Search

Seismic data from several long-running broadband sensors around the globe will be used to investigate the statistical significance and geologic interpretation of negative velocity discontinuities in the upper-most mantle. Several previous studies have identified negative polarity arrivals in S-wave receiver function data which are variously interpreted as lithosphere-asthenosphere and/or mid-lithosphere boundaries.One-dimensional joint-inversion is applied using the LitMod framework, which is a Bayesian statistical method driven by a Markov Chain Mote Carlo algorithm.LitMod uses a thermodynamically consistent physical model of the mantle and thus provides important constraints for the interpretation of the receiver function results.Joint inversions combine Rayleigh wave phase velocity measurements, both P and S-wave receiver functions, absolute elevation, and geoid height.Particular attention is given to the calculation and inversion of S-wave receiver function data, which represents a new addition to the LitMod framework.

The present-day thermochemical structure of the subcontinental mantle holds key information on its origin and evolution and informs exploration strategies, natural hazard management and evolutionary model of the Earth system. As such, unravelling the nature of the continental lithosphere, its modification through time and its interactions with the sublithospheric mantle and the atmosphere/hydrosphere constitute some of the main goals of modern geoscience. Despite its fundamental importance, imaging the fine-scale thermochemical structure of the lithosphere using indirect (remote) data is plagued with difficulties, which has traditionally left the analysis of xenoliths and xenocrysts as the only reliable approach.In recent years, however, ‘simulation-based’ inverse methods that integrate multiple geophysical and geochemical datasets within an internally- and thermodynamically-consistent platform have opened new and promising ways to address this ‘grand challenge’. In this presentation, I will discuss i) some recent progress, case studies and future directions on the mapping of the thermochemical structure of the continental lithosphere, and ii) their predictive power for the energy and critical minerals sectors and possible implications for planetary exploration in general.

The chemical composition of erupted basalts provides a record of the thermo-chemical state of their source region and the melting conditions that lead to their formation. Here we present the first probabilistic inversion framework capable of inverting both trace and major element data of mafic volcanic rocks to constrain mantle potential temperature, depth of melting, and major and trace element source composition. The inversion strategy is based on the combination of (i) a two-phase multi-component reactive transport model, (ii) a thermodynamic solver for the evolution of major elements and mineral/liquid phases, (iii) a disequilibrium model of trace element partitioning and (iv) an adaptive Markov chain Monte Carlo algorithm. The mechanical and chemical evolution of melt and solid residue are therefore modelled in an internally- and thermodynamically-consistent manner. We illustrate the inversion approach and its sensitivity to relevant model parameters with a series of numerical experiments with increasing level of complexity. We show the benefits and limitations of using major and trace element compositions separately before demonstrating the advantages of a joint inversion. We show that such joint inversion has great sensitivity to mantle temperature, pressure range of melting and composition of the source, even when realistic uncertainties are assigned to both data and predictions. We further test the reliability of the approach on a real dataset from a well-characterised region: the Rio Grande Rift in western North America. We obtain estimates of mantle potential temperature ( ∼ 1340 °C), lithospheric thickness ( ∼ 60 km) and source composition that are in excellent agreement with numerous independent geochemical and geophysical estimates. In particular, this study suggests that the basalts in this region originated from a moderately hot upwelling and include the contribution from a slightly depleted source that experienced a small degree of melt or fluid metasomatism. This component is likely associated with partial melting of the lower portions of the lithosphere. The flexibility of both the melting model and inversion scheme developed here makes the approach widely applicable to assessing the thermo-chemical structure and evolution of the lithosphere-asthenosphere system and paves the way for truly joint geochemical-geophysical inversions.

The thermochemical structure of the lithosphere exerts control on melting mechanisms in the mantle as well as the location of volcanism and ore deposits. Imaging the complex interactions between the lithosphere and asthenospheric mantle requires the joint inversion of multiple data sets and their uncertainties. In particular, the combination of seismic velocity and electrical conductivity with data proxies for bulk composition and elusive minor phases is a crucial step towards fully understanding large-scale lithospheric structure and melting. We apply a novel probabilistic approach for joint inversions of 3D magnetotelluric and seismic data to image the lithosphere beneath southeast Australia. Results show a highly heterogeneous lithospheric structure with deep conductivity anomalies that correlate with the location of Cenozoic volcanism. In regions where the conductivities have been at odds with sub-lithospheric temperatures and seismic velocities, we observe that the joint inversion provides conductivity values consistent with other observations. The results reveal a strong relationship between metasomatized regions in the mantle and i) the limits of geological provinces in the crust, which elucidates the subduction-accretion process in the region; ii) distribution of leucitite and basaltic magmatism; iii) independent geochemical data, and iv) a series of lithospheric steps which constitute areas prone to generating small-scale instabilities in the asthenosphere. This scenario suggests that shear-driven upwelling and edge-driven convection are the dominant melting mechanisms in eastern Australia rather than mantle plume activity, as conventionally conceived. Our study offers an integrated lithospheric model for southeastern Australia and provides insights into the feedback mechanism driving surface processes.

The recycling of oceanic lithosphere into the deep mantle at subduction zones is one of the most fundamental geodynamic processes on Earth. During the closure of an ocean, ancient oceanic slabs are thought to be consumed entirely in subduction zones due to their negative buoyancy. Yet, it is recently suggested that small pieces of oceanic slabs could be trapped along paleo-subduction zones. What remains far more enigmatic is whether significant portions of paleo-oceanic lithosphere could eventually avoid the fate of subduction and be accreted to continental lithosphere, thus contributing to continental growth through time. We present seismic evidence for a preserved paleo-oceanic lithosphere beneath the Junggar region in northwestern China. We show that unsubducted oceanic lithosphere in the West Junggar has been preserved beneath the Junggar Basin, becoming a piece of the Eurasian continent. This scenario is likely to have occurred in other continents throughout Earth’s history, providing an additional and commonly underestimated contribution to the growth of continental lithosphere.

