Using PP and SS precursors to Image Lithospheric and Upper-mantle Discontinuities Beneath the Line Islands in the Central Pacific Basin

We interpret PP precursors for upper mantle structure beneath the Line Islands (LI), north of French Polynesia. Petrological and geochemical observations, as well as SS precursors, are used to verify our interpretation. Although PP and SS bouncepoint precursors have not hitherto been effective for detailed imaging of the lithosphere, array processing of EarthScope Transportable Array data and simultaneous iterative deconvolution improves frequency content to 4Hz and affords a more detailed view of the upper mantle. Using PP precursors, the 520 is observable beneath the LI region and is shallowest directly beneath the islands. To the SW of the island chain, the 410 is apparent at 430 km, then shallows by 6 km toward the NE. We observe a negative phase underlying a relatively flat 90-km-deep positive phase representing the LAB (Lithosphere-Asthenosphere Boundary. Overlying this is a negative lens-shaped phase that reaches its greatest thickness just NE of the center of the LI, and pinches out with distance from the LI axis. This negative phase suggests either a lens of partial melt or depleted mantle due to a past melt. A strong arch-shaped positive phase (~60 km depth) overlies this lens; it appears to be breached at its apex, and is overlain by a negative phase (~40 km) that is relatively thin to the SW and thickens to the NE. We believe this breached arch to be the vestige of an ancestral LAB that fractured as it became a conduit for ancient melt entering the lithosphere. The mechanism for formation of the LI appears similar to that in continental rifts. Petrological and geochemical observations support this interpretation. The use of PP precursors to investigate upper mantle structure at this scale beneath oceanic crust is novel, as PP data have historically been severely low-pass filtered and PP-waves, which travel through water, require a correction for ocean-bottom response. To do so, we generate a synthetic response for water depth at each bouncepoint and then deconvolve it from the PP data. If the synthetic response is correct, the side effects should be sufficiently small that interpretation is not affected; otherwise, side lobes appear that would interfere with the shallow structural response. We are confident in our results using PP data, and in our correction for ocean-bottom response, as we do not observe side lobes in most locations, and we see shallow negative anomalies only where they make geological sense. We are currently processing SS data to validate our interpretation of PP-precursors for upper mantle structure beneath LI. SS data do not require a correction for ocean-bottom response, and since S-waves are shorter in wavelength than P-waves there is a possibility of greater resolution with SS data. Additionally, SS data are more sensitive to partial melt and could help distinguish anomalous regions due to partial melt from those due to compositional change.