopleths of the maximum pyrope and the corresponding grossular contents in P–T pseudosections. The post‐peak decompression includes three substages with different fluid behaviours: the first substage is dominated by lawsonite dehydration with the growth of paragonite; the second substage is marked by the decompressional path proceeding in fluid‐absent fields; and the third substage occurs when mineral assemblages become again saturated with water at 10–9 kbar or 12–10 kbar where glaucophane is metastably transformed into Amp1, followed by the growth of Amp2. The late‐retrograde stage is defined to be cooling with decompression on the basis of the growth of Amp3, the coronae around paragonite and the matrix albite. The peak P–T estimates for garnet–epidote–amphibole schists suggest a geotherm of ~7℃/km, much cooler than those in the previous researches in the Yuli belt, and well consistent with the results from numerical thermal modelling. We argue that HP rocks from ancient orogenic belts can robustly constrain the thermal structure of subduction zones if their peak P–T conditions as well as the post‐peak decompressional evolution are well recovered.

Deep decoupling in subduction zones: Observations and temperature limits

AbstractThe plate interface undergoes two transitions between seismogenic depths and subarc depths. A brittle-ductile transition at 20–50 km depth is followed by a transition to full viscous coupling to the overlying mantle wedge at ∼80 km depth. We review evidence for both transitions, focusing on heat-flow and seismic-attenuation constraints on the deeper transition. The intervening ductile shear zone likely weakens considerably as temperature increases, such that its rheology exerts a stronger control on subduction-zone thermal structure than does frictional shear heating. We evaluate its role through analytic approximations and two-dimensional finite-element models for both idealized subduction geometries and those resembling real subduction zones. We show that a temperature-buffering process exists in the shear zone that results in temperatures being tightly controlled by the rheological strength of that shear zone’s material for a wide range of shear-heating behaviors of the shallower brittle region. Higher temperatures result in weaker shear zones and hence less heat generation, so temperatures stop increasing and shear zones stop weakening. The net result for many rheologies are temperatures limited to ≤350–420 °C along the plate interface below the cold forearc of most subduction zones until the hot coupled mantle is approached. Very young incoming plates are the exception. This rheological buffering desensitizes subduction-zone thermal structure to many parameters and may help explain the global constancy of the 80 km coupling limit. We recalculate water fluxes to the forearc wedge and deep mantle and find that shear heating has little effect on global water circulation.

Dehydration at subduction zones and the geochemistry of slab fluids

Subducting oceanic slabs undergo metamorphic dehydration with the increase of temperature and pressure during subduction. Dehydration is an essential step for element recycling, and slab fluids are critical agents for mediating slab-mantle interaction. Dehydration is mainly controlled by the thermal structure of subduction zones and the stability of hydrous minerals. At fore-arc depths, slab dehydration produces aqueous fluid with dissolved salts such as NaCl. As subduction proceeds deeper, the content of silicate components increases. At sub-arc and post-arc depths, a hydrous silicate melt is likely to form, or a supercritical fluid could arise from complete miscibility between silicates and H2O. The partitioning of elements between slab fluid and the residual solid rock is controlled by the type of fluid, and generally it is the supercritical fluid that is the most capable of mobilizing trace elements, being an effective carrier even for high field strength elements. Understanding the chemistry of slab fluids relies on sophisticated integration of experiments, theoretical computation and investigation of natural rock samples. This contribution focuses on the content and speciation of key volatiles, including carbon, nitrogen and sulfur, in slab fluids as well as important fluid properties such as oxygen fugacity and acidity. The properties of slab fluids show complicated variation under the control of mineral assemblages and T-P conditions. Slab fluids at great depths of subductions have been inferred to be modestly alkaline and not necessarily very oxidizing as often assumed. Further progress in the research of slab dehydration and the chemistry and properties of slab fluids demands urgently the development of innovative experimental and computational technology including in situ analytical methods at high T-P.

