Key points are not available for this paper at this time.
The arrival times of compressional (P) and shear (S) waves from approximately 580 microearthquakes recorded by a temporary array in the Pamir‐Hindu Kush region in central Asia are used to deduce one‐ and three‐dimensional velocity structures of this region. The results for one‐dimensional structures imply that the Moho is at 70±5 depth. Also, there is a velocity reversal near 160 km depth, which is inferred to be the beginning of the low velocity zone. This reversal continues to depths of approximately 230 km. Below 230 km, velocities are somewhat higher than those of normal mantle at similar depths (9. 3 km/s versus 8. 4 km/s for P waves). The outstanding feature of the results for three‐dimensional velocity structures is a broad (>40 km), centrally located region with 8 to 10% lower velocities than those in the surrounding regions. This low velocity region envelopes the seismic zone at depths between 70 and 150 km. The region may actually extend beyond these depths, but the results for shallower and deeper structure lack sufficient resolution to decide. Several tests, using both hypothetical and real data, were performed to estimate the reliability of the three‐dimensional solutions. The results of these tests suggest that the inferred velocities are reasonably accurate representations of the average velocities in the blocks, although one must be cautious of the effects of averaging in interpreting the solution. The low velocity region is inferred to be a manifestation of substantial quantities of subducted continental crust. Therefore, while subduction has occurred in the Pamir‐Hindu Kush, the results of the three‐dimensional inversions suggest that continental rather than only oceanic lithosphere has been subducted to depths of at least 150 km. Appendices are available with the entire article on microfiche. Order from American Geophysical Union, 2000 Florida Avenue, N. W. , Washington, D. C. , 20009. Document J81‐013; 1. 00. Payment must accompany order.
S. W. Roecker (Wed,) studied this question.