Abstract.Phase transformations in mantle mineralogies probably cause the transition zone seismic discontinuities, nominally at 410, 520, and 660 km depth. , 1923]. This paper focuses on the mantle discontinuities in the transition zone, the region between about 400 and 800 km depth. Whereas in seismology's nascency new discontinuity discoveries progressively refined our basic knowledge of Earth structure, in its maturity the discontinuities themselves become investigative tools rather than investigative targets. The goal of this review is to show how changes in discontinuity depth have been and can be used to investigate the mantle's properties. The argument develops along two lines. The first establishes what the major discontinuities at 410, 520, and 660 km depth are. Significantly, the topography on the discontinuities gives strong evidence to discriminate among different ideas about how they arise. The second strand describes the sort of scientific craftsmanship that can be achieved using the spatial variation of discontinuity depths as a tool, a tool that turns out to be surprisingly versatile, under continual refinement, and increasingly common in a research seismologist's tool kit.
Detecting DiscontinuitiesSeismic discontinuities are detected by the additional arrivals they cause in the seismic wave field by reflection and refraction of the direct wave at the discontinuity. Reflection typically leads to an earlier arrival than would be expected in a uniform medium. Its timing relative to the direct arrival yields the discontinuity's depth. A refraction usually leads to more than one arrival, where only one would be expected in a uniform medium. In this case, one deduces the discontinuity depth from the distance from the seismic source where the multiple arrivals appear. The Earth's major discontinuities were initially discovered this way, and others continue to be discovered like this [Oldham, 1906;Lehmann, 1936;Song and Helmberger, 1998].There are quite diverse methods for measuring discontinuity depths, probably best described through sketches (Figure 2) (see Shearer [1991] for more exotic ones). One way to characterize the methods is by the position where the seismic wave interacts with the discontinuity: near the source, near the receiver, or elsewhere. Near-source studies provide the best spatial resolution because there is little opportunity for the seismic wave field to spread out before interacting with the discontinuity (its Fresnel zone width is small). Spatial resolution for near-receiver studies is worse since the interaction is at least 410 km from the receiver. Interac-