Bayesian inversion of surface heat flow in subduction zones: a framework to refine geodynamic models based on observational constraints

SUMMARY Surface heat flow has been widely used to constrain the thermal structure of subduction zones. However, the forward modelling approaches in previous geodynamic studies have only provided limited information on the model parameters controlling the thermal structure, which makes model validation difficult. Here we apply a probabilistic inversion technique based on Bayes’ theorem to surface heat flow data from Tohoku in Japan and Cascadia to simultaneously infer five model parameters that appear to have the greatest influence on the thermal structure of subduction zones. The surface heat flow is predicted via 2-D steady-state thermomechanical modelling. The Metropolis algorithm is used to obtain the posterior probability distributions. A comparison of our results with previous estimates indicates that our activation energy for the shear viscosity of dislocation creep is lower in both regions, and our radiogenic heat production rate in the upper continental crust is lower in Cascadia. These findings suggest that our geodynamic models cannot explain the surface heat flow observations with the acceptable ranges of model parameter values. We therefore need to refine the models by including, for example, the effects of recent backarc extension, vigorous thermal convection beneath the overriding plate and fluid circulation in the uppermost part of the oceanic crust. The approach presented here also allows us to determine trade-offs between the parameters. This study provides a framework to validate and refine geodynamic models based on various types of observations.

Seismic Evidence for Water Transportation in the Forearc off Northern Japan

AbstractThe water cycle plays an essential role in arc volcanism, earthquake generation, mantle rheology, and thermal structure of subduction zones. Previous seismic studies have revealed strong structural heterogeneities in the megathrust zone in Northeast Japan and Hokkaido. However, water transportation in the forearc region remains poorly understood due to the lack of long‐term seismic observatories at the seafloor. Using high‐quality data recently recorded by the permanent ocean‐bottom‐seismometer network (S‐net) in the Pacific Ocean off Northern Japan, we study the fine three‐dimensional P wave velocity (Vp) structure of the crust and upper mantle beneath the Kuril and Tohoku forearc region. Our results reveal high velocities in the mantle‐wedge corner, implying a low degree of serpentinization there. We suggest that the forearc mantle in the study region is cold and anhydrous. The slab interface under the forearc area may have a low permeability, which controls the fluid flux to the mantle wedge and the overriding plate. A low‐Vp layer atop the subducting Pacific plate is interpreted as the subducted oceanic crust. Dehydration of the subducted oceanic crust occurs at depths of 80–120 km, providing a large volume of water to the overriding mantle wedge to produce arc volcanoes, and part of the water may migrate upward to the shallow area. Large megathrust earthquakes (Mw ≥ 6.0) mainly occurred around low‐Vp patches in the megathrust zone. Destruction of the low‐permeability slab interface would result in fluid flow upward, which may trigger large megathrust earthquakes and seismicity in the mantle wedge under the forearc.

Seismicity and mineral destabilizations in the subducting mantle up to 6 GPa, 200 km depth

In the subducting oceanic lithosphere, a significant part of the seismicity is triggered in the mantle, especially along the upper and lower Wadati-Benioff planes. Several studies have investigated the potential involvement of dehydration reactions in the triggering mechanism of mantle earthquakes. Recent experimental results reveal that, under subduction conditions, mechanical instabilities nucleate in strong stable mineral aggregates during the destabilization of minor amounts of antigorite, i.e. the high-temperature serpentine, through a stress transfer, without any fluid overpressure. Here I confront these laboratory results to seismological observations. On one hand, most of the natural hydrous magnesium silicates seem to be known, with experimentally deduced stability limits up to 7 GPa, at least, available as relatively accurate estimates. On the other hand, recent achievements in thermal structure of subduction zones combined with precise hypocentre relocation give access to pressure and temperature conditions at earthquakes hypocentres. A series of P-T diagrams summarizes the stability limit of minerals that may be part of natural peridotites or serpentinites with variable compositions at pressures from 0.5 to 6 GPa and temperatures from 200 to 950 °C, and compares it with seismicity. Both hydrous and anhydrous phases are considered. A myriad of minor metamorphic reactions could participate in a transformation-driven stress transfer, even if the stability limits of serpentine minerals seem to correlate with most of the observed seismicity.

