520-km discontinuity

A less prominent feature within the mantle transition zone is a seismic discontinuity around a depth of 520 km (the 520). This discontinuity has been found in several long-period studies (Shearer, 1990, 1991; Flanagan and Shearer, 1998; Revenaugh and Jordan, 1991a). However, short period travel time studies (Cummins et al., 1992; Jones et al., 1992) and studies of short period precursors to pP (Vidale and Benz, 1992) show no evidence of a discontinuity at this depth.
The controversial discussion on the existence of this discontinuity even in long-period data is corroborated by Bock (1994) as an artefact of the processing. He shows that the signals interpreted as resulting from underside reflections from the 520 by Shearer (1990) can also be explained as an artifact of the processing of the long period data. The long period filtering produces sidelobes of the signals from the 410 and the 660-km discontinuity at the appropriate travel time of the 520. The different studies detecting the 520 show a very small velocity and/or density increase across the 520. Shearer (1990) estimates a shear impedance increase of $\sim$3% whereas Revenaugh and Jordan (1991a) find values smaller than 2% (i.e. less than 30% - 60% of the impedance contrast across the 410) which is close to the minimum resolution of the ScS reverberation method.
The phase transition of the $\beta$-spinel $\rightarrow$ $\gamma$-spinel form within the olivine component coincides with the pressure and temperature conditions at the 520. This transition leads to a gradual change of velocity and density with a thickness of $\sim$ 50 km (Rigden et al., 1991). The velocity and density changes are small ($\Delta$v$_p$ = 1% - 2%, $\Delta$v$_s$ = 0.8% - 1.5%, $\Delta\rho$ = 2.5% - 3.0%) but could account for the seismologically observed reflections.
Within the garnet component of the mantle material the transition to Ca-perovskite appears at a pressure of 18 GPa (Ita and Stixrude, 1992), which could partly explain this transition.
Aside from these mineralogical changes, Gasparik (1993) interprets the discontinuity as the result of chemical layering. He explains the 520 as the base of a garnet layer. This view is based mainly on geochemical arguments and differs significantly from the generally accepted view of a chemical largely homogeneous mantle.
Due to the sparse seismological evidence, the small impedance contrast of the mineralogical phase transition and the large depth interval where the phase transition occurs, the existence of the 520 in seismological data is still controversial.