Understanding Mars' deep interior is essential to reconstruct its geological history, thermal evolution, and present-day dynamics. To this end, the NASA InSight mission has provided unprecedented seismic observations. However, strong trade-offs between temperature and composition in seismic interpretations continue to limit our ability to resolve interior models. To address this challenge, we account for electromagnetic induction data from Mars Global Surveyor as an additional, independent constraint. We develop a joint probabilistic inversion framework that simultaneously fits seismic body wave arrival times, electrical conductivity, the Love number, and the moment of inertia. A key feature of our approach is the integration of Mars' long-term thermal evolution within the forward model, along with mineral physics and petrology data, to better constrain geodynamical parameters. We explore three different mantle compositions (Sanloup et al., 1999, https://doi.org/10.1016/s0031-9201(98)00175-7; Taylor, 2013, https://doi.org/10.1016/j.chemer.2013.09.006; Yoshizaki & McDonough, 2020, https://doi.org/10.1016/j.gca.2020.01.011) and consider both radially homogeneous and heterogeneous (with a basal molten layer (Samuel et al., 2023, https://doi.org/10.1038/s41586-023-06601-8)) mantle scenarios. For homogeneous mantle models, two families of solutions emerge regardless of the bulk composition: one with low Mg content and high potential temperature, which better reproduces electrical conductivity data due to a thicker lithosphere, and another with high Mg content and lower potential temperature. Models with a heterogeneous mantle reproduce electrical conductivity data less accurately, due to thinner lithospheres, and the mantle composition of Yoshizaki and McDonough (2020, https://doi.org/10.1016/j.gca.2020.01.011) appears to be less consistent with the full data set. To further refine models of Mars' interior, future efforts should focus on acquiring electromagnetic data with reduced uncertainties and seismically constraining more precisely the depth of mantle discontinuities associated with mineral phase transitions.

Abstract Mercury's unique interior structure and magnetic field generation remain to be fully understood. We construct models to further constrain Mercury's interior and test the hypothesis that iron snow within the liquid core drives the dynamo. We build upon previous models by incorporating an updated iron‐sulfur‐silicon (Fe‐S‐Si) core alloy composition and use a Monte Carlo approach to explore the parameter space consistent with geodetic and geophysical constraints. A high normalized moment of inertia (MoI) of , in combination with thermal and geochemical constraints, favors models with an Earth‐like mantle density of kg/, an inner core km in radius, and a core silicon content of at least 6 wt%. In contrast, a lower value of MoI favors models with lower mantle densities of kg/, an inner core radius in the range of 850–1,450 km, and a core silicon content less than 8 wt%. We also show that the formation of iron snow requires a sulfur concentration greater than wt%. However, the expected geochemistry of the core restricts the sulfur content to less than 2 wt%. This inconsistency suggests the absence of snow layers in Mercury's present‐day core and that its dynamo is not driven by sulfur‐induced iron crystallization.

Seismic data offer a unique window into planetary interiors. Here, following a note from Russell et al. (2026), we provide further analysis to solidify our identification of a seismic phase detected by the InSight mission as one which interacts with a basal mantle layer (BML) on Mars (Samuel et al., 2023a). We show that other mantle discontinuities or other seismic structures invoked in the aforementioned note are unlikely to produce this seismic signal. We further illustrate how our proposed layered structure is compatible with other available geophysical observations and geodynamical constraints, and can be used to make inferences about Martian evolution.

We present updated estimates of Basic Earth Parameters (BEP) from VLBI Celestial Pole Offset (CPO) time series spanning 1980-2025 using an ensemble Markov Chain Monte Carlo (MCMC) Bayesian inversion. Building upon , we incorporate recent advances in ocean tidal modeling and update several aspects of the algorithm. Key improvements include: (1) implementation of a cubic spline representation for Free Core Nutation (FCN) amplitude variations, which significantly reduces multimodality in FCN-related parameter in MCMC sampling compared to a linear representation; (2) integration of updated Ocean Tidal Angular Momentum (OTAM) values from FES2014 ocean tidal atlas , without the empirical 0.7 scaling factor previously applied in the construction of the last adopted nutation model MHB2000; and (3) utilization of five diverse CPO series from different analysis centers spanning up to 45 years of observations. Our estimated mean values of the parameters show good consistency across different CPO series, with values of the Earth's dynamical ellipticity at the edge of the 1 sigma range of MHB2000 . Notable findings include a larger absolute value for the imaginary part of the core-mantle boundary (CMB) coupling constant ( K CMB ), approaching the 2 sigma boundary of , which may reflect contributions from other coupling mechanisms in addition to electromagnetic coupling, including possible topographic coupling through "form drag" effect caused by wave interactions with irregular boundaries . The real part of the Inner Core Boundary (ICB) coupling constant ( K ICB ) is approximately half the MHB2000 value, potentially indicating the need to revisit hydrostatic assumptions for the inner core given recent seismic evidence of viscous deformation . Compliance estimates suggest that frequency extrapolation methods from seismic to nutation bands may be considered unchanged in the mantle, but should be revised at the ICB. The enhanced FCN free-mode modeling captures amplitude variations that differ from empirical models, particularly after 2000, although the physical interpretation of these differences requires further investigation. The systematic discrepancies across multiple parameters suggest that current nutation theory needs substantial updates to incorporate more realistic models of core-mantle coupling and inner-core-outer-core coupling.