Northeast China hosts one of the largest Cenozoic intraplate volcanic regions in the world. However, the mechanisms that generate the volcanism, its spatial-temporal distribution, and compositional signatures remain highly debated due to the lack of high-resolution images of the mantle's thermochemical structure. We jointly inverted new surface-wave dispersion data, surface heat flow, geoid height, and elevation data to image the fine-scale thermal and compositional structures beneath northeast China and infer regions of partial melting in the mantle. Our model reveals a complex circulation pattern in the asthenosphere and a highly variable lithospheric structure. Combining predictions from our model with independent geochemical data from recent basaltic volcanism, we demonstrate that the generation, location, and composition of intraplate volcanism in this region are controlled by the interaction between shallow asthenospheric circulation and lithospheric thickness. The modeling approach and correlations between basaltic composition and mantle state identified in our study are globally applicable to assessing mantle conditions over time in other continental regions.

The formation and preservation of compositional heterogeneities inside the Earth affect mantle convection patterns globally and control the long-term evolution of geochemical reservoirs. However, the distribution, nature, and size of reservoirs in the Earth’s mantle are poorly constrained. Here, we invert measurements of travel times and amplitudes of seismic waves interacting with mineralogical phase transitions at 400–700-km depth to obtain global probabilistic maps of temperature and bulk composition. We find large basalt-rich pools (up to 60% basalt fraction) surrounding the Pacific Ocean, which we relate to the segregation of oceanic crust from slabs that have been subducted since the Mesozoic. Segregation of oceanic crust from initially cold and stiff slabs may be facilitated by the presence of a weak hydrated layer in the slab or by weakening upon mineralogical transition due to grain-size reduction.

Aggregates of corundum crystals with skeletal to hopper morphology occur in pyroclastic rocks erupted from Cretaceous basaltic volcanoes on Mt Carmel, N. Israel. The rapid growth of the crystals trapped volumes of the parental Al2O3-supersaturated melt; phenocrysts of tistarite (Ti2O3) in the trapped melts indicate crystallization at oxygen fugacities 6–7 log units below the Iron-Wüstite buffer (fO2 = ΔIW − 6 to − 7), induced by fluxes of mantle-derived CH4-H2 fluids. Cathodoluminescence images reveal growth zoning within the individual crystals of the aggregates, related to the substitution of Ti3+ in the corundum structure. Ti contents are 2 wt%. Numerical modelling indicates that the first skeletal crystals grew in an open system, from a moving magma. The subsequent linear increase in Ti reflects growth in a partially closed system, with decreasing porosity; the exponential increase in Ti close to melt pockets reflects closed-system growth, leading to dramatic increases in incompatible-element concentrations in the residual melts. We suggest that the corundum aggregates grew in melt/fluid conduits; diffusion modelling implies timescales of days to years before crystallization was terminated by explosive eruption. These processes probably operate in explosive volcanic systems in several tectonic settings.

Simulation-based probabilistic inversions of 3D magnetotelluric (MT) data are arguably the best option to deal with the non-linearity and non-uniqueness of the MT problem. However, the computational cost associated with the modeling of 3D MT data has so far precluded the community from adopting and/or pursuing full probabilistic inversions of large MT datasets. In this contribution, we present a novel and general inversion framework, driven by Markov chain Monte Carlo (MCMC) algorithms, which combines i) an efficient parallel-in-parallel structure to solve the 3D forward problem, ii) a reduced order technique to create fast and accurate surrogate models of the forward problem, and iii) adaptive strategies for both the MCMC algorithm and the surrogate model. In particular, and contrary to traditional implementations, the adaptation of the surrogate is integrated into the MCMC inversion. This circumvents the need of costly offline stages to build the surrogate and further increases the overall efficiency of the method. We demonstrate the feasibility and performance of our approach to invert for large-scale conductivity structures with two numerical examples using different parameterizations and dimensionalities. In both cases, we report staggering gains in computational efficiency compared to traditional MCMC implementations. Our method finally removes the main bottleneck of probabilistic inversions of 3D MT data and opens up new opportunities for both stand-alone MT inversions and multi-observable joint inversions for the physical state of the Earth’s interior., Facultad de Ciencias Astronómicas y Geofísicas

In the context of whole-lithosphere structure, the joint inversion of magnetotelluric (MT) with seismic data is particularly interesting as they provide complementary information on the thermal structure, fluid pathways and water content. Both data sets can put tight constrains on the first-order thermal structure and mineralogical structure of the lithosphere, but only MT is strongly sensitive to anomalous features such as hydrogen content, minor conductive phases and/or small volumes of fluid or melt. This makes joint inversions of MT with other observables a powerful means to detect fluid pathways in the lithosphere including the locus of partial melting, ore deposits and hydrated (or metasomatized) lithologies. This unique potential of joint inversions of MT with other datasets has given impetus to the acquisition of collocated MT and seismic data over large regions. Concrete examples are the US Array, Sinoprobe in China, and the AusLAMP/AusArray in Australia. These multi-disciplinary programs are providing high-quality seismic and MT data with unprecedented resolution and coverage, allowing the pursuit of large-scale 3D joint inversions to image the structure, dynamics and evolution of the whole lithosphere and upper mantle. Within probabilistic approaches the solution to the inverse problem is given by the so-called posterior probability density function which provides complete information about the unknown parameters and their uncertainties conditioned on the data and modelling assumptions. Joint probabilistic inversions of MT and seismic data have been successfully implemented in the context of 1D MT data only. For the cases of 2D and 3D MT data, however, the large computational cost of the MT forward problem has been the main impediment for pursuing probabilistic inversions, as the number of forward solutions required are typically on the order of 105 – 107. To overcome this limitation, we have recently presented a novel strategy [2,3], called RB+MCMC, that computes 3D MT surrogate models and uses complementary parameterizations to couple different data sets. This strategy reduces the computational cost of the 3D MT forward solver and allow us to perform full joint probabilistic inversions of MT and other datasets for the 3D imaging of deep thermochemical anomalies. In this contribution, we first illustrate the benefits and general capabilities of our method for 3D joint probabilistic inversions of MT with other datasets using whole-lithosphere synthetic models. Last, as part of the Exploring for the Future program, we present results of the first joint probabilistic inversion of 3D MT in southeast Australia using the AusLAMP data and a seismic velocity model derived from teleseismic tomography [4]. These results demonstrate the capabilities of our conceptual and numerical framework for 3D joint probabilistic inversions of MT with other geophysical data sets and open up exciting opportunities for elucidating the Earth’s interior in other regions. References[1] Afonso, J.C. et al., (2016), Journal of Geophysical Reseach, 121, doi:10.1002/2016JB013049[2] Manassero, M. C., et al., (2020), Geophysical Journal International, 223(3), doi: 10.1093/gji/ggaa415[3] Manassero, M. C., et al., (2021), doi: 10.1029/2021JB021962[4] Rawlinson, N., et al., (2016), Tectonophysics, doi: 10.1016/j.tecto.2015.11.034

