SUMMARY We present a tomographic model of the crust, upper mantle and transition zone beneath the South Atlantic, South America and Africa. Taking advantage of the recent growth in broadband data sampling, we compute the model using waveform fits of over 1.2 million vertical-component seismograms, obtained with the automated multimode inversion of surface, S and multiple S waves. Each waveform provides a set of linear equations constraining perturbations with respect to a 3-D reference model within an approximate sensitivity volume. We then combine all equations into a large linear system and solve it for a 3-D model of S- and P-wave speeds and azimuthal anisotropy within the crust, upper mantle and uppermost lower mantle. In South America and Africa, our new model SA2019 reveals detailed structure of the lithosphere, with structure of the cratons within the continents much more complex than seen previously. In South America, lower seismic velocities underneath the transbrasilian lineament (TBL) separate the high-velocity anomalies beneath the Amazon Craton from those beneath the São Francisco and Paraná Cratons. We image the buried portions of the Amazon Craton, the thick cratonic lithosphere of the Paraná and Parnaíba Basins and an apparently cratonic block wedged between western Guyana and the slab to the west of it, unexposed at the surface. Thick cratonic lithosphere is absent under the Archean crust of the São Luis, Luis Álves and Rio de La Plata Cratons, next to the continental margin. The Guyana Highlands are underlain by low velocities, indicating hot asthenosphere. In the transition zone, we map the subduction of the Nazca Plate and the Chile Rise under Patagonia. Cratonic lithosphere beneath Africa is more fragmented than seen previously, with separate cratonic units observed within the West African and Congo Cratons, and with cratonic lithosphere absent beneath large portions of Archean crust. We image the lateral extent of the Niassa Craton, hypothesized previously and identify a new unit, the Cubango Craton, near the southeast boundary of the grater Congo Craton, with both of these smaller cratons unexposed at the surface. In the South Atlantic, the model reveals the patterns of interaction between the Mid-Atlantic Ridge (MAR) and the nearby hotspots. Low-velocity anomalies beneath major hotspots extend substantially deeper than those beneath the MAR. The Vema Hotspot, in particular, displays a pronounced low-velocity anomaly under the thick, high-velocity lithosphere of the Cape Basin. A strong low velocity anomaly also underlies the Cameroon Volcanic Line and its offshore extension, between Africa and the MAR. Subtracting the global, age-dependent VS averages from those in the South Atlantic Basins, we observe areas where the cooling lithosphere is locally hotter than average, corresponding to the location of the Tristan da Cunha, Vema and Trindade hotspots. Beneath the anomalously deep Argentine Basin, we image unusually thick, high-velocity lithosphere, which suggests that its anomalously great depth can be explained, at least to a large extent, by isostatic, negative lithospheric buoyancy.

[1] Equation-of-state (EOS) modeling, whereby the seismic properties of a specified thermochemical structure are constructed from mineral physics constraints, and compared with global seismic data, provides a potentially powerful tool for distinguishing between plausible mantle structures. However, previous such studies at lower mantle depths have been hampered by insufficient evaluation of mineral physics uncertainties, overestimation of seismic uncertainties, or biases in the type of seismic and/or mineral physics data used. This has led to a wide, often conflicting, variety of models being proposed for the average lower mantle structure. In this study, we perform a thorough reassessment of mineral physics and seismic data uncertainties. Uncertainties in both the type of EOS, and mineral elastic parameters, used are taken into account. From this analysis, it is evident that the seismic variability due to these uncertainties is predominantly controlled by only a small subset of the mineral parameters. Furthermore, although adiabatic pyrolite cannot be ruled out completely, it is problematic to explain seismic velocities and gradients at all depth intervals with such a structure, especially in the interval 1660-2000 km. We therefore consider a range of alternative thermal and chemical structures, and map out the sensitivity of average seismic velocities and gradients to deviations in temperature and composition. Compositional sensitivity is tested both in terms of plausible end-member compositions (e.g., MORB, chondrite), and via changes in each of the five major mantle oxides, SiO 2 , MgO, FeO, CaO, and Al 2 O 3 . Fe enrichment reduces both P and S velocities significantly, while Si enrichment (and Mg depletion) increases P and S velocities, with a larger increase in P than in S. Using purely thermal deviations from adiabatic pyrolite, it remains difficult to explain simultaneously all seismic observations. A superadiabatic temperature gradient does improve the seismic fit in the lowermost mantle, but should be accompanied by concurrent bulk chemistry changes. Our results suggest that the most plausible way to alter bulk chemistry in the lowermost mantle, simultaneously fitting density, bulk velocity and shear velocity constraints, is an increasing contribution of a hot, basalt-enriched component with depth.

