Evidence of recycled basalt in mantle plumes: Receiver function observations of the X-discontinuity beneath hotspots

The X-discontinuity (230-350 km depth) has been observed across a wide range of tectonic settings predominantly using reflected seismic phases with broad, upper mantle Fresnel zones. Consequently, observation across many settings has led to the proposal of multiple causal mechanisms of the X-discontinuity, but an individual cause explaining all observations remains enigmatic. Reflected seismic phases are ill-suited to target specific regions of the Earth because their sensitivity is predominantly controlled by source-receiver distributions. Conversely, seismic discontinuities can be interrogated using P wave receiver functions (RFs) possessing short wavelength sensitivity (~100 km) to structures almost directly beneath the station. Below well-instrumented regions, targeted high spatial resolution RF observations of the X-discontinuity can help to constrain its properties and unique causal mechanism. Here we specifically focus on the X-discontinuity below hotspots globally.

We compute RFs at 28 hotspots and 6 cratons globally using P-to-s converted teleseismic phases between 30-90˚ epicentral distance. Stacking RFs in the depth and time-slowness domains reveals 15 robust observations of the X-discontinuity beneath hotspots between 244-344 km depth. A further 10 potential observations are found beneath hotspots, while null observations are limited to the 6 stacks beneath cratons. Elevated mantle potential temperatures expected beneath hotspots restrict observations of the X-discontinuity to two causes: the coesite to stishovite phase transition in eclogite or the presence of carbonated silicate melt, both requiring basalt enrichment in the upper mantle. Further, we find a correlation between X-discontinuity observations and the Large Low Velocity Provinces in the lowermost mantle, suggesting that mantle plumes may entrain significant volumes of recycled basalt from the lowermost mantle.