Mercury's annual longitudinal libration (88 days) and its mean rotation rate have been determined based on independent observations from the ground-based radar (Margot et al., 2012), camera and/or laser altimetry (Stark et al., 2015; Bertone et al., 2021), and radio science (Mazarico et al., 2014; Genova et al., 2019; Konopliv et al., 2020). Although consistent, the precision of the libration measurements precludes identification of a large solid inner core (Van Hoolst et al., 2012). At the same time, the measured rotation rates are largely inconsistent. Deviation from the resonant rotation rate is caused by the planet-induced long-term librations which can be amplified if their periods are close to that of a free libration mode (Yseboodt et al., 2013).We devise an alternative and innovative approach aimed at precisely tracking how the rotation angle varies with time so that various libration terms can be analyzed quantitatively. The approach involves two self-registration processes of the MESSENGER Mercury Laser Altimeter (MLA) profiles (Xiao et al., 2024). We focus on a small polar region from 81°N to 84°N. In the first step, we carry out the self-registration by shifting the individual profiles laterally and radially to get rid of the slow-varying orbit, pointing, and timing errors, which can be treated as near-constant. In contrast to the aforementioned near-constant shifts, offsets in the rotation angles can lead to non-linear rotation-like distortions of the profiles. Offsets in the orientation angles of the spin axis can shift the profiles as a whole, ensuring that our approach is insensitive to the a priori orientation state. Then in the second step, we update the inertial coordinates of the profiles and perform the second self-registration in which adjustments are made to the rotation angles at the acquisition times of each of the profiles. However, as the periapsis of the spacecraft has drifted throughout the mission, the ground track does not exactly cross the North Pole and an offset in the rotation angle can also shift the centroid of the profile. In the light of this, the above two-step process needs to be iterated till convergence. Finally, we obtain the updated rotation angle per profile uncontaminated by external error sources.We have experimented with various a priori rotation and orientation values, i.e., Stark2015, IAU2015 (Archinal et al., 2018), Genova2019, and Bertone2021. An example of the obtained variation of the rotation with time is shown in Figure 1. The long-term libration most likely to be amplified and captured is that with a period of around 6 years, induced by Venus (5.66 y), or by Jupiter (5.93 y), or by the Earth (6.57 y). The superposition of multiple long-period terms is also possible. We will carry out close-loop simulations to assess uncertainty and consider interior and libration modelings to interpret the scientific implications.Figure 1: Rotation variation with time using the IAU2015 model as a priori values. Correction is with respect to Mercury’s resonant rotation.References:Archinal et al., 2018. Celest. Mech. Dyn. Astron.. Bertone et al., 2021. JGR. Mazarico et al., 2014. JGR. Genova et al., 2019. GRL. Konopliv et al., 2020. Icarus. Margot et al., 2012. JGR. Stark et al., 2015. GRL. Van Hoolst et al., 2012. EPSL. Xiao et al., 2024. Authorea Preprints. Yseboodt et al., 2013. Icarus.

