The composition of the subcontinental lithospheric mantle (SCLM) is broadly related to the tectonothermal age of the overlying crust, suggesting a secular change in SCLM-forming processes. Most estimated compositions of Archean SCLM, based on well-studied suites of xenoliths and xenocrysts, are depleted garnet lherzolites with high orthopyroxene/olivine. However, these compositions make it difficult to account for the high shear-wave velocities measured in the cores of large cratons, and predict deeper geoid anomalies and higher elevations than are observed in most cratons. Global and regional seismic tomography indicates that most cratonic xenolith suites represent material from the lower-velocity margins of lithospheric blocks. This implies that previous compositional estimates are strongly biased toward metasomatized material. We suggest that most Archean SCLM originally consisted of highly depleted dunites/harzburgites, similar to the Archean orogenic massifs of western Norway. Incorporation of such rocks in the cold upper parts of the cratonic SCLM satisfies the seismic and gravity data, suggesting that large volumes of these rocks are preserved in the cores of cratons, but are poorly sampled by volcanic rocks. The roots of most Proterozoic shields probably consist of refertilized Archean SCLM; the juvenile SCLM beneath Proterozoic and Phanerozoic mobile belts reflects only moderate depletion of Primitive Mantle compositions. Rather than a gradual evolution in SCLM-forming processes, we suggest a sharp dichotomy between Archean and younger tectonic regimes. The differences in buoyancy and viscosity between these two types of SCLM have played a major role in the construction, preservation and recycling of continental crust. If originally Archean SCLM is more widespread than currently recognized, models of crustal growth rates and recycling may need to be revised.

Traditional inversion techniques applied to the problem of characterizing the thermal and compositional structure of the upper mantle are not well suited to deal with the nonlinearity of the problem, the trade‐off between temperature and compositional effects on wave velocities, the nonuniqueness of the compositional space, and the dissimilar sensitivities of physical parameters to temperature and composition. Probabilistic inversions, on the other hand, offer a powerful formalism to cope with all these difficulties, while allowing for an adequate treatment of the intrinsic uncertainties associated with both data and physical theories. This paper presents a detailed analysis of the two most important elements controlling the outputs of probabilistic (Bayesian) inversions for temperature and composition of the Earth's mantle, namely thea prioriinformation on model parameters,ρ(m), and the likelihood function,L(m). The former is mainly controlled by our current understanding of lithosphere and mantle composition, while the latter conveys information on the observed data, their uncertainties, and the physical theories used to relate model parameters to observed data.The benefits of combining specific geophysical datasets (Rayleigh and Love dispersion curves, body wave tomography, magnetotelluric, geothermal, petrological, gravity, elevation, and geoid), and their effects onL(m), are demonstrated by analyzing their individual and combined sensitivities to composition and temperature as well as their observational uncertainties. The dependence of bulk density, electrical conductivity, and seismic velocities to major‐element composition is systematically explored using Monte Carlo simulations. We show that the dominant source of uncertainty in the identification of compositional anomalies within the lithosphere is the intrinsic nonuniqueness in compositional space. A general strategy for definingρ(m) is proposed based on statistical analyses of a large database of natural mantle samples collected from different tectonic settings (xenoliths, abyssal peridotites, ophiolite samples, etc.). This strategy relaxes more typical and restrictive assumptions such as the use of local/limited xenolith data or compositional regionalizations based on age‐composition relations. We demonstrate that the combination of ourρ(m) with aL(m) that exploits the differential sensitivities of specific geophysical observables provides a general and robust inference platform to address the thermochemical structure of the lithosphere and sublithospheric upper mantle. An accompanying paper deals with the integration of these two functions into a general 3‐D multiobservable Bayesian inversion method and its computational implementation.

Large peridotite massifs are scattered along the 1500 km length of the Yarlung–Zangbo Suture Zone (southern Tibet, China), the major suture between Asia and Greater India. Diamonds occur in the peridotites and chromitites of several massifs, together with an extensive suite of trace phases that indicate extremely low fO₂ (SiC, nitrides, carbides, native elements) and/or ultrahigh pressures (UHP) (diamond, TiO₂ II, coesite, possible stishovite). New physical and isotopic (C, N) studies of the diamonds indicate that they are natural, crystallized in a disequilibrium, high-T environment, and spent only a short time at mantle temperatures before exhumation and cooling. These constraints are difficult to reconcile with previous models for the history of the diamond-bearing rocks. Possible evidence for metamorphism in or near the upper part of the Transition Zone includes the following: (1) chromite (in disseminated, nodular and massive chromitites) containing exsolved pyroxenes and coesite, suggesting inversion from a high-P polymorph of chromite; (2) microstructural studies suggesting that the chromitites recrystallized from fine-grained, highly deformed mixtures of wadsleyite and an octahedral polymorph of chromite; (3) a new cubic Mg-silicate, with the space group of ringwoodite but an inverse-spinel structure (all Si in octahedral coordination); (4) harzburgites with coarsely vermicular symplectites of opx + Cr–Al spinel ± cpx; reconstructions suggest that these are the breakdown products of majoritic garnets, with estimated minimum pressures to > 13 GPa. Evidence for a shallow pre-metamorphic origin for the chromitites and peridotites includes the following: (1) trace-element data showing that the chromitites are typical of suprasubduction-zone (SSZ) chromitites formed by magma mixing or mingling, consistent with Hf-isotope data from magmatic (375 Ma) zircons in the chromitites; (2) the composition of the new cubic Mg-silicate, which suggests a low-P origin as antigorite, subsequently dehydrated; (3) the peridotites themselves, which carry the trace element signature of metasomatism in an SSZ environment, a signature that must have been imposed before the incorporation of the UHP and low-fO₂ phases. A proposed P–T–t path involves the original formation of chromitites in mantle-wedge harzburgites, subduction of these harzburgites at c. 375 Ma, residence in the upper Transition Zone for >200 Myr, and rapid exhumation at c. 170–150 Ma or 130–120 Ma. Os-isotope data suggest that the subducted mantle consisted of previously depleted subcontinental lithosphere, dragged down by a subducting oceanic slab. Thermomechanical modeling shows that roll-back of a (much later) subducting slab would produce a high-velocity channelized upwelling that could exhume the buoyant harzburgites (and their chromitites) from the Transition Zone in < 10 Myr. This rapid upwelling, which may explain some characteristics of the diamonds, appears to have brought some massifs to the surface in forearc or back-arc basins, where they provided a basement for oceanic crust. This model can reconcile many apparently contradictory petrological and geological datasets. It also defines an important, previously unrecognized geodynamic process that may have operated along other large suture zones such as the Urals.