3D Density Research Articles

3D gravity inversion based on mixed-norm regularization in spherical coordinates with application to the lunar Moscoviense Basin

Three-dimensional gravity inversion has been widely used to infer density structures and tectonic movements of the earth and Moon. However, two problems (the nonuniqueness and low depth resolution) of current gravity inversion methods still exist and are not completely resolved, which affect the reliability of the inversion results and the corresponding geologic interpretation. To improve the depth resolution of the 3D gravity inversion, we propose an efficient inversion method in spherical coordinates based on a mixed smooth and focused regularization with depth-weighting functions. In the inversion, we also use the kernel matrix equivalence strategy and the fast kernel-vector multiplication method based on a fast Fourier transform in each iteration to increase computational efficiency. Two synthetic inversion examples indicate that the proposed method can recover more complex density structures compared with the often used smooth-constraint inversion approach. Finally, the proposed inversion approach is applied to the latest lunar gravitational field model GL1500E to study the 3D density structures of the Moscoviense Basin. The inverted results indicate that the density structure with high-positive anomalies is principally distributed at depths ranging from 5 to 70 km, forming a giant asymmetrical high-density structure inclining from the southwest to the northeast. We may conclude that the Moscoviense Basin was formed by the double impact, where the second collision occurred in the southwest of the basin center of the first collision.

Biochar-derived flower-like Co-Mo2C spheres/g-C3N4 photocatalyst: Engineering morphology configuration and electronic structure tuning

As a class of highly recalcitrant water contaminants, pharmaceuticals and personal care products (PPCPs) have raised considerable concerns. Here, we develop a biochar-derived flower-like Co-Mo2C sphere/g-C3N4 (CMCN) heterostructure photocatalyst for typical PPCPs degradation. The flower-like Co-Mo2C nanostructure is derived from biochar, an inexpensive and ubiquitous plant byproduct. The open-up nanoflake structure with sharp edges can induce the fast charge collection and separation. DFT calculations and experimental characterizations are conducted to fundamentally understand the enhanced intrinsic photocatalytic activity of the as-prepared CMCN photocatalysts, especially the role of Co atoms. The delocalized charge density of CMCN could be rational tailored via the increasing density of state and the difference in work functions between g-C3N4 and Co-Mo2C. The electron density of Mo 3d in CMCN is increased through integrating it with Co atoms, which possess numerous delocalized electrons. Compared with the Mo2C/g-C3N4 (MCN) bulk, the CMCN catalyst displays superior photocatalytic activity for the typical PPCPs degradation, mainly arising from the multifaceted open-up surfaces of Co-Mo2C nanoflower, the higher density of electron and more efficient charge transfer ability, as well as the prolonged life time of photogenerated charge carriers. This work provides a cheap and mild route for synthesizing Co-doped Mo2C nanoflower catalyst and expands the scope of constructing high active g-C3N4 based heterostructure.

Growth Dynamics of Ultrathin Films of Benzo[1,2-b:4,5-b’]dithiophene Derivatives on Au(111): A Photoelectron Spectroscopy Investigation

Ultrathin films of a stereoisomeric mixture of benzo[1,2-b:4,5-b']dithiophene derivatives were grown by thermal evaporation in vacuum on Au(111), and they were studied in situ by photoelectron spectroscopy. X-ray photons from a non-monochromatic Mg Kα conventional X-ray source and UV photons from a He I discharge lamp equipped with a linear polarizer were used. He I photoemission results were compared with density functional theory (DFT) calculations: density of states (DOS) and 3D molecular orbital density distribution. Au 4f, C 1s, O 1s, and S 2p core-level components suggest a surface rearrangement as a function of film nominal thickness, with the variation of the molecular orientation, from flat-laying at the initial deposition to tilted toward the surface normal at coverages exceeding 2 nm. Eventually, the DFT results were exploited in assigning of the valence band experimental structures. Moreover, polarization-dependent photoemission confirmed the tilted arrangement of the molecules, starting at 2 nm. A variation of the work function of 1.4 eV with respect to the clean substrate was measured, together with a valence band offset of 1.3 eV between the organic layer and gold.