Heat transfer through convection in Mercury's large core may be limited to a liquid layer between a solid inner core and a stably stratified outer liquid layer. Convection in the thin mantle may even have entirely stopped. Here, we consider the transition from convective to conductive heat transport in a coupled thermal evolution model of the mantle and core and assess implications for the generation of the magnetic field.We argue that a conductive temperature profile best describes the temperature in regions of the core with a subadiabatic heat flux. Implementing an adiabat in these regions in a model of the evolution of the core, as is often done, implicitly assumes the existence of a mechanism that transports heat downward. Such a mechanism not only consumes power that could otherwise be available for sustaining dynamo action, but is also unlikely to be effective.We show that a thermally convective layer deep in Mercury's liquid core below a thermally stratified layer is more likely to persist until present if light elements depress the liquidus of the core by several hundred degree compared to iron. Substantial partitioning of light elements into the liquid core can drive strong compositional convection in the upper part of Mercury's core, but this may not be in line with dynamo studies that are consistent with the observed magnetic field. Therefore, thermal evolution scenarios with light elements in the core that depress the core liquidus significantly but do not strongly fractionate into the core liquid are the most consistent with the present-day core dynamo.Present-day dynamo action below a thermally stratified layer does not necessarily imply that the mantle is currently convective. If the mantle has a high concentration of radiogenic elements and a low viscosity, it must be convecting, but mantle convection can have ended before the present for a more viscous mantle with low concentration of radiogenic elements.

Cooling and crystallization of Mercury's magma ocean likely formed a layered mantle composed of various proportions of minerals such as olivine, orthopyroxene, clinopyroxene, sulfides, and plagioclase, each with distinct thermal properties (e.g. thermal diffusivity, thermal conductivity, heat capacity, and melting temperature). Planetary thermal evolution models often consider an homogeneous mantle and treat these properties as constant or only varying with pressure and/or temperature. Their dependence on composition and modal proportions is usually neglected, but can have a large impact on the modeled evolution.Recent experimental studies gave access to the thermal conductivity and diffusivity of olivine, orthopyroxene and clinopyroxene. We calculated the thermal conductivity and diffusivity profiles of Mercury’s mantle assuming it is made of the Mg-rich endmembers forsterite, enstatite or diopside (i.e. the most likely phases occurring in the reduced interior of Mercury). We used a 1D parameterized model to simulate the thermal evolution of the planet with conductivity values varying from 1 to 4 Wm-1K-1, covering the above range of different mineralogies. We investigated several scenarios with (1) homogeneous conductivity over the whole mantle; (2) two layers characterized by different conductivity values. We then analyzed the results in terms of crust production and duration of mantle melting.At pressures and temperatures relevant for Mercury's mantle, enstatite and diopside have higher conductivities and diffusivities than forsterite. This has a direct impact on the thermal evolution of the planet and on the melting of a fertile layer. Indeed, the more conductive the mantle is, the shorter its melting duration. Therefore, a mantle characterized by the conductivity of enstatite or diopside would promote a shorter melting time than one with conductivity of forsterite. In a two-layer mantle, melting duration is lower when conductivity of the top layer is higher compared to the bottom layer. The melting duration would thus be shorter for a mantle with a refractory olivine-like mantle conductivity at the base and an enstatite- and diopside-bearing fertile mantle-like conductivity in the upper part of the stratigraphic column. Besides the thermal conductivity, other parameters such as solidus temperature and heat production rate will be taken into account to obtain a consistent picture of the influence of mineralogical-dependent parameters on Mercury's evolution.Accounting for variations in thermal conductivity and diffusivity due to heterogeneity in the mantle is therefore crucial in modeling planetary interiors. These factors significantly affect key parameters like crust thickness and the duration of volcanism.

Understanding the crystallization of metallic cores is necessary to constrain the structure and thermal evolution of terrestrial bodies in our solar system and beyond. Core cooling is also closely related to the generation and sustainability of a magnetic field. The core crystallization regime depends primarily on the depth of intersection of the core temperature with the liquidus ([1], and refs therein). Core composition, pressure, and thermal profile are the major parameters controlling the depth of intersection. If the temperature gradient across the core is steeper than that of the liquidus, solidification starts at the top, the “top-down” crystallization regime. At low pressure (≤10 GPa) relevant to small terrestrial planets, moons, and possibly some asteroids, the eutectic temperature decreases with increasing pressure (e.g., [2] for the Fe-S system), favoring an onset of crystallization at the top of the core. Top-down crystallization has been proposed to exist in several planets and moons in the Solar System, such as Mercury [2], [3], Mars ([4], [5]), and Ganymede [6], [7], [8], [9].In this study, which was performed by the International Space Science Institute (ISSI) Team “A new non-equilibrium model of iron snow in planetary cores”, we investigate the effect of non-equilibrium as well as the effect of the core composition on top-down crystallization. We find that the time scale of phase relaxation is significantly shorter than the time scales usually employed in one-dimensional evolution models. Consequently, the assumption of equilibrium in these models remains valid. Nevertheless, the time scales associated with crystallization, melting, and crystal settling may be similar to the phase relaxation time scale, which warrants a closer investigation. Additionally, if the amount of supercooling required to initiate nucleation is large [11], non equilibrium could play a much larger role. In terms of core chemistry we studied two different core alloys (Fe-S and Fe-C) motivated by silicate-metal partitioning experiments (reviewed by [12]) at various concentrations in the framework of the equilibrium top-down crystallization model. We find that the time scales of growing either the snow zone (iron-rich compositions) or the flotation crust (iron-poor compositions) can vary significantly between the Fe-S and Fe-C system. Furthermore, the exact concentration of sulfur or carbon has an impact on the thermodynamic parameters, subsequently affecting the entropy available to the dynamo.References:[1] Breuer et al., 2015. [2] Chen et al., 2008. [3] Dumberry & Rivoldini, 2015. [4] Stewart et al., 2007. [5] Davies & Pommier, 2018. [6] Hauck et al., 2006. [7] Christensen, 2015. [8] Rückriemen et al., 2015. [9] Rückriemen et al., 2018. [10] Loper, 1992. [11] Huguet et al., 2018. [12] Pommier et al., 2022.