The abrupt changes in mineralogical properties across the Earth’s mantle transition zone substantially impact convection and thermochemical fluxes between the upper and lower mantle. While the 410-km discontinuity at the top of the mantle transition zone is detected with all types of seismic waves, the 660-km boundary is mostly invisible to underside P-wave reflections (P660P). The cause for this observation is debated. The dissociation of ringwoodite and garnet into lower-mantle minerals both contribute to the ‘660’ visibility; only the garnet reaction favours material exchanges across the discontinuity. Here, we combine large datasets of SS and PP precursors, mineralogical modelling and data-mining techniques to obtain a global thermal map of the mantle transition zone, and explain the lack of P660P visibility. We find that its prevalent absence requires a chemically unequilibrated mantle, and its visibility in few locations is associated with potential temperatures greater than 1,800 K. Such temperatures occur in approximately 0.6% of Earth, indicating that the 660 is dominated by ringwoodite decomposition, which tends to impede mantle flow. We find broad regions with elevated temperatures beneath the Pacific surrounded by major volcanic hotspots, indicating plume retention and ponding of hot materials in the mantle transition zone. The mantle transition zone is poorly, mechanically mixed, and acts to impede mantle flow, according to seismic observations integrated with detailed mineral-physics models.

We examine REE (Rare-Earth Element) and isotopic (Sr–Hf–Nd–Pb) signatures in OIB (Ocean Island Basalts) as a function of lithospheric thickness and show that the data can be divided into thin- (12 Ma) sub-sets. Comparison to geophysically constrained thermal plate models indicates that the demarcation age (∼12 Ma) corresponds to a lithospheric thickness of about 50 km. Thick-plate OIB show incompatible element and isotopic enrichments, whereas thin-plate lavas show MORB-like or slightly enriched values. We argue that enriched signatures in thick-plate OIB originate from low-degree melting at depths below the dry solidus, while depleted signatures in MORB and thin-plate OIB are indicative of higher-degree melting. We tested quantitative explanations of REE systematics using melting models for homogeneous fertile peridotite. Using experimental partition coefficients for major upper mantle minerals, our equilibrium melting models are not able to explain the data. However, using a new grain-scale disequilibrium melting model for the same homogeneous lithology the data can be explained. Disequilibrium models are able to explain the data by reducing the amount of incompatible element partitioning into low degree melts. To explore new levels of detail in disequilibrium phenomena, we employ the Monte-Carlo Potts model to characterize the textural evolution of a microstructure undergoing coarsening and phase transformation processes simultaneous with the diffusive partitioning of trace elements among solid phases and melt in decompressing mantle. We further employ inverse methods to study the thermochemical properties required for models to explain the OIB data. Both data and theory show that OIB erupted on spreading ridges contain signatures close to MORB values, although E-MORB provides the best fit. This indicates that MORB and OIB are produced by compositionally indistinguishable sources, although the isotopic data indicate that the source is heterogeneous. Also, a posteriori distributions are found for the temperature of the thermomechanical lithosphere-asthenosphere boundary (TLAB), the temperature in the source of OIB (Tp, oib) and the extent of equilibrium during melting (i.e. grain size). TLAB has been constrained to 1200–1300°C and Tp, oib is constrained to be

Special issue Physical properties and observations of the lithosphere-asthenosphere system.-- 21 pages, 15 figures, 3 tables, 1 appendix, The topography of the Iberian Peninsula is characterized by the presence of Variscan and Alpine orogenic belts and foreland basins, but what sets it apart from the rest of Western Europe are the large elevated flat surfaces (700 m above sea-level on average) in its central parts. The origin and support of such high average topography, whether isostatic or dynamic in nature, is a matter of intense debate. To understand Iberian topography, it is key to have a reliable image of the present-day lithospheric thermochemical structure. So far, this structure remains poorly constrained, particularly at mantle level. The goal of this paper is to derive robust estimates of the thermal, compositional and density structure of the lithosphere beneath the Iberian Peninsula from an integrated geophysical-petrological probabilistic inversion of surface wave, elevation, geoid anomaly and heat flow data. Our inversion reveals an average lithospheric thickness of 80–100 km in the Iberian Peninsula with only moderate lateral variations. The most prominent lithospheric thickness change is a steep decrease from the central to the easternmost Pyrenees. The thinnest lithosphere in our models is found below the south-eastern Mediterranean margin (, This project has been funded by Spanish Ministry of Science projects CGL2012-37222 (J. Fullea) and PGC2018-095154-B-I00 (A.M Negredo). J. Fullea is supported by an Atracción Talento senior fellowship (2018-T1/AMB/11493) funded by Comunidad Autonoma de Madrid (Spain). I. Palomeras is funded by the Beatriz Galindo fellowship (BEAGAL18/00090) co-founded by the Spanish Ministry of Education and University of Salamanca (Spain). ICM-CSIC is a Centre of Excellence Severo Ochoa (Spanish Ministry of Science and Innovation, Project CEX2019-000928-S

We present new thermochemical models of the lithosphere and upper mantle beneath the Superior craton and surrounding regions. The study area is dominated by the Archean Superior Province, surrounded by Proterozoic orogenic belts such as the Trans-Hudson Orogen (THO) to the north and the Grenville Orogen to the southeast. Portions of the Rae and Hearne cratons north of the THO are also studied, as is the Mid-continent Rift to the south. Over a period of ∼3 Ga, the region has seen assembly and modification by accretionary and orogenic events, periods of rifting, and the influence of a number of mantle hotspots. Here, we use a probabilistic inverse method to jointly invert Rayleigh wave dispersion data, Vp data, geoid anomalies, surface heat flow, and absolute elevation. The output is a 3D model of the seismic, temperature, bulk density, and compositional structure of the whole lithosphere beneath the Superior craton.The resulting model will provide new opportunities for joint studies of the structure of the upper mantle and will shed light on the thermal and compositional variations beneath the region. In this presentation, we will discuss the results from our model and several robust features that carry important geological and geodynamical implications for this region.