No abstract

Understanding the enigmatic intraplate volcanism in the Tristan da Cunha region requires knowledge of the temperature of the lithosphere and asthenosphere beneath it. We measured phase‐velocity curves of Rayleigh waves using cross‐correlation of teleseismic seismograms from an array of ocean‐bottom seismometers around Tristan, constrained a region‐average, shear‐velocity structure, and inferred the temperature of the lithosphere and asthenosphere beneath the hotspot. The ocean‐bottom data set presented some challenges, which required data‐processing and measurement approaches different from those tuned for land‐based arrays of stations. Having derived a robust, phase‐velocity curve for the Tristan area, we inverted it for a shear wave velocity profile using a probabilistic (Markov chain Monte Carlo) approach. The model shows a pronounced low‐velocity anomaly from 70 to at least 120 km depth. VnormalS in the low velocity zone is 4.1–4.2 km/s, not as low as reported for Hawaii (∼4.0 km/s), which probably indicates a less pronounced thermal anomaly and, possibly, less partial melting. Petrological modeling shows that the seismic and bathymetry data are consistent with a moderately hot mantle (mantle potential temperature of 1,410–1,430°C, an excess of about 50–120°C compared to the global average) and a melt fraction smaller than 1%. Both purely seismic inversions and petrological modeling indicate a lithospheric thickness of 65–70 km, consistent with recent estimates from receiver functions. The presence of warmer‐than‐average asthenosphere beneath Tristan is consistent with a hot upwelling (plume) from the deep mantle. However, the excess temperature we determine is smaller than that reported for some other major hotspots, in particular Hawaii.

Seismic‐wave velocities offer essential constraints on the temperature, thickness, and composition of the lithosphere of cratons. We invert broadband, Rayleigh‐wave phase and Love‐wave phase velocities measured across the Kaapvaal Craton and Limpopo Belt for depth distributions of shear‐wave velocity and radial anisotropy, from the upper‐crust down to deep upper mantle. Our probabilistic, Bayesian inversion addresses model nonuniqueness by means of direct parameter‐space sampling. An increase in Vs between the Moho and 100–150 km depths occurs across the region and can be explained by the gradual emergence of garnet below 80 km, due to the spinel peridotite‐garnet peridotite transformation and due to the exsolution of garnet from mantle orthopyroxene. Lateral variations in this Vs gradient can provide new information on lateral compositional variations. Cold cratonic lithosphere is manifest in very high shear velocities, up to 4.8 km/s. The depth extent of the shear‐velocity anomaly and the inferred lithospheric thickness increase from ∼200 km beneath the central and southwestern Kaapvaal to ∼300 km beneath the Limpopo Belt. Curiously, surface elevation decreases monotonically with the increasing lithospheric thickness. The relationship between the lithospheric thickness and topography depends on the lithospheric composition and, with the crustal structure taken into account, our results imply that the bottom part of the Limpopo lithosphere (200–300 km) is weakly‐to‐moderately depleted (Mg# 89.7–90.8). Our results also show that the central‐southwestern Kaapvaal lithosphere is thinner than it was (according to kimberlites) 100–200 m.y. ago. It may have been thinned by the same mantle plume that, initially, triggered the kimberlite eruptions.