Data from the MErcury Surface, Space ENvironment GEochemistry and Ranging (MESSENGER) spacecraft revealed that Mercury’s surface is volatile-rich and iron depleted. In particular the high sulfur concentration in surface lavas and their low iron content are indicative that Mercury formed under highly reducing conditions. Consequently, its core likely contains a significant amount of silicon along a smaller fraction of sulfur. Additionally, the low surface reflectance and spectral measurements support the presence of substantial amounts of carbon on its surface. Several lines of evidence indicate that Mercury was carbon-saturated early in its evolution and for this reason carbon might be abundant in its core.Unlike silicon, carbon and sulfur have a strong decreasing effect on the melting temperature of iron and as such, even small amounts of these elements imply a relatively low present-day core liquidus, affecting the inner core radius and magnetic field generation. The partitioning behaviour of light elements between solid inner- and liquid outer impact the density structure of the core and the gravitational coupling strength between the mantle and inner core. As a result, the amplitude of the longitudinal libration of Mercury can be affected by its core composition. Here we study the effect of the core composition on Mercury’s present-day thermal state and assess how its forced and free libration are affected.

The rotation rates of Ganymede and Callisto, the two largest satellites of Jupiter, are on average equal to their orbital mean motion but cannot be constant as a result of the varying gravitational torque exerted by Jupiter on the satellites. For a Keplerian orbit, the period of the torque and of the rotation variations is equal to the orbital period. Gravitational interaction with the other Galilean satellites and the Sun induces deviations from a purely Keplerian orbital motion, leading to changes in the gravitational torque of Jupiter on the satellites with respect to the mean Keplerian orbital motion and therefore to additional rotation variations. Here we discuss small variations from the average rotation on different time scales and assess the potential of using rotation as a probe of the interior structure. The ESA JUICE (JUpiter ICy moons Explorer) mission will measure the rotation and tides of Ganymede and Callisto in the early 30s, and will in particular very accurately determine those quantities for Ganymede during the orbital phase of the spacecraft around that satellite starting in 2032. We report on different theoretical aspects of the rotation for realistic models of the interior of the satellites, include tidal deformations and take into account the low-degree gravity field and topography of Ganymede and Callisto. We assess the advantages of a joint use of rotation and tides to constrain the satellite's interior structure, in particular its ice shell and ocean.

Abstract Using body wave arrival times from 31 seismic events recorded on Mars by the InSight mission, combined with topography and gravity field modeling, we constrained lateral variations of crustal thickness through a Bayesian inversion approach. The parameterization of the seismic structure relies on quantities that influence the thermochemical evolution of Mars, enabling the seismic velocities and densities in the different planetary envelopes to be consistently linked through common physical assumptions. Compared to a 1D structure, models with lateral variations of crustal thickness show two possible interpretations of the thermal evolution of Mars, with either a hot or cold scenario at the present‐day. We found the hot scenario to be more compatible with InSight's radiotracking data and the tidal Love number. We relocated the marsquakes and derived maps of seismicity recorded by InSight, which is mostly located along or North of the boundary between the Northern lowlands and the Southern highlands.

