Within the last decade a new picture of the oceanic crust has emerged from advances in seismic experimental design, instrumentation, and analysis techniques. In this new picture, layer 2 is a region in which velocity increases rapidly with depth. While there is evidence of finer structure within layer 2, the exact nature of this structure is still poorly resolved. Layer 3 is much more homogeneous vertically than layer 2 and appears to have gentle vertical velocity gradients and occasional low-velocity zones in v•, and vs. Although it has been observed at several sites, the widespread existence of a high-velocity basal crustal layer is in doubt. The thickness of the crust-mantle transition has been observed to vary between 0 and 2 km from site to site, and even at fiat lying, uncomplicated sites, seismic evidence for lateral heterogeneities on a scale of a few kilometers can be found. This seismic picture of the crust is in good agreement with the seismic velocities of rocks from ophiolite complexes and is consistent with the theoretically expected behavior of seismic velocities in porous, water-saturated rocks at elevated pressures. A review of velocity results obtained from use of synthetic seismogram modeling techniques is given, and the types of synthetic techniques suitable for marine work are described. anic crust and comment on both its current limitations and its advantages over the classical thick layered view of the crust. As useful companions to this paper we suggest the works of Kennett [1977] and Lewis [1978].Our main points in this paper are as follows: 1. Although the travel time data from marine refraction experiments can generally be well explained by oceanic crustal velocity models consisting of a small number of homogeneous layers, this cannot be construed to imply the necessary existence of homogeneous layers, velocity interfaces, or reflectors within the crust.
Crustal velocity models in which velocity varies smoothly with depth in most depth regions generally explain seismic wave amplitude variations better than the classical thick homogeneously layered models.3. In oceanic 'layer 2,' seismic velocities increase quite rapidly with depth (velocity gradients are -1-2 s-'). Finer structure has been observed within layer 2, but it appears to be highly variable laterally. The layer 2-layer 3 transition may be a velocity discontinuity at some sites and a broad gradient zone at other sites. The finer structure of layer 2 is only • Now at U.S. Geological Survey, marginally within the ability of explosion seismology to resolve.
Oceanic 'layer 3' has more gentle vertical seismic velocity gradients (---0.1 s-') and possible low-velocity zones inboth compressional (P) and shear (S) wave velocity. While a high-velocity (vp = 7.2-7.7 km/s) basal crustal layer may exist at some sites, there is little clear evidence that such a layer is widespread throughout the oceans. 5. The width of the crust-mantle transition is quite variable._6. Velocity gradients within the crust are more naturally interpreted petrol...