Seismic resolution of lowermost mantle structure

Abstract:

Earth's lowermost mantle is characterized by a seismic reflector at 200 to 300 kilometers above the core-mantle boundary. This so-called D'' discontinuity may be caused by a phase transition from perovskite to post-perovskite, by a sudden onset in seismic anisotropy, by an accumulation of scatterers, by remnants of subducted slabs or dense primitive materials, or may bear witness to chemical reactions with core material. We discuss these possible interpretations of D'' structure in the light of array seismologically attainable observables and their accuracies. Synthetic modeling of the seismic response of a wide range of lowermost mantle models shows e.g. that the maximum observable thickness of a second-order discontinuity corresponds to the maximum thickness of the D'' discontinuity observed in previous studies. Thus, the observation of sharp D'' reflectors cannot rule out the predicted wide phase transition in perovskite at lowermost mantle conditions. Synthetic results also show that steep topography of one sharp discontinuity may cause multiple apparent reflectors a different depths, possibly providing explanations for apparent double reflectors or complex structure near the edges of large low shear wave velocity provinces. Regarding topography, we propose an additional array seismological observable for measuring the lateral variability of discontinuity depths utilizing both traveltime and slowness measurements at regional seismic arrays. However, despite improving the seismic resolution of lowermost mantle structure, none of the initially proposed possible mechanisms that may cause the D" reflector can be falsified. Seismological studies may rely on further mineralphysical, geochemical and dynamical specification of the elastic parameters corresponding to the individual thermal and compositional range of each model including their respective accuracies, to reliably support or falsify or even discriminate the physical cause for the observed lowermost mantle complexity.