IntroductionIn light of the low oxygen fugacity inferred for Mercury, it is expected that Si is the dominant light constituent of Mercury’s core, potentially combined with smaller amounts of C and S. Depending on specific assumptions on Mercury’s oxidation state and bulk composition (for example, assuming a CB or EH chondrite-like bulk composition), geochemical estimates for the amount of Si, C, and S in Mercury’s core differ [1,2]. The concentration of light elements in Mercury’s core can be assessed independently from geochemical inferences using geophysical measurements. Geodetic constraints on the interior mass distribution of Mercury are derived from Mercury’s (close proximity to a) Cassini state 1 [3]. The compositional range of Mercury’s core for which the mass distribution of the planet is in agreement with geodetic measurements can be examined by planetary interior structure modelling [4-6]. Such interior structure modelling requires a composition-dependent parametrisation of the thermodynamic properties of metallic core alloys. Experimental measurements of density and of ternary Fe-Si-C liquids at high pressures have been lacking, although they are crucial to constrain the compositional range of Mercury’s core, and of bulk Mercury, using geodetic information of the planet.We experimentally measured the density and P-wave velocity (VP) of Fe-Si-C liquid metallic mixtures at high pressure, and investigated the feasible range of Mercury’s core composition in light of present geodetic constraints by interior structure modelling.MethodsDensity measurements of Fe-Si-C liquid metals are performed at high-pressure and high-temperature by X-ray absorption techniques with a Paris-Edinburgh (PE) press at beamline ID-27 of the European Synchrotron Radiation Facility (ESRF) in Grenoble, France. Binary Fe-Si sample powder is loaded into a cylindrical diamond (C) sample container, which saturates the sample in C. X-ray absorption profiles of the C-saturated Fe-Si-C metallic liquid sample are obtained in-situ, and of the quenched sample after the experiment ended and brought to ambient conditions. The density of the recovered samples is measured by hydrostatic weighing at the IPGP in Paris, France. The density of the liquid sample at experimental conditions is derived by combining the density of the quenched sample and the post-situ X-ray absorption profile with the in-situ X-ray absorption profile, using the Beer-Lambert law for X-ray absorption.Ultrasonic VP measurements of Fe-Si liquid metals are performed at high-pressure and high-temperature with a PE press at beamline 16-BMB of the Advanced Photon Source (APS) synchrotron facility, in Argonne, Illinois. Ultrasonic signals of 20 MHz, 25 MHz, and 30 MHz are generated and received in-situ by a lithium niobite transducer located at the top of the setup. Vertical wave reflections at the top and at the bottom of the sample are recorded and used to determine the vertical two-way travel time of the ultrasonic wave through the sample. The sample length is determined from radiographic images that are collected in situ. The ultrasonic sound speed (VP) through the sample is calculated using the sample length and the two-way travel time of sound waves.The samples that are successfully recovered from both the density experiments and the VP experiments are analysed by electron microprobe at the Institute for Mineralogy in Münster, Germany. A ternary ideal mixing model (FESIC) is fitted to the Fe-Si-C liquid metal density and data obtained by us and to data available in literature. This mixing model is implemented in spherically symmetric interior structure models of Mercury that are constrained by the normalized moment of inertia of Mercury I=0.346±0.014 MR2 and that of its outer solid shell Im=0.1475±0.006 MR2 [3]. The thermal profile is taken linear in the upper part (assumingly conductive) of the core, and adiabatic lower in the core. An Fe-Si-C solid inner core is considered where temperature is below the melting temperature of the liquid alloy. The mantle and an assumed 40 km thick curst are parametrized as single density spherical shells with fixed density ratio.ResultsFigure 1 compares the obtained density and VP measurements with the FESIC mixing model. The results confirm previous studies that VP of Fe-rich liquid metal increases and the density decreases with increasing concentration of Si. For the C-saturated density measurements, the decrease of C with increasing Si concentration confirms that the solubility of C in Fe-Si metals decreases with increasing concentration of Si. In the VP experiments, the sample was not loaded in a C-saturated environment. Nonetheless, we did detect a small amount of C in the sample with ~10wt%Si.The bulk-core compositions of interior structure models, constrained by the I and Im mentioned above, are compared to estimated core composition from by geochemical methods [1] from assuming a bulk composition of Mercury similar to an EH chondrite or CB chondrites (figure 2). The Si and C concentrations of Mercury’s core that are estimated from assuming a CB-like bulk composition of Mercury are consistent with the core composition of interior structure models with the central value of I (0.3458·MR2). A EH chondrite-like bulk composition of Mercury is thought to have a large core concentration of Si, and can only be met with an obliquity of Mercury at the upper end of the uncertainty limit, the planet is cold, and the inner core is large. The latter is inconsistent with the deep-dynamo explanation for Mercury’s broad-scale low-intensity magnetic field [7]. This study is published and openly accessible [8].AcknowledgementsWe acknowledge the Marie Sklodowska-Curie grant No 845354 awarded to JSK and the BRAIN-be program grant (BR/143/A2/COME-IN) to TVH. The X-ray absorption measurements were performed at the European Synchrotron Radiation Facility (ESRF), Grenoble, France. The APS is operated by Argonne National Laboratory under Contract No. DE-AC02-06CH11357.[1] Steenstra E. S. and W. van Westrenen (2020) Icarus, 340, 113621 [2] Vander Kaadden et al. (2020) JGR:Planets, 125(5), e2019JE006239 [3] Margot et al. (2012) JGR:Planets, 117(E12), E00L09 [4] Hauck et al. (2013), JGR:planets, 118(6), 1204-1220 [5] [Rivoldini A. and T. Van Hoolst (2013) Earth Planet. Sci. Lett., 377-378, 62-72 [6] Knibbe J. S. and W. van Westrenen (2015) JGR:Planets, 120(11), 1904-1923. [7] Christensen, U. (2006) nature, 444(7122), 1056-1058. [8] Knibbe et al. (2021), JGR:Planets, 126(1), e2020JE006651.