Rotational Voxels Statistics Histogram for both real-valued and binary feature representations of 3D local shape

3D Local feature description is an active and fundamental task in 3D computer vision. However, most of the existing descriptors fail to simultaneously achieve satisfactory performance among descriptiveness, robustness, efficiency, and compactness. To address these limitations, we first propose a real-valued descriptor named Rotational Voxels Statistics Histogram (RoVo), which exploits the novel 3D multi-pose processing mechanism proposed in this paper to calculate the 3D voxel density distribution in different 3D poses. Moreover, through well-designed binary encoding algorithms, we conduct the seamless extension of the real-valued RoVo descriptor to three binary representations that have different performance characteristics. Extensive evaluation experiments validate the superiority of the real-valued and three binary RoVo descriptors concerning descriptiveness, robustness, and efficiency. Furthermore, the three binary RoVo descriptors extend the performance of high compactness. Lastly, we perform the experiments of 3D scene registration and 3D object recognition to intuitively present the effectiveness of the four proposed RoVo descriptors.

3D radiative transfer modelling and virial analysis of starless cores in the B10 region of the Taurus molecular cloud

ABSTRACT Low-mass stars like our Sun begin their evolution within cold (10 K) and dense (∼105 cm−3) cores of gas and dust. The physical structure of starless cores is best probed by thermal emission of dust grains. We present a high-resolution dust continuum study of the starless cores in the B10 region of the Taurus Molecular Cloud. New observations at 1.2 and 2.0 mm (12 and 18 arcsec resolution) with the NIKA2 instrument on the IRAM 30m have probed the inner regions of 14 low-mass starless cores. We perform sophisticated 3D radiative transfer modelling for each of these cores through the radiative transfer framework pandora, which utilizes RADMC-3D. Model best-fits constrain each cores’ central density, density slope, aspect ratio, opacity, and interstellar radiation field strength. These ‘typical’ cores in B10 span central densities from 5 × 104 to 1 × 106 cm−3, with a mean value of 2.6 × 105 cm−3. We find the dust opacity laws assumed in the 3D modelling, as well as the estimates from Herschel, have dust emissivity indices, β’s, on the lower end of the distribution constrained directly from the NIKA2 maps, which averages to β = 2.01 ± 0.48. From our 3D density structures and archival NH3 data, we perform a self-consistent virial analysis to assess each core’s stability. Ignoring magnetic field contributions, we find nine out of the 14 cores (64 per cent) are either in virial equilibrium or are bound by gravity and external pressure. To push the bounded cores back to equilibrium, an effective magnetic field difference of only ∼15 $\mu$G is needed.

Thiocarbohydrazone and Chalcone-Derived 3,4-Dihydropyrimidinethione as Lipid Peroxidation and Soybean Lipoxygenase Inhibitors.