Convective stirring of chemical heterogeneities introduced through oceanic plate subduction results in the marble cake model of mantle composition. A convenient description invokes a chemically unequilibrated mixture of oceanic basaltic crust and harzburgitic lithosphere. Such a composition is required to explain joint observations of shear and compressional waves reflected underneath transition zone (TZ) discontinuities1. The formation of basaltic reservoirs at TZ depth results from complex interaction between phase-change induced chemical segregation, subducted slab downward entrainment, and plume upward advection. However, the dominant mechanism to create and maintain the reservoirs is debated, because both present-day reservoir location and the amount of basalt in these reservoirs are unconstrained. Here, Bayesian inversion of SS- and PP-precursors reflection data indicates that the TZ comprises a global average basalt fraction f = 0.32 ± 0.11. We find the most enriched basaltic reservoirs (f = 0.5-0.6) are associated with recent subduction in the circum-Pacific region. We investigate the efficiency of plate subduction to maintain such reservoirs using global-scale thermochemical convection models2.[1] Waszek, L., Tauzin, B., Schmerr, N.C., Ballmer, M., & Afonso, J.C. (in review). A poorly mixed mantle and its thermal state inferred from seismic waves.[2] Yan, J., Ballmer, M. D., & Tackley, P. J. (2020). The evolution and distribution of recycled oceanic crust in the Earth's mantle: Insight from geodynamic models. Earth and Planetary Science Letters, 537, 116171.

Joint probabilistic inversions of magnetotelluric (MT) and seismic data have great potential for imaging the thermochemical structure of the lithosphere as well as mapping fluid/melt pathways and regions of mantle metasomatism. In this contribution, we present a novel probabilistic (Bayesian) joint inversion scheme for 3D MT and surface-wave dispersion data particularly designed for large-scale lithospheric studies. The approach makes use of a recently developed strategy for fast solutions of the 3D MT forward problem (Manassero et al., 2020, https://doi.org/10.1093/gji/ggaa415) and combines it with adaptive Markov chain Monte Carlo (MCMC) algorithms and parallel-in-parallel strategies to achieve extremely efficient simulations. To demonstrate the feasibility, benefits and performance of our joint inversion method for imaging the temperature and conductivity structures of the lithosphere, we apply it to two numerical examples of increasing complexity. The inversion approach presented here is timely and will be useful in the joint analysis of MT and surface wave data that are being collected in many parts of the world. This approach also opens up new avenues for the study of trans-lithospheric and trans-crustal magmatic systems, the detection of metasomatized mantle, and the incorporation of MT into multi-observable inversions for the physical state of the Earth's interior. We thank Farshad Salajegheh for providing part of his Matlab codes for plotting results 850 Special thanks to Kate Selway and Anandaroop Ray for their suggestions at different 851 stages of this work. The 3D rendering views were created using ParaView (Ahrens et al., 852 2005). 853 MCM thanks support from an International Macquarie Research Excellence Schol854 arship (iMQRES). MCM and JCA acknowledge support from ARC Grant DP160103502, 855 ARC Linkage Grant LP170100233, the ARC Centre of Excellence Core to Crust Fluids 856 Systems (http://www.ccfs.mq.edu.au) and the Centre for Earth Evolution and Dynam857 ics, Geoscience Australia and the European Space Agency via the “3D Earth - A Dy858 namic Living Planet”. FZ acknowledges support from CONICET through grant PIP 112- 859 201501-00192. SZ has been funded by the Spanish Ministry through grant DPI2017-85139- 860 C2-2-R, by Catalan government through grant 2017-SGR-1278 and by the EU’s Hori861 zon 2020 research and innovation programme under the Marie Sklodowska-Curie grant 862 agreement No 777778.