Shear heating reconciles thermal models with the metamorphic rock record of subduction

Some commonly referenced thermal-mechanical models of current subduction zones imply temperatures that are 100-500 °C colder at 30-80-km depth than pressure-temperature conditions determined thermobarometrically from exhumed metamorphic rocks. Accurately inferring subduction zone thermal structure, whether from models or rocks, is crucial for predicting metamorphic reactions and associated fluid release, subarc melting conditions, rheologies, and fault-slip phenomena. Here, we compile surface heat flow data from subduction zones worldwide and show that values are higher than can be explained for a frictionless subduction interface often assumed for modeling. An additional heat source--likely shear heating--is required to explain these forearc heat flow values. A friction coefficient of at least 0.03 and possibly as high as 0.1 in some cases explains these data, and we recommend a provisional average value of 0.05 ± 0.015 for modeling. Even small coefficients of friction can contribute several hundred degrees of heating at depths of 30-80 km. Adding such shear stresses to thermal models quantitatively reproduces the pressure-temperature conditions recorded by exhumed metamorphic rocks. Comparatively higher temperatures generally drive rock dehydration and densification, so, at a given depth, hotter rocks are denser than colder rocks, and harder to exhume through buoyancy mechanisms. Consequently--conversely to previous proposals--exhumed metamorphic rocks might overrepresent old-cold subduction where rocks at the slab interface are wetter and more buoyant than in young-hot subduction zones.

Thermal evolution of an ancient subduction interface revealed by Lu–Hf garnet geochronology, Halilbağı Complex (Anatolia)

The thermal structure of subduction zones exerts a major influence on deep-seated mechanical and chemical processes controlling arc magmatism, seismicity, and global element cycles. Accretionary complexes exposed inland may comprise tectonic blocks with contrasting pressure–temperature (P–T) histories, making it possible to investigate the dynamics and thermal evolution of former subduction interfaces. With this aim, we present new Lu–Hf geochronological results for mafic rocks of the Halilbağı Complex (Anatolia) that evolved along different thermal gradients. Samples include a lawsonite–epidote blueschist, a lawsonite–epidote eclogite, and an epidote eclogite (all with counter-clockwise P–T paths), a prograde lawsonite blueschist with a “hairpin”-type P–T path, and a garnet amphibolite from the overlying sub-ophiolitic metamorphic sole. Equilibrium phase diagrams suggest that the garnet amphibolite formed at ∼0.6–0.7 GPa and 800–850 °C, whereas the prograde lawsonite blueschist records burial from 2.1 GPa and 420 °C to 2.6 GPa and 520 °C. Well-defined Lu–Hf isochrons were obtained for the epidote eclogite (92.38 ± 0.22 Ma) and the lawsonite–epidote blueschist (90.19 ± 0.54 Ma), suggesting rapid garnet growth. The lawsonite–epidote eclogite (87.30 ± 0.39 Ma) and the prograde lawsonite blueschist (ca. 86 Ma) are younger, whereas the garnet amphibolite (104.5 ± 3.5 Ma) is older. Our data reveal a consistent trend of progressively decreasing geothermal gradient from granulite-facies conditions at ∼104 Ma to the epidote-eclogite facies around 92 Ma, and the lawsonite blueschist-facies between 90 Ma and 86 Ma. Three Lu–Hf garnet dates (between 92 Ma and 87 Ma) weighted toward the growth of post-peak rims (as indicated by Lu distribution in garnet) suggest that the HP/LT rocks were exhumed continuously and not episodically. We infer that HP/LT metamorphic rocks within the Halilbağı Complex were subjected to continuous return flow, with “warm” rocks being exhumed during the tectonic burial of “cold” ones. Our results, combined with regional geological constraints, allow us to speculate that subduction started at a transform fault near a mid-oceanic spreading centre. Following its formation, this ancient subduction interface evolved thermally over more than 15 Myr, most likely as a result of heat dissipation rather than crustal underplating.