The potential of the 4,6-diphenyl-3,4-dihydropyrimidine-2(1H)-thione (abbreviated as KKII5) and (E)-N'-benzylidenehydrazinecarbothiohydrazide (abbreviated as DKI5) compounds as possible drug leads is investigated. KKII5 and DKI5 are synthesized in high yield of up to 97%. Their structure, binding in the active site of the LOX-1 enzyme, and their toxicity are studied via joint experimental and computational methodologies. Specifically, the structure assignment and conformational analysis were achieved by applying homonuclear and heteronuclear 2D nuclear magnetic resonance (NMR) spectroscopy (2D-COSY, 2D-NOESY, 2D-HSQC, and 2D-HMBC) and density functional theory (DFT). The obtained DFT lowest energy conformers were in agreement with the NOE correlations observed in the 2D-NOESY spectra. Additionally, docking and molecular dynamics simulations were performed to discover their ability to bind and remain stabile in the active site of the LOX-1 enzyme. These in silico experiments and DFT calculations indicated favorable binding for the enzyme under study. The strongest binding energy, -9.60 kcal/mol, was observed for dihydropyrimidinethione KKII5 in the active site of LOX-1. ADMET calculations showed that the two molecules lack major toxicities and could serve as possible drug leads. The redox potential of the active center of LOX-1 with the binding molecules was calculated via DFT methodology. The results showed a significantly smaller energy attachment of 2.8 eV with KKII5 binding in comparison to DKI5. Thus, KKII5 enhanced the ability of the active center to receive electrons compared to DKI5. This is related to the stronger binding interaction of KKII5 relative to that of DK15 to LOX-1. The two very potent LOX-1 inhibitors exerted IC50 19 μΜ (KKII5) and 22.5 μΜ (DKI5). Furthermore, they both strongly inhibit lipid peroxidation, namely, 98% for KKII5 and 94% for DKI5.

Validation of a stereological method for estimating particle size and density from 2D projections with high accuracy.

Stereological methods for estimating the 3D particle size and density from 2D projections are essential to many research fields. These methods are, however, prone to errors arising from undetected particle profiles due to sectioning and limited resolution, known as 'lost caps'. A potential solution developed by Keiding, Jensen, and Ranek in 1972, which we refer to as the Keiding model, accounts for lost caps by quantifying the smallest detectable profile in terms of its limiting 'cap angle' (ϕ), a size-independent measure of a particle's distance from the section surface. However, this simple solution has not been widely adopted nor tested. Rather, model-independent design-based stereological methods, which do not explicitly account for lost caps, have come to the fore. Here, we provide the first experimental validation of the Keiding model by comparing the size and density of particles estimated from 2D projections with direct measurement from 3D EM reconstructions of the same tissue. We applied the Keiding model to estimate the size and density of somata, nuclei and vesicles in the cerebellum of mice and rats, where high packing density can be problematic for design-based methods. Our analysis reveals a Gaussian distribution for ϕ rather than a single value. Nevertheless, curve fits of the Keiding model to the 2D diameter distribution accurately estimate the mean ϕ and 3D diameter distribution. While systematic testing using simulations revealed an upper limit to determining ϕ, our analysis shows that estimated ϕ can be used to determine the 3D particle density from the 2D density under a wide range of conditions, and this method is potentially more accurate than minimum-size-based lost-cap corrections and disector methods. Our results show the Keiding model provides an efficient means of accurately estimating the size and density of particles from 2D projections even under conditions of a high density.

Made-to-measure modelling of globular clusters

ABSTRACT We present the first application of the made-to-measure method for modelling dynamical systems to globular clusters. Through the made-to-measure algorithm, the masses of individual particles within a model cluster are adjusted while the system evolves forward in time via a gravitational N-body code until the model cluster is able to reproduce select properties of an observed cluster. The method is first applied to observations of mock isotropic and anisotropic clusters while fitting against the cluster’s 3D or projected density profile, density weighted mean-squared velocity profile, or its density profile with individual mean-squared velocity profiles assuming $10 {{\ \rm per\ cent}}$ uncertainty in each of these observables. We find that a cluster’s 3D density profile can easily be reproduced by the made-to-measure method, with minor discrepancies in the outer regions if fitting against a cluster’s projected surface density or projected kinematic properties. If an observed cluster is anisotropic, only fitting against the cluster’s density profile and individual mean-squared velocity profiles will fully recover the full degree of anisotropy. Partial anisotropy can be recovered as long as two kinematic properties are included in the fit. We further apply the method to observations of the Galactic globular cluster M4 and generate a complete 6D representation of the cluster that reproduces observations of its surface density profile, mean-squared proper motion velocity profile, and mean-squared line of sight velocity profile. The M2M method predicts M4 is primarily isotropic with a mass of $9.2\pm 0.4\times 10^4\, \mathrm{ M}_{\odot }$ and a half-mass radius of 3.7 ± 0.1 pc.