Besides standard thermo-mechanical conservation laws, a general description of mantle magmatism requires the simultaneous consideration of phase changes (e.g. from solid to liquid), chemical reactions (i.e. exchange of chemical components) and multiple dynamic phases (e.g. liquid percolating through a deforming matrix). Typically, these processes evolve at different rates, over multiple spatial scales and exhibit complex feedback loops and disequilibrium features. Partially as a result of these complexities, integrated descriptions of the thermal, mechanical and chemical evolution of mantle magmatism have been challenging for numerical models. Here we present a conceptual and numerical model that provides a versatile platform to study the dynamics and nonlinear feedbacks inherent in mantle magmatism and to make quantitative comparisons between petrological and geochemical datasets. Our model is based on the combination of three main modules: (1) a Two-Phase, Multi-Component, Reactive Transport module that describes how liquids and solids evolve in space and time; (2) a melting formalism, called Dynamic Disequilibirum Melting, based on thermodynamic grounds and capable of describing the chemical exchange of major elements between phases in disequilibrium; (3) a grain-scale model for diffusion-controlled trace-element mass transfer. We illustrate some of the benefits of the model by analyzing both major and trace elements during mantle magmatism in a mid-ocean ridge-like context. We systematically explore the effects of mantle potential temperature, upwelling velocity, degree of equilibrium and hetererogeneous sources on the compositional variability of melts and residual peridotites. Our model not only reproduces the main thermo-chemical features of decompression melting but also predicts counter-intuitive differentiation trends as a consequence of phase changes and transport occurring in disequilibrium. These include a negative correlation between Na2O and FeO in melts generated at the same Tp and the continued increase of the melt's CaO/Al2O3 after Cpx exhaustion. Our model results also emphasize the role of disequilibrium arising from diffusion for the interpretation of trace-element signatures. The latter is shown to be able to reconcile the major- and trace-element compositions of abyssal peridotites with field evidence indicating extensive reaction between peridotites and melts. The combination of chemical disequilibrium of major elements and sluggish diffusion of trace elements may also result in weakened middle rare earth to heavy rare earth depletion comparable with the effect of residual garnet in mid-ocean ridge basalt, despite its absence in the modelled melts source. We also find that the crystallization of basalts ascending in disequilibrium through the asthenospheric mantle could be responsible for the formation of olivine gabbros and wehrlites that are observed in the deep sections of ophiolites. The presented framework is general and readily extendable to accommodate additional processes of geological relevance (e.g. melting in the presence of volatiles and/or of complex heterogeneous sources, refertilization of the lithospheric mantle, magma channelization and shallow processes) and the implementation of other geochemical and isotopic proxies. Here we illustrate the effect of heterogeneous sources on the thermo-mechanical-chemical evolution of melts and residues using a mixed peridotite-pyroxenite source. © 2020 The Author(s) 2020. Published by Oxford University Press. All rights reserved., B.O. and J.C.A. acknowledge funding from Macquarie University, ARC Grant DP160103502 and ARC Linkage Grant LP170100233. B.O., J.C.A. and R.T. also acknowledge support from the ARC Centre of Excellence Core to Crust Fluids Systems (CCFS), the Centre for Earth Evolution and Dynamics (CEED) and the European Space Agency via the ‘3D Earth—A Dynamic Living Planet’

Multiobservable thermochemical tomography (MTT) is a recent computational approach to obtain estimates of the physical state (e.g. temperature distribution, compositional structure and rock properties) for the upper mantle [1]. It allows to jointly invert multiple independent datasets (e.g. gravity, seismic, magnetotelluric) within a thermodynamically-constrained and fully probabilistic framework. Evaluation of the plausibility of different physical states of the mantle with Markov Chain Monte Carlo (MCMC) simulations requires the solution of complex forward problems (e.g. Stokes flow, Maxwell’s equations, etc.) millions of times, making MTT computationally demanding for large-scale inverse problems. Furthermore, the number of parameters in a global study can easily reach several millions, making it increasingly difficult to 1) locate the regions of high probability and 2) sample these regions appropriately.In order to overcome these limitations, we have combined and implemented a number of techniques, such as reduced-order modelling and efficient parallelization of both the forward problems and the MCMC algorithms, which dramatically accelerate the solution of the forward problems. Our software, LitMod1D_4INV and LitMod3D_4INV, allow to compute a proposal in less than 1 second, even when solving multiple complex forward problems together. We develop a multi-level parallel MPI driver for a collection of advanced MCMC sampling strategies to locate and sample high-probability regions efficiently. The massive amounts of data generated by large-scale MTT inversions need to be managed efficiently. We output results to open-source freeware formats, such as HDF5, TileDB, designed for big data problems. We emphasize that our methods and approaches are not only useful for MTT, but for any demanding inverse problem.In this contribution, we will present applications of our software to complex, large-scale MTT problems and discuss its benefits, limitations and future improvements.[1] J.C. Afonso, N. Rawlinson, Y. Yang, D. L. Schutt, A. G. Jones, J. Fullea, W. L. Griffin, 3‐D multiobservable probabilistic inversion for the compositional and thermal structure of the lithosphere and upper mantle: III. Thermochemical tomography in the Western‐Central U.S., Journal of Geophysical Research, 121, doi:10.1002/2016JB013049, 2016

Markov chain Monte Carlo (MCMC) methods have become standard in Bayesian inference and multi-observable inversions in almost every discipline of the Earth sciences. In the case of geodynamic and/or coupled geophysical-geodynamic inverse problems, however, the computational cost associated with the solution of large-scale 3-D Stokes forward problems has rendered probabilistic formulations impractical. Here we present a novel and extremely efficient method to produce ultrafast solutions of the 3-D Stokes problem for MCMC simulations. Our approach combines the individual benefits of Reduced Basis techniques, goal-oriented error formulations, and MCMC algorithms to produce an accurate and computationally efficient surrogate for the forward problem. Importantly, the surrogate adapts itself during the MCMC simulation according to the history of the chain and the goals of the inversion. This maximizes the efficiency of the forward problem and removes the need for preinversion off-line computations to build a surrogate. We demonstrate the benefits and limitations of the method with several numerical examples and show that in all cases the computational cost is of the order of

The deep mantle African Superswell is considered to contribute to the topographic uplift of the Southern African Plateau, but dynamic support estimates vary wildly depending on the approach and data used. One reason for these large disparities is that the role of lithospheric structure, key in modulating deep dynamic contributions to elevation, is commonly ignored or oversimplified in convection studies. We use multiple high-quality geophysical data coupled with xenolith-based geochemical constraints to compute the isostatic lithospheric contribution to the elevation of the Plateau, facilitating isolation of the current dynamic component from the total observed elevation. We employ a multi-observable stochastic algorithm to invert geoid anomaly, surface-wave dispersion data, magnetotelluric data and surface heat flow to predict elevation in a fully thermodynamically and internally-consistent manner. We find that a compositionally-layered 230 ±7 km thick lithosphere is required to simultaneously fit all four data types, in agreement with abundant independent xenolith evidence. Our stochastic modelling indicates a lithospheric contribution to elevation of the order of 670 m, which implies dynamic support arising from the convecting sub-lithospheric mantle of ∼650 m. Our results have important implications for the understanding of lithospheric-deep mantle feedback mechanisms and for calibrating dynamic topography estimates from global convection studies.

