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.

Seismic data from central Tibet have been combined to image the subsurface structure and understand the evolution of the collision of India and Eurasia. The 410- and 660-kilometer mantle discontinuities are sharply defined, implying a lack of a subducting slab beneath the plateau. The discontinuities appear slightly deeper beneath northern Tibet, implying that the average temperature of the mantle above the transition zone is about 300 degrees C hotter in the north than in the south. There is a prominent south-dipping converter in the uppermost mantle beneath northern Tibet that might represent the top of the Eurasian mantle lithosphere underthrusting the northern margin of the plateau.

Fault plane solutions and well‐determined focal depths of medium‐sized earthquakes, topography, and Landsat imagery in conjunction with seismicity maps, cross sections, and available geological information are used to investigate the present tectonics of the Himalayan continental collision zone. Most of the accurately located epicenters of events along the Himalayan arc (78°E–95°E) that occurred between 1961 and 1981 are concentrated in a narrow zone, about 50 km wide, lying between the northerly dipping Main Boundary (MBT) and Main Central (MCT) thrusts. Most of these events are located just south of the MCT. Though the epicenters of the events are, in general, well located, their depths as determined by teleseismic travel time data are very unreliable. Events with accurately determined depths obtained from identification of surface‐reflected phases define a simple, planar zone from about 10‐km and 20‐km depth, with an apparent dip of about 15°. This result is all the more remarkable considering that the events used were located along about an 1800‐km length of the Himalyan arc. Except for one, all available focal mechanisms of events within this zone indicate shallow ( ≲30°), north dipping thrusts. This shallow, north dipping zone apparently defines a part of the detachment that separates the underthrusting Indian plate from the Lesser Himalayan crustal block. The spatial extent and the geometry of this interplate thrust zone strongly indicate that the MBT and nearby subsidiary surface and blind thrusts, rather than the MCT, are currently the most active structures of the Himalayan arc. We suggest that the great Himalayan earthquakes (M>8) occur along the same detachment surface as defined by the thrust‐type, medium‐sized events. Events located to the south of the MBT and beneath the Ganges foredeep show normal faulting with T axes perpendicular to the Himalayan trend. The above results suggest that the Indian continental plate is underthrusting the Himalayan crustal blocks in a relatively coherent and simple geometry and that this geometry is not much different from that observed along oceanic subduction zones. The November 19, 1980, earthquake that occurred near the MCT (near 88.5°E) shows a predominantly strike‐slip focal mechanism. One of the nodal planes of this mechanism is transverse to the Himalayan structural grain, and moreover, this plane has a trend similar to that of the recently mapped Yadong‐Gulu rift in the Tethyan Himalaya and in southern Tibet just northeast of the earthquake. We interpret this predominantly left‐lateral, strike‐slip mechanism to indicate a possible genetic relationship between transverse structural features in the Underthrusting Indian plate (the Kishangang basement fault) and the upper Himalayan blocks and Tibet.

Besides the improved method we also use more data than Kind et al. [1996]. We have added teleseismic recordings of the dense wide-angle German Depth Profiling of Tibet and the Himalayas (GEDEPTH) deployment, which proved to be very successful because of the close spacing of these stations. We have also added data from the permanent broadband station Lhasa (LSA), permitting a laterally extended view into the lithosphere and upper mantle.The passive seismological part of INDEPTH II lasted from May until October 1994. Fifteen Reftek recording stations were operated in that time period. Nine stations were equipped with Guralp 3T broadband seismometers, and six with 1-Hz Mark L-4 seismometers. They were installed from the high Himalaya to approximately 150km north of the 27,491

Volcanism that occurs far from plate margins is di cult to explain with the current paradigm of plate tectonics. The Changbaishan volcanic complex, located on the border between China and North Korea, lies approximately 1,300 km away from the Japan Trench subduction zone and is unlikely to result from a mantle plume rising from a thermal boundary layer at the base of the mantle. Here we use seismic images and three-dimensional waveform modelling results obtained from the NECESSArray experiment to identify a slow, continuous seismic anomaly in the mantle beneath Changbaishan. The anomaly extends from just below 660 km depth to the surface beneath Changbaishan and occurs within a gap in the stagnant subducted Pacific Plate. We propose that the anomaly represents hot and buoyant sub-lithospheric mantle that has been entrained beneath the sinking lithosphere of the Pacific Plate and is now escaping through a gap in the subducting slab. We suggest that this subduction-induced upwelling process produces decompression melting that feeds the Changbaishan volcanoes. Subductioninduced upwelling may also explain back-arc volcanism observed at other subduction zones.