Estimating concentrations of cracks and platelets from their traces in 2D cross-sections

The effect of cracks on the overall elastic and conductive properties is expressed in terms of the 3D crack density parameter. Similarly, in composites, the effect of thin platelets (perfectly rigid, in the context of elasticity, or perfect conductor/insulator, in the context of conductivity) on the said properties is expressed in terms of the same parameter. However, it may be difficult to estimate experimentally and it cannot, generally, be related to volume fractions. Its estimation from the 2D density of line traces of cracks/platelets observed on a surface (say, specimen face) is derived in a simple form – thus allowing estimates of the said properties directly from the traces on specimen faces. This 2D–3D relation is verified by direct numerical simulations. The related statistical aspects are discussed.

A three-dimensional chemical simulation with irregular density distributions of L1544

ABSTRACTThe dust continuum and molecular distributions observed on the sky plane always show irregular shapes of molecular clouds. However, it is hard to directly reproduce the observed distributions using symmetrical models. In this work, for the first time, we present a three-dimensional (3D) gas-grain chemical simulation using an irregular 3D density structure derived by the Abel inversion taking the starless core L1544 as an example. We found that most (∼70 per cent) of the observed features (molecular distributions, peak positions, and column density values of 16 species) can be reproduced directly. The previously reached conclusion of non-uniform illumination at C3H2 peak is confirmed as the contribution from the gas component with density of a few 104 cm−3 at distance of ∼3000–8000 au from the core centre along the line of sight.

2D-3D reconstruction of the proximal femur from DXA scans: Evaluation of the 3D-Shaper software.

Introduction: Osteoporosis is currently diagnosed based on areal bone mineral density (aBMD) computed from 2D DXA scans. However, aBMD is a limited surrogate for femoral strength since it does not account for 3D bone geometry and density distribution. QCT scans combined with finite element (FE) analysis can deliver improved femoral strength predictions. However, non-negligible radiation dose and high costs prevent a systematic usage of this technique for screening purposes. As an alternative, the 3D-Shaper software (3D-Shaper Medical, Spain) reconstructs the 3D shape and density distribution of the femur from 2D DXA scans. This approach could deliver a more accurate estimation of femoral strength than aBMD by using FE analysis on the reconstructed 3D DXA. Methods: Here we present the first independent evaluation of the software, using a dataset of 77 ex vivo femora. We extend a prior evaluation by including the density distribution differences, the spatial correlation of density values and an FE analysis. Yet, cortical thickness is left out of this evaluation, since the cortex is not resolved in our FE models. Results: We found an average surface distance of 1.16mm between 3D DXA and QCT images, which shows a good reconstruction of the bone geometry. Although BMD values obtained from 3D DXA and QCT correlated well (r 2 = 0.92), the 3D DXA BMD were systematically lower. The average BMD difference amounted to 64mg/cm3, more than one-third of the 3D DXA BMD. Furthermore, the low correlation (r 2 = 0.48) between density values of both images indicates a limited reconstruction of the 3D density distribution. FE results were in good agreement between QCT and 3D DXA images, with a high coefficient of determination (r 2 = 0.88). However, this correlation was not statistically different from a direct prediction by aBMD. Moreover, we found differences in the fracture patterns between the two image types. QCT-based FE analysis resulted mostly in femoral neck fractures and 3D DXA-based FE in subcapital or pertrochanteric fractures. Discussion: In conclusion, 3D-Shaper generates an altered BMD distribution compared to QCT but, after careful density calibration, shows an interesting potential for deriving a standardized femoral strength from a DXA scan.