LitMod2D_2.0 can be downloaded from https://doi.org/10.20350/digitalCSIC/9063 or GitHub https://github.com/ajay6763/LitMod2D_2.0_package_dist_users., LitMod2D integrates geophysical and petrological data sets to produce the thermal, density, and seismic velocity structure of the lithosphere and upper mantle. We present a new LitMod2D_2.0 package with improvements focused on (i) updated anelastic attenuation correction for anharmonic seismic velocities, (ii) chemical composition in the sublithospheric mantle, and (iii) incorporation of sublithospheric mantle anomalies. Sublithospheric mantle anomalies can be defined with different chemical composition, temperature, seismic velocities, and a combination of them, allowing the application of LitMod2D_2.0 to regions affected by mantle upwelling, subduction, delamination, and metasomatism. We demonstrate the potential application of LitMod2D_2.0 to such regions and the sensitivity of thermal and compositional anomalies on density and seismic velocities through synthetic models. Results show nonlinearity between the sign of thermal and seismic velocity anomalies, and that S wave velocities are more sensitive to temperature whereas P wave velocities are to composition. In a synthetic example of subduction, we show the sensitivity of sublithospheric mantle anomalies associated with the slab and the corner flow on surface observables (elevation, geoid height, and gravity anomalies). A new open-source graphic user interface is incorporated in the new package. The output of the code is simplified by writing only the relevant physical parameters (temperature, pressure, material type, density, and seismic velocities) to allow the user using predefined post-processing codes from a toolbox (flexure, mineral assemblages, synthetic passive seismological data, and tomography) or designing new ones. We demonstrate a post-processing example calculating synthetic seismic tomography, Rayleigh surface-wave dispersion curves, and P wave receiver functions from the output file of LitMod2D_2.0. © 2020. The Authors., This work has been supported by EU Marie Curie Initial Training Network (SUBITOP) (674899-SUBITOP-H2020-MSCA-ITN-2015), the Agencia Estatal de Investigación through projects MITE (CGL2014-59516) and GeoCAM (PGC2018-095154-B-100), and the Agency for Management of University and Research Grants of Catalonia (AGAUR-2017-SGR-847). Thermodynamic databases used can be found in cited references (Afonso & Zlotnik,?2011; Stixrude & Lithgow-Bertelloni,; Xu et al.,). We thank two anonymous reviwers for insightful comments and suggestions on an earlier version of this manuscript. We also thank F. Salajeghegh for sharing his GUI matlab routines. This work has been done using the facilities of the Laboratory of Geodynamic Modeling from ICTJA-CSIC. JCA's work is supported by ARC Linkage Project LP170100233 and ARC Discovery Project DP190102940. LitMod2D_2.0 can be downloaded from https://doi.org/10.20350/digitalCSIC/9063 or GitHub https://github.com/ajay6763/LitMod2D_2.0_package_dist_users.

To address uncertainties in the values and mathematical form of the radiative thermal conductivity k_(rad) in the mantle, we developed new models for the transport, scattering, and absorption of thermal radiation in semitransparent multiphase polycrystalline assemblages. We show that the Rosseland diffusion equation correctly describes the diffusion of thermal radiation and infer the form of the effective spectral coefficients through numerical experimentation. We show that the scattering coefficient depends on the grain size and on interphase contact statistics in complicated ways, but that simplifications can be employed in practice. The effective opacity of a composite random material is a harmonically weighted mixture in the limit of infinitely large grain size and an arithmetically weighted mixture in the limit of infinitesimal grain size. Using existing absorption spectra for major upper mantle minerals, we estimate k_(rad) as a function of temperature, grain size, and petrology. In mantle assemblages, the scattering effect is important for small grain sizes (

The Arizona Transition Zone is a narrow band that separates two of the main and most contrasting tectonic provinces in western US, namely the southern Colorado Plateau and the southern Basin and Range provinces. As such, the internal crustal structure and physical state of this transitional zone hold clues for understanding (i) the amalgamation of these provinces, (ii) the partitioning of deformation due to both past and present-day stress fields, and (iii) the role of thermal versus compositional effects in controlling surface observables. Here we employ and expand a novel multiobservable probabilistic inversion method and jointly invert fundamental mode Rayleigh phase velocities, receiver functions, surface heat flow, geoid height, and absolute elevation to obtain an internally consistent 3-D model of the temperature, density, Vs, and Vp of the Arizona Transition Zone and the southern portions of the Colorado Plateau and Basin and Range. Our results confirm a significant crustal thickening from ∼28 km in the SW of the Arizona Transition Zone and southern Basin and Range to ∼48 km beneath the southern Colorado Plateau. Inverted temperatures agree well with the location of recent volcanism and indicate that the lithosphere-asthenosphere boundary is not deeper than ∼70 km in most of the region. We find that major pre-Cambrian surface structures and/or shear zones separate crustal domains with distinct bulk properties, suggesting that the juxtaposed crustal blocks still retain, at least in part, their original characteristics. However, widespread intrusions of significant volumes of mafic magmas have affected these blocks at different depths, locally overprinting their original compositions and creating highly heterogeneous crustal sections. A dominant and large-scale internal crustal pattern of SW dipping planes/structures is evident in our models, coinciding with the orientation of deep faults previously inferred from earthquake focal mechanisms. While we cannot categorically corroborate the presence of melt or aqueous fluids within the crust, our results are compatible with these scenarios beneath some parts of the Basin and Range, the Mogollon-Datil, and Springerville volcanic fields.

We present a 3D thermochemical model of the North China Craton (NCC) from the surface down to 350 km by jointly inverting surface wave phase velocity data, geoid height, surface heat flow and absolute elevation with a multi-observable probabilistic inversion method. Our model reveals a thin (~ 65–100 km) and chemically fertile lithosphere (87 90) lithospheric mantle is imaged beneath the central TNCO and Ordos Block, reaching depths > 260 km. This lithospheric “keel” is surrounded to the east by a high-temperature sublithospheric anomaly that originates at depths > 280 km. The spatial distribution of this anomaly and its correlation with the location of recent volcanism in the region suggest that the anomaly represents a deep mantle upwelling being diverted by the cratonic keel and spreading onto regions of shallow lithosphere. Our results indicate that the present-day thermochemical structure beneath the NCC is the result of a complex interaction between a large-scale return flow associated with the subduction of the Pacific slab and the shallow lithospheric structure.

