INDEPTH geophysical and geological observations imply that a partially molten midcrustal layer exists beneath southern Tibet. This partially molten layer has been produced by crustal thickening and behaves as a fluid on the time scale of Himalayan deformation. It is confined on the south by the structurally imbricated Indian crust underlying the Tethyan and High Himalaya and is underlain, apparently, by a stiff Indian mantle lid. The results suggest that during Neogene time the underthrusting Indian crust has acted as a plunger, displacing the molten middle crust to the north while at the same time contributing to this layer by melting and ductile flow. Viewed broadly, the Neogene evolution of the Himalaya is essentially a record of the southward extrusion of the partially molten middle crust underlying southern Tibet.

The Cenozoic collision between the Indian and Asian continents formed the Tibetan plateau, beginning about 70 million years ago. Since this time, at least 1,400 km of convergence has been accommodated by a combination of underthrusting of Indian and Asian lithosphere, crustal shortening, horizontal extrusion and lithospheric delamination. Rocks exposed in the Himalaya show evidence of crustal melting and are thought to have been exhumed by rapid erosion and climatically forced crustal flow. Magnetotelluric data can be used to image subsurface electrical resistivity, a parameter sensitive to the presence of interconnected fluids in the host rock matrix, even at low volume fractions. Here we present magnetotelluric data from the Tibetan-Himalayan orogen from 77 degrees E to 92 degrees E, which show that low resistivity, interpreted as a partially molten layer, is present along at least 1,000 km of the southern margin of the Tibetan plateau. The inferred low viscosity of this layer is consistent with the development of climatically forced crustal flow in Southern Tibet.

Magnetotelluric exploration has shown that the middle and lower crust is anomalously conductive across most of the north-to-south width of the Tibetan plateau. The integrated conductivity (conductance) of the Tibetan crust ranges from 3000 to greater than 20,000 siemens. In contrast, stable continental regions typically exhibit conductances from 20 to 1000 siemens, averaging 100 siemens. Such pervasively high conductance suggests that partial melt and/or aqueous fluids are widespread within the Tibetan crust. In southern Tibet, the high-conductivity layer is at a depth of 15 to 20 kilometers and is probably due to partial melt and aqueous fluids in the crust. In northern Tibet, the conductive layer is at 30 to 40 kilometers and is due to partial melting. Zones of fluid may represent weaker areas that could accommodate deformation and lower crustal flow.

The lithosphere-asthenosphere boundary (LAB) is a first-order structural discontinuity that accommodates differential motion between tectonic plates and the underlying mantle. Although it is the most extensive type of plate boundary on the planet, its definitive detection, especially beneath cratons, is proving elusive. Different proxies are used to demarcate the LAB, depending on the nature of the measurement. Here we compare interpretations of the LAB beneath three The seismic LAB beneath cratons is typically regarded as the base of a high-velocity mantle lid, although some workers infer its location based on a distinct change in seismic anisotropy.Surface-wave inversion studies provide depth-constrained velocity models, but are relatively insensitive to the sharpness of the LAB. The S-receiver function method is a promising new seismic technique with complementary characteristics to surface-wave studies, since it is 2 sensitive to sharpness of the LAB but requires independent velocity information for accurate depth estimation. Magnetotelluric (MT) observations have, for many decades, imaged an "electrical asthenosphere" layer at depths beneath the continents consistent with seismic lowvelocity zones. This feature is most easily explained by the presence of a small amount of water in the asthenosphere, possibly inducing partial melt. Depth estimates based on various proxies considered here are similar, lending confidence that existing geophysical tools are effective for mapping the LAB beneath cratons.

[1] 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 the a priori information on model parameters, (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.[2] 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 on L(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 defining (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 our (m) with a L(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.