Lithosphere Density Model of the Middle Urals Segment

For the Middle and North Urals segments and adjacent territories of the East European Platform and West Siberian Plate (within the graticular trapezoid with geographic coordinates 53–65° N, 48–72° E), ten deep seismic sounding (DSS) profiles were reprocessed using two-dimensional (2D) seismic tomography method, and gradient sections of crustal velocities were constructed in the form of grid functions. In the same format, density sections were constructed. The coefficients of density–velocity empirical correlation were calculated using the solution algorithm of 2D inverse problem of gravimetry. The method and calculation technology of a three-dimensional (3D) density distribution with tying to 2D data from reference seismic sections are embedded in the procedure of quantitative interpretation of potential fields with construction of 3D geophysical models. A stable solution of the 3D inverse problem of gravimetry is sought on the correctness set of the family of horizontal layers with 2D density distribution.

NMR and DFT studies of monounsaturated and ω-3 polyunsaturated free fatty acids in the liquid state reveal a novel atomistic structural model of DHA

Structures of low-energy conformers of caproleic acid (CA; 10:1 cis-9), oleic acid (OA; 18:1 cis-9), α-linolenic acid (ALA; 18:3 cis-9,12,15 ω-3), eicosapentaenoic acid (EPA; 20:5 cis-5,8,11,14,17 ω-3), and docosahexaenoic acid (DHA; 22:6 cis-4,7,10,13,16,19 ω-3) in the liquid state, based on detailed 1D NOE and density functional theory calculations of 1H NMR chemical shifts are presented. Transient 1D NOE experiments with variable mixing time showed significant through-space proximity of the CH2–COOH protons and the terminal CH3– groups. Variable temperature 1H NMR experiments revealed strong intermolecular centro-symmetric cyclic hydrogen bond interactions of the carboxylic groups of the monounsaturated oleic and caproleic acids (δ(COOH) ∼ 12.0–12.4 ppm), ALA and EPA (δ(COOH) ∼ 11.0 ppm). On the contrary, the carboxylic proton of DHA is strongly shielded with δ(COOH) ∼ 8.5 ppm. DFT calculations were interpreted in terms of aggregates of dimerized fatty acids with parallel and antiparallel interdigitated structures. The antiparallel arrangement was found to be in excellent agreement with experimental 1D NOE NMR data. For the dimeric DHA, a flip-flop process between a classical intermolecular centro-symmetric hydrogen bond through carboxylic groups and a novel intramolecular hydrogen bond between the carboxylic group and the terminal ω-3 double bond is demonstrated.

The upper-mantle density variation beneath the South Atlantic Ocean and its implications for hotspots and large igneous provinces

The South Mid-Atlantic Ridge (SMAR), the most typical slow-spreading ridge on the earth, is a significant area to study the oceanic crust-mantle evolution. We used gravity data and the S-wave velocity model to derive a 3D density model of the upper-mantle beneath the South Atlantic Ocean to better understand the relationship between ridges, hotspots, and large igneous provinces. The density variations in the upper-mantle are related to thermal and compositional changes. The segment near the Agulhas-Falkland Fracture Zone (AFFZ) is characterized by low-density but inconspicuous low-velocity anomalies reflecting the depleted composition. Its composition is different from the northern part of the SMAR. A low-density layer at 100 km that extends 200 km from the ridge to the Discovery hotspot and that disappears near the AFFZ, is also present. The higher degree of mantle melting beneath the segment near the AFFZ is most probably influenced by the residual plume materials from the Discovery hotspot. Deep low densities beneath the Rio Grande Rise (RGR) and the Walvis Ridge (WR) indicate a higher degree of melting and their formation is probably influenced by the Tristan da Cunha plume. When the RGR is driven away from the plume, its original hot materials cool down and present as cold and dense uppermost mantle deposits. As the plume is located in the African plate, the WR was continuously formed as a seamount chain by intraplate volcanism of mantle plume.