We present a new crust and upper mantle cross section of the western India-Eurasia collision zone by combining geological, geophysical, and petrological information within a self-consistent thermodynamic framework. We characterize the upper mantle structure down to 410 km depth from the thermal, compositional, and seismological viewpoints along a profile crossing western Himalayan orogen and Tibetan Plateau, Tarim Basin, Tian Shan, and Junggar Basin, ending in the Chinese Altai Range. Our results show that the Moho deepens from the Himalayan foreland basin (~40 km depth) to the Kunlun Shan (~90 km depth), and it shallows to less than 50 km beneath the Tarim Basin. Crustal thickness between the Tian Shan and Altai mountains varies from ~66 km to ~62 km. The depth of the lithosphere-asthenosphere boundary (LAB) increases from 230 km below the Himalayan foreland basin to 295 km below the Kunlun Shan. To NE the LAB shallows to ~230 km below the Tarim Basin and increases again to ~260 km below Tian Shan and Junggar region and to ~280 km below the Altai Range. Lateral variations of the seismic anomalies are compatible with variations in the lithospheric mantle composition retrieved from global petrological data. We also model a preexisting profile in the eastern India-Eurasia collision zone and discuss the along-strike variations of the lithospheric structure. We confirm the presence of a noticeable lithospheric mantle thinning below the Eastern Tibetan Plateau, with the LAB located at 140 km depth, and of mantle compositional differences between the Tibetan Plateau and the northern domains of Qilian Shan, Qaidam Basin, and North China.

International audience; We reassess the mantle transition zone structure below the northeast Asia margin in the context of subduction of the Pacific plate below the Eurasian continent. We use two independent approaches of teleseismic imaging, namely, compressional-to-shear converted waves (receiver functions) and shear wave underside reflections (SS precursors), and compare them within their statistical uncertainties. We find localized complexity in the interfaces marking solid phase changes in mantle minerals, in terms of both apparent topography and reflectivity. The 660-km discontinuity is doubled, with approximate to 80-km maximum vertical distance between the interfaces, over an 890 x 350 km(2) region between 36-44 degrees N and 130-133 degrees E at the tip of the subducted Pacific plate. A similar complexity exists on the 410-km discontinuity, coinciding with the presence of a deep cluster of seismicity below the Japan Sea. Both methods suggest the presence of low-velocity zones atop the 410, within the mantle transition zone, and below the 660. This complex seismic signature is related to the Pacific plate and interpreted in light of the subduction thermal regime and phase equilibria for a pyrolitic mantle composition. Phase changes manifest themselves as broad zones of velocity gradients with localized doubled or multiple first-order discontinuities, associated with transitions in the olivine, pyroxene, and garnet systems. An average pyrolitic composition and local temperatures of 1000-1300K can explain the observed velocity gradients and multiple discontinuities. We show that the dissolution of stishovite, a high-pressure polymorph of SiO2, into the higher-pressure perovskite mineral, is a possible explanation for the low-velocity zones at the top of the lower mantle. Plain Language Summary In the early 1990s, the idea emerged that the study of seismic discontinuities near 410- and 660-km depths could inform us about the vertical circulation of material in the mantle. These two discontinuities were assigned to solid phase changes of a silicate mineral rich in magnesium (90%) and iron (10%), olivine. Advances in the field of mineralogy and high-pressure physics suggested, however, that other minerals, garnet and pyroxenes, should contribute to the general seismic signature. Due to the limited imaging capability of seismological methods and station deployments, however, a one-to-one correspondence between observed seismic structure and mineralogy has remained elusive, at best controversial. We provide here from two independent approaches compelling evidence for localized complexity in the seismic structure near 410- and 660-km depths, which we relate, through mineralogical modeling, to phase changes in olivine, garnet, pyroxene, and silica systems. This work provides the observational basis for assessing the viability of geodynamic and thermochemical models of the mantle beneath the northeast Asia margin.

A 3-D crustal and upper mantle S-wave velocity model of NE China is constructed by inversion of phase velocity dispersion curves at 6–140 s periods from ambient noise tomography and two-plane surface wave tomography. The seismic data used in this study are collected from 120 China Earthquake Administration (CEA) permanent stations and 127 portable stations of NECESSArray. We observe strong low S-wave velocity beneath the Changbaishan volcano in the upper mantle to at least 200-km depth, which is interpreted as a mantle upwelling beneath the Changbaishan volcano that is consistent with the body wave tomographic image. The Songliao Basin is dominated by a high velocity extending to at least 200-km depth. Built upon the observed velocity anomalies, we propose a sub-lithosphere mantle convection model for NE China in which the upwelling of upper mantle materials from the mantle transition zone to the Changbaishan volcano could induce a local sub-lithosphere convection in the upper mantle and the strong high velocity of the upper mantle beneath the Songliao Basin corresponds to the downwelling limb of this convection cell. The downwelling beneath the Songliao Basin could also induce secondary local convection in the asthenosphere to the west, leading to local asthenospheric upwelling beneath the Abaga and Halaha volcanoes in the Xing'an–Mongolia Orogenic Belt.

We have estimated the power of ventilated hydrothermal heat transport, and its spatial distribution, using a set of recently developed plate models which highlight the effects of axial hydrothermal circulation and thermal insulation by oceanic crust. Testing lithospheric cooling models with these two effects, we estimate that global advective heat transport is about 6.6 TW, significantly lower than most previous estimates, and that the fraction of that extracted by vigorous circulation on the ridge axes (< 1 My old) is about 50 % of the total, significantly higher than previous estimates. These new estimates originate from the thermally insulating properties of oceanic crust in relation to the mantle. Since the crust is relatively insulating, the effective properties of the lithosphere are "crust dominated" near ridge axes (a thermal blanketing effect yielding lower heat flow) and gradually approach mantle values over time. Thus, cooling models with crustal insulation predict low heat flow over young seafloor, implying that the difference of modeled and measured heat flow is due to the heat transport properties of the lithosphere, in addition to ventilated hydrothermal circulation as generally accepted. These estimates may bear on important problems in the physics and chemistry of the Earth because the magnitude of ventilated hydrothermal power affects chemical exchanges between the oceans and the lithosphere, thereby affecting both thermal and chemical budgets in the oceanic crust and lithosphere, the subduction factory, and the convective mantle.