Thermal impact of magmatism in subduction zones

Magmatism in subduction zones builds continental crust and causes most of Earth's subaerial volcanism. The production rate and composition of magmas are controlled by the thermal structure of subduction zones. A range of geochemical and heat flow evidence has recently converged to indicate that subduction zones are hotter at lithospheric depths beneath the arc than predicted by canonical thermomechanical models, which neglect magmatism. We show that this discrepancy can be resolved by consideration of the heat transported by magma. In our one- and two-dimensional numerical models and scaling analysis, magmatic transport of sensible and latent heat locally alters the thermal structure of canonical models by ∼300 K, increasing predicted surface heat flow and mid-lithospheric temperatures to observed values. We find the advection of sensible heat to be larger than the deposition of latent heat. Based on these results we conclude that thermal transport by magma migration affects the chemistry and the location of arc volcanoes.

Diapir versus along-channel ascent of crustal material during plate convergence: Constrained by the thermal structure of subduction zones

Subduction channel processes are crucial for understanding the material and energy exchange between the Earth’s crust and mantle. Crustal rocks can be subducted to mantle depths, interact with the mantle wedge, and then exhume to the crustal depth again, which is generally considered as the mechanism for the formation of ultrahigh-pressure metamorphic rocks in nature. In addition, the crustal rocks generally undergo dehydration and melting at subarc depths, giving rise to fluids that metasomatize and weaken the overlying mantle wedge. There are generally two ways for the material ascent from subarc depths: one is along subduction channels; the other is through the mantle wedge by diapir. In order to study the conditions and dynamics of these contrasting material ascent modes, systematic petrological-thermo-mechanical numerical models are constructed with variable thicknesses of the overriding and subducting continental plates, ages of the subducting oceanic plate, as well as the plate convergence rates. The model results suggest that the thermal structures of subduction zones control the thermal condition and fluid/melt activity at the slab-mantle interface in subcontinental subduction channels, which further strongly affect the material transportation and ascent mode. The thick overriding continental plate and the low-angle subduction style induced by young subducting oceanic plate both contribute to the formation of relatively cold subduction channels with strong overriding mantle wedge, where the along-channel exhumation occurs exclusively to result in the exhumation of HP-UHP metamorphic rocks. In contrast, the thin overriding lithosphere and the steep subduction style induced by old subducting oceanic plate are the favorable conditions for hot subduction channels, which lead to significant hydration and metasomatism, melting and weakening of the overriding mantle wedge and thus cause the ascent of mantle wedge-derived melts by diapir through the mantle wedge. This may correspond to the origination of continental arc volcanism from mafic to ultramafic metasomatites in the bottom of the mantle wedge. In addition, the plate convergence rate can also affect the material ascent mode, e.g., diapiric extrusion versus along-channel exhumation, by changing the amount of supracrustal rocks carried into the subduction channels, which further regulate the fluid/melt activity and thermo-rheological properties.

Geochemistry: Ions surprise in Earth's deep fluids.

Models indicate that there are strong gradients in element concentrations and in the pH of fluids at the slab–mantle interface — a major discontinuity deep within Earth. This transforms our view of global geochemical transport. See Letter p.420 Matthieu Galvez and co-authors present thermodynamic predictions of fluid–rock equilibria that tie together models of subduction-zone thermal structure, mineralogy and fluid speciation. They find that the pH of fluids in subducted crustal lithologies is uniform and confined to a mildly alkaline range, controlled by rock volatile and chlorine contents, but that the pH of mantle wedge fluids exhibits marked sensitivity to minor variations in rock chemistry. They conclude that this sensitivity of fluid chemistry to carbon, alkali metals and halogens illustrates a feedback between Earth's atmosphere–ocean chemistry and the speciation of subduction-zone fluids via the hydrothermally altered oceanic lithosphere.