A Fast Forward and Inversion Strategy for Three-Dimensional Gravity Field

Obtaining a three-dimensional (3D) density distribution within a reasonable time is one of the most critical problems in gravity exploration. In this paper, we present an efficient 3D forward modeling and inversion method for gravity data. In forward modeling, the 3D model is discretized into multiple horizontal layers, with the gravity field at a point on the surface being the sum of the gravity fields from all layers. To calculate the gravity field from each horizontal layer, we use the fast Fourier transform (FFT) method and the Block Toeplitz with Toeplitz Blocks (BTTB) matrix, which dramatically reduces both the computation time and storage requirement. In the inversion, the observed gravity data are separated into multiple gravity components of different depths using the cutting separation method. An iterative method is used to adjust the model to fit the above gravity component for each cutting radius. The initial model is constructed from the transformation of gravity components. These methods were applied to both synthetic data and field data. The numerical simulation validated the proposed methods, and the inversion results of field data were consistent with information obtained from well logging. The computational time and memory usage were also reasonable.

Dislocations as natural quantum wires in diamond

We study the electronic properties of the glide set of dislocations in diamond from first principles using hybrid exchange correlation functionals and find that the atomic-scale dislocation core states give rise to a prototypical one-dimensional (1D) band structure, i.e., natural quantum wires. The position and character of the core states varies strongly with local structure, where mixed dislocations with dangling bonds exhibit a 1D metallic band with a characteristic 1D density of states ($1/\sqrt{E}$). This 1D Fermi gas is spatially localized to a single atomic diameter orbital chain along the dislocation core. When the dangling bonds within the core are reconstructed, the 1D metallic band disappears. In contrast, pure edge dislocations in diamond reveal a 1D semiconductor with a direct band gap of 3.21 eV. These calculations provide a possible explanation to the long-standing observation of a blue luminescence band correlated with dislocations in diamond. This opens the door to using dislocations as 1D quantum phases with functional (electronic and optical) properties arising from the atomic-scale core states.

A probabilistic approach for quantitative evaluation of fatigue cracks using nonlinear vibration responses

Fatigue cracks can cause nonlinearities in the vibration responses of cracked structural components. Nevertheless, there is a noticeable barrier to quantitatively evaluating the fatigue cracks using vibration responses owing to the difficulty in extracting crack-induced weak dynamic nonlinearities and quantifying probabilities of crack statuses. To overcome this barrier, this study proposes a probabilistic approach for quantitatively evaluating fatigue cracks. Crack-induced nonlinearities in the vibration responses are extracted by the singular spectrum analysis (SSA) and embodied in status features (SFs). The dispersed distributions of the SF clusters in the status space can visually characterize different crack statuses. In particular, the mechanism of the SSA for extracting crack-induced dynamic nonlinearities is interpreted in this study. In addition, the 3D probability density functions of the SFs are formed, based on which a status probability index is established to quantify the relative probabilities of different known crack statuses. Hereby, the statuses of fatigue cracks can be quantified in a probabilistic manner. The approach is experimentally validated on a steel cantilever beam that bears a fatigue crack, attached to which an electromagnetic shaker is used to generate transverse harmonic excitations. The excitation amplitudes are increased to produce increasing open extents of the fatigue crack. Simultaneously, steady-state velocity responses are acquired from the free end of the beam through non-contact measurement using a Doppler laser vibrometer. The results reveal that the crack statuses corresponding to different open extents are visually characterized and their relative probabilities are quantified.

Do-It-Yourself 5-Color 3D Printing of Molecular Orbitals and Electron Density Surfaces

