Search

The stable continents have a puzzling structure. Recent seismological studies have revealed a marked drop in seismic velocity at middle-lithosphere depths, but a generally small velocity drop at the lithosphere-asthenosphere boundary. The mid-lithosphere discontinuity has previously been attributed to changes in composition and/or crystal alignment (anisotropy) caused by metasomatic alteration, as well as to partial melting and/or accretion of intruded materials that occurred after the formation of the continents. We show that these models cannot easily explain the global presence of a large seismic velocity drop in the middle lithosphere and a small velocity change at the lithosphere-asthenosphere boundary. These models are also difficult to reconcile with long-term continental stability and, in particular, observations of nearly depth-invariant ages of rocks in the continental lithosphere that do not support the notion of late alteration events. Instead, we propose an elastically accommodated grain-boundary sliding model that predicts a substantial velocity drop at the mid-lithosphere discontinuity and a weak seismic signal at the lithosphere-asthenosphere boundary, as observed, without invoking late-stage modifications to the lithosphere. In this model, the mid-lithosphere discontinuity is a general feature of the stable continents, the precise depth of which depends primarily on temperature and water content. Consequently, the depth of the mid-lithosphere discontinuity may provide clues to the evolution of continents. [ABSTRACT FROM AUTHOR]