[1] The Zagros of Iran form one of the youngest collisional orogenic belts on Earth. At shallow depths, shortening across the Zagros is accommodated by folding in the sediments, high-angle thrust faulting in the basement and thickening of the lower crust, but how shortening is accommodated by the lithospheric mantle has been uncertain largely because the upper mantle seismic structure has been poorly known. We map the lateral variations in upper mantle shear wave speed beneath this region using a large, multimode surface wave data set. The upper mantle is slow for most of the Middle East, but a high shear wave speed lid extending to $225 km depth exists beneath the Zagros. We use a T(V s , z) relation to convert the shear wave speed profiles to temperature profiles and fit these with geotherms to identify the base of the lithosphere. The upper mantle temperatures from the seismic model are consistent with temperatures derived from geochemical modeling. The lithosphere is less than $120 km thick over the region except for a thick lithospheric root beneath the Zagros, implying that shortening in the mantle is accommodated by lithospheric thickening. The composition of the volcanic rocks from above the area of the thickened lithosphere has depleted magma source regions with densities $60 kg m À3 less than the MORB source. Elsewhere in the Middle East the volcanic source regions have compositions and densities similar to that of MORB. The shortening across the Zagros is accommodated by lithospheric thickening but the cool thickened lithosphere has been stabilized from delamination by depletion.Components: 6600 words, 8 figures.

S U M M A R YWe present new observations of the frequency-dependent propagation efficiency of the seismic phase S n over the Tibetan Plateau. Our measurements are the ratio of the S n amplitude to the P g coda amplitude on Tibetan regional seismograms, and we map the lateral variation in the maximum and mean values of this ratio across the Plateau. Good path density and azimuthal coverage allow the area of S n blockage identified by previous studies to be better constrained. An important result is that at low frequencies (∼0.2 Hz), S n propagates efficiently over the entire Plateau, whereas at higher frequencies (∼1 Hz), S n is blocked for a region of the northern Plateau. As we take measurements at higher frequencies, we find that the southern boundary of the region of inefficient propagation migrates to the south. The observation that low frequency S n propagates efficiently across the whole Plateau suggests that at the length-scales sampled by these waves, the upper mantle has an overall positive shear velocity gradient, which could indicate that the high-velocity seismic lid, or seismic lithosphere, beneath Tibet is still intact, and has not delaminated as some researchers have previously proposed. The observation that high frequency S n does not propagate beneath northern Tibet suggests that at the shorter length-scales sampled by the higher frequency waves, the upper mantle has a negative shear velocity gradient. This could indicate that at shallow depths the sub-Moho upper mantle is hot and may contain some melt.

No abstract

BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles in the biological, ecological, and environmental sciences published by nonprofit societies, associations, museums, institutions, and presses.

No abstract