This report outlines an approach for preparing 5-color 3D printed plastic models of molecular orbitals and electron density surfaces using a hobby-grade 3D printer. Instructions are provided for preparing 3D orbital and electron density surface (EDS) models using solid or mesh representations in ground state and transition state structures. We show that the information content of 3D orbital and surface models can be enhanced with text annotation, bisection, strut and dashed bond placement, and composite orbital-EDS model constructions. Example prints illustrate orbital concepts in organic and inorganic chemistry, such as the ethyl cation (σ–p hyperconjugation), methane (hybridization and C–H bonding), ethane (σ–σ* hyperconjugation in the staggered and eclipsed conformations), 2-hydroxytetrahydropyran (the anomeric effect/n−σ* hyperconjugation), the water dimer (hydrogen bonding/n−σ* overlap), ethylene, 1,3-butadiene, and benzene (π molecular orbitals), Re2Cl82–, and U2(COT)2 (metal–metal quadruple bonds). Transition state models illustrate orbital interactions in the SN2 reaction of cyanide with methyl, ethyl, and isopropyl chloride and in a hydroboration reaction of BH3 with propene. Composite EDS models of methyl and isopropyl chloride (for exploring their relative reactivities as SN2 electrophiles) and methylcyclohexane (visualizing 1,3-diaxial interactions) are described. Student perceptions of a multicolor 3D orbital print used in an introductory organic chemistry laboratory course are reported.

Near Real-Time 3D Reconstruction and Quality 3D Point Cloud for Time-Critical Construction Monitoring

Improving the rapidity of 3D reconstruction is vital for time-critical construction tasks such as progress monitoring and hazard detection, but the majority of construction studies in this area have focused on improving its quality. We applied a Direct Sparse Odometry with Loop Closure (LDSO)-based 3D reconstruction method, improving the existing algorithm and tuning its hyper-parameter settings, to achieve both near real-time operation and quality 3D point cloud simultaneously. When validated using a benchmark dataset, the proposed method showed notable improvement in 3D point cloud density, as well as loop closure robustness, compared to the original LDSO. In addition, we conducted a real field test to validate the tuned LDSO’s accuracy and speed at both object and site scales, where we demonstrated our method’s near real-time operation and capability to produce a quality 3D point cloud comparable to that of the existing method. The proposed method improves the accessibility of the 3D reconstruction technique, which in turn helps construction professionals monitor their jobsite safety and progress in a more efficient manner.

Density structure beneath the Rungwe volcanic province and surroundings, East Africa from shear-wave velocity perturbations constrained inversion of gravity data

Density perturbations in the subsurface are the main driver of mantle convection and can contribute to lithospheric deformation. However, in many places the density structure in the subsurface is poorly constrained. Most geodynamic models rely on simplified equations of state or use linear seismic velocity perturbations to density conversions. In this study, we investigate the density structure beneath the Rungwe Volcanic Province (RVP), which is the southernmost volcanic center in the Western Branch of the East African Rift (EAR). We use shear-wave velocity perturbations (dlnvs) as a reference model to perform constrained inversions of satellite gravity data centered on the RVP. We use the code jif3D with a dlnvs-density coupling criterion based on mutual information to generate a 3D density model beneath the RVP up to a depth of 660 km. Our results reveal a conspicuous negative density anomaly (∼−200 kg/m3) in the sublithospheric mantle (at depths ranging from ∼100 km to ∼250 km) beneath the central part of the Malawi Rift extending to the west, beneath the Niassa Craton, coincident with locations with positive shear-wave velocity perturbations (+7%). We calculate a 3D model of the velocity-to-density conversion factor (f) and find negative f-values beneath the Niassa Craton which suggests the observed negative density anomaly is mostly due to compositional variations. Apart from the Niassa Craton, there are generally positive f-values in the study area, which suggest dominance of temperature control on the density structure. Although the RVP generally shows negative density anomalies and positive f-values, at shallow depths (<120 km), f≈ 0, which suggests important contributions of both temperature and composition on the density structure possibly due to the presence of plume material. The negative buoyancy of the Niassa Craton contributes to its long stability, while constituting a barrier to the southward flow of plume material, thus restricting the southward continuation of magmatism in the Western Branch of the EAR. The presence of a negative-density anomaly where dlnvs are positive is incompatible with models based on the use of simple dlnvs to density conversion factors. These results have implications on how dlnvs models are converted to density perturbations.