Tectonic plate and lower-mantle motions are often considered independent. Plate tectonic reconstructions reveal long-lived interactions between mantle plumes and mid-ocean ridges that imply feedback between plate boundaries and the deep mantle.

The past ten years have been marked by dramatic advances in four seemingly isolated research fields: Thermodynamic modelling of minerals and rocks at high PT conditions, numerical simulation of the thermomechanical behaviour of the Earth’s interior, efficient decomposition techniques to solve complex simulation-based problems, and probabilistic data analysis and inversion methods. All these disciplines/techniques have individually created true “revolutions” in the way we understand and model natural systems, including the interior of the Earth. Nevertheless, a more profound understanding is still ahead of us from the formal combination of these disciplines/techniques into a single operational framework to study the physical state of the Earth’s interior. In this contribution, I will present and discuss the concept of multi-observable probabilistic tomography or “thermochemical tomography”. This new kind of tomography is particularly designed for studies of the fundamental thermodynamic variables of the Earth’s lithosphere, namely temperature, pressure and chemical composition. Once these variables are known, all physical parameters of interest (e.g. seismic velocities, density, viscosity, conductivity, etc), as well as traditional tomography images, are also retrieved in a thermodynamically-consistent way. The method is built on a simulation-based inversion technique where multiple satellite (e.g. gravity gradients, geoid height, etc) and land-based (e.g. seismic, plate motions, heat flow, etc) datasets can be jointly inverted to maximize the physical consistency of the resulting Earth model. Assembling this large problem required a collaborative effort between thermodynamicists, mineral physicists, geophysicists and geochemists, and marks the first step towards a full coupling between geophysics, geodynamics, thermodynamics, and geochemistry. I will present results for both synthetic and real case studies, which serve to highlight the advantages and limitations of this approach.

The origin of intraplate carbonatitic to alkaline volcanism in Africa is controversial. A tectonic control, i.e., decompression melting associated with far-field stress, is suggested by correlation with lithospheric sutures, repeated magmatic cycles in the same areas over several million years, synchronicity across the plate, and lack of clear age progression patterns. Conversely, a dominant role for mantle convection is supported by the coincidence of Cenozoic volcanism with regions of lithospheric uplift, positive free-air gravity anomalies, and slow seismic velocities. To improve constraints on the genesis of African volcanism, here we report the first radiometric and isotopic results for the Catanda complex, which hosts the only extrusive carbonatites in Angola. Apatite (U-Th-Sm)/He and phlogopite 40Ar/39Ar ages of Catanda aillikite lavas indicate eruption at ca. 500-800 ka, more than 100 m.y. after emplacement of abundant kimberlites and carbonatites in this region. The lavas share similar high-μ (HIMU)-like Sr-Nd-Pb-Hf isotope compositions with other young mantle-derived volcanics from Africa (e.g., Northern Kenya Rift; Cameroon Line). The position of the Catanda complex in the Lucapa corridor, a long-lived extensional structure, suggests a possible tectonic control for the volcanism. The complex is also located on the Bié Dome, a broad region of fast Pleistocene uplift attributed to mantle upwelling. Seismic tomography models indicate convection of deep hot material beneath regions of active volcanism in Africa, including a large area encompassing Angola and northern Namibia. This is strong evidence that intraplate late Cenozoic volcanism, including the Catanda complex, resulted from the interplay between mantle convection and preexisting lithospheric heterogeneities.

International audience; Mantle lithologies in orogenic massifs and xenoliths commonly display strikingly different Hf- and Nd-isotope compositions compared to oceanic basalts. While the presence of pyroxenites has long been suggested in the source region of mantle-derived magmas, very few studies have reported their combined HfNd isotope compositions. We here report the first LuHf data along with ReOs data and S concentrations on the Cabo Ortegal Complex, where the pyroxenite-rich Herbeira massif has been interpreted as remnants of a delaminated arc root. The pyroxenites, chromitites and their host harzburgites show a wide range of whole-rock 187Re/188Os and 187Os/188Os (0.16–1.44), indicating that Re was strongly mobilized, partly during hydrous retrograde metamorphism but mostly during supergene alteration that preferentially affected low-Mg#, low Cu/S pyroxenites. Samples that escaped this disturbance yield an isochron age of 838 ± 42 Ma, interpreted as the formation of Cabo Ortegal pyroxenites. Corresponding values of initial 187Os/188Os (0.111–0.117) are relatively unradiogenic, suggesting limited contributions of slab-derived Os to primitive arc melts such as those parental to these pyroxenites. This interpretation is consistent with radiogenic Os in arc lavas being mostly related to crustal assimilation. Paleoproterozoic to Archean Os model ages confirm that Cabo Ortegal pyroxenites record incipient volcanic arc magmatism on the continental margin of the Western African Craton, as notably documented by zircon UPb ages of 2.1 and 2.7 Ga. LuHf data collected on clinopyroxene and amphibole separates and whole-rock samples are characterized by uncorrelated 176Lu/177Hf and 176Hf/177Hf (0.2822–0.2855), decoupled from Nd-isotope compositions. This decoupling is ascribed to diffusional disequilibrium during melt-peridotite interaction, in good agreement with the results of percolation-diffusion models simulating the interaction of an arc melt with an ancient melt-depleted residue. These models notably show that HfNd isotopic decoupling such as recorded by Cabo Ortegal pyroxenites and peridotites (ΔƐHf(i) up to +97) is enhanced during melt-peridotite interaction by slow diffusional re-equilibration and can be relatively insensitive to chromatographic fractionation. Finally, we discuss the hypothesis that arc-continent interaction may provide preferential conditions for such isotopic decoupling and propose that its ubiquitous recognition in peridotites reflects the recycling of sub-arc mantle domains derived from ancient, reworked SCLM.