Slab Control on the Northeastern Edge of the Mid‐Pacific LLSVP Near Hawaii

Abstract

At the core‐mantle boundary, most observed ultralow velocity zones (ULVZs) cluster along the edges of the large low shear velocity provinces (LLSVPs) and provide key information on the composition, dynamics, and evolution of the lower mantle. However, their detailed structure near slab‐like structures beneath the mid‐Pacific remains particularly challenging because of the lack of station coverage. While most studies of ULVZs concentrate on SKS‐complexity, here we report on the multipathing of ScS, which expands the sampling for ULVZs. We find the strongest multipathing along a ULVZ patch located just south of Hawaii and the far northeastern edge of the LLSVP, in a zone ~200 km in width and extending 600 km southward. The anomalous ScS travel times and distorted Sdiff waveforms further reveal patches interrupted by observed enhanced D″ indicative of slab‐debris influence on the complexity of the northeastern boundary of the mid‐Pacific LLSVP.

Fine‐Scale Structure of Earth's Deep Mantle Resolved Through Statistical Analysis of Hawaiian Basalt Geochemistry

“…This is a result of differences in age and composition of the material, observations that are consistent with what is understood about the long‐lived thermochemical stability of the LLSVPs relative to the more efficient mixing of the convecting mantle outside of the LLSVPs (Ballmer et al., 2017; Garnero et al., 2016; McNamara et al., 2010). Geodynamic models have shown that the arrival of subducted slabs in the lower mantle can “sweep” mantle heterogeneities towards the LLSVPs, resulting in both the stacking of heterogeneous material around and perhaps into the LLSVPs (Bower et al., 2013; Dannberg & Gassmöller, 2018; Doucet et al., 2020; Garnero et al., 2016; Hassan et al., 2016; McNamara et al., 2010; Sun et al., 2019). The combination of this dynamic, lower mantle “sweeping” (e.g., Bower et al., 2013) and lower mantle wind against the highly viscous, dense, FeO‐rich dregs of a primordial magma ocean and recycled slab “graveyard” results in the concentration of mantle heterogeneity both into, and along, the edges of the LLSVPs, which we have resolved with our dataset of >800 high‐precision isotopic analyses from the Hawaiian Islands.…”

Lower Mantle Dynamics Perceived With 50 Years of Hindsight From Plate Tectonics

“…Such P waves propagating along the boundary can refract P wave energy into the core, and P waves from the core can be refracted from it back into the mantle over long distances along the boundary. P wave speeds in these ultralow‐speed layers are 10% lower than surrounding or overlying regions (e.g., Garnero & Helmberger, , ; Mori & Helmberger, ), and S wave speeds are 30% lower (e.g., Garnero et al, ; Helmberger et al, ), though recently, Sun et al () suggested an S wave speed anomalous low by only 15% but spread over a layer 80 km thick. Williams and Garnero () recognized that partial melt contributes to the markedly low speeds.…”

“…At the risk of failing to heed my own admonition, I ask: Does the dynamics of the lower mantle face a “chicken or egg” problem? Wolf et al () posed this well by noting “At one extreme, [the large topographic relief of the LLSVPs] might represent chemically dense passive piles, which are dynamically propped up by external convective stresses, while at the other, they could be free standing and internally convecting metastable piles, whose topography is a direct reflection of the thermophysical properties of the pile material.” Indeed, many treat sinking cold material beneath the subduction zones surrounding the Pacific as pushing hot material around, which requires ascent elsewhere and is manifested by upwelling comprising the LLSVPs (e.g., Bower et al, ; Davies et al, ; Gonnermann et al, ; Jellinek & Manga, ; Li et al, ; McNamara & Zhong, , ; Sun et al, ; Yao & Wen, ). Others, however, recognize that heating at the base of the mantle might lead to convection in which the bottom boundary layer rises and evolves into a quadrupole pattern that could become a long‐lasting stable feature (e.g., Busse, ; Forte & Mitrovica, ; Gurnis et al, ), and the surrounding ring of higher‐speed material might then simply mark the return flow.…”

Lateral Velocity Gradients in the African Lower Mantle Inferred From Slowness Space Observations of Multipathing

“…LLVP boundary studies using travel time residuals and waveforms are common and, from their observations, have estimated the gradients at the boundaries of LLVPs to range from 3% δV s per 50 km (0.0044 km s −1 km −1 ) (Ni et al., 2002) to 2% δV s per 300 km (0.00048 km s −1 km −1 ) (Ritsema et al., 1998) (See Table 1 for published estimates of African LLVP S wave velocity gradients). Combining travel time residuals, multipathing identification, and forward modeling to observe and infer the properties of structures is well established and has been applied to a variety of structures (Silver & Chan, 1986; Sun et al., 2019, 2010, 2017) and algorithms developed to identify multipathing automatically in the waveforms (Sun et al., 2009; Zhao et al., 2015). Although regional seismology studies only use the waveform to infer the effects of deep Earth structure on the wavefield, they do not analyze all information available such as the direction and inclination of the arrival.…”