The InSight mission collected four years of seismic data from Mars, providing unprecedented data on the planet’s interior. Two seismic events – one marsquake and one meteorite impact – took place on the opposite side of Mars to InSight’s broadband seismometer [1]. These two far-side events were initially located using mantle-transiting PP and SS waves, but we demonstrate that their waveforms contain additional seismic arrivals which are sensitive to the properties of the Martian core. Using multiple seismic methods, we obtain SKS differential travel times for S0976a, a distant marsquake, and S1000a, a distant impact. SKS travels through Mars’ core as compressional waves and is therefore sensitive to its elastic properties. We use these differential travel times to build the first seismically informed models of Mars’ core [2]. The core velocity of Mars is low: 4.9-5.0 km/s at the CMB. We use our seismic results to estimate the fraction of light elements in Mars’ core, finding a high fraction of sulphur is needed, together with lesser amounts of oxygen, carbon and hydrogen. In addition to the SKS signals observed, the waveforms from S1000a contain evidence for a seismic wave which diffracts along a molten silicate layer at the base of the Martian mantle and also reflects from the core-mantle boundary. The presence of a basal silicate molten layer has implications for the areotherm, and could reduce both estimates of metallic core’s radius and its fraction of light elements [3]. The SKS detections reported provide information about the physical properties of the liquid core of Mars. A solid inner core at the centre of Mars is unlikely; we assess what structures inside the Martian core can be excluded based on existing data and models. [1] Horleston et al., 2022. Seism. Rec., 2(2), 88-99 [2] Irving et al., 2023. PNAS, 120, e2217090120 [3] Samuel et al., 2023, Nature, 622, 712–717

The detection of deep reflected S waves on Mars inferred a core size of 1,830 ± 40 km (ref. 1 ), requiring light-element contents that are incompatible with experimental petrological constraints. This estimate assumes a compositionally homogeneous Martian mantle, at odds with recent measurements of anomalously slow propagating P waves diffracted along the core–mantle boundary 2 . An alternative hypothesis is that Mars’s mantle is heterogeneous as a consequence of an early magma ocean that solidified to form a basal layer enriched in iron and heat-producing elements. Such enrichment results in the formation of a molten silicate layer above the core, overlain by a partially molten layer 3 . Here we show that this structure is compatible with all geophysical data, notably (1) deep reflected and diffracted mantle seismic phases, (2) weak shear attenuation at seismic frequency and (3) Mars’s dissipative nature at Phobos tides. The core size in this scenario is 1,650 ± 20 km, implying a density of 6.5 g cm −3 , 5–8% larger than previous seismic estimates, and can be explained by fewer, and less abundant, alloying light elements than previously required, in amounts compatible with experimental and cosmochemical constraints. Finally, the layered mantle structure requires external sources to generate the magnetic signatures recorded in Mars’s crust.

The local structure and density of ternary Fe-C-S liquid alloys have been studied using a combination of in situ X-ray diffraction and absorption experiments between 1 and 5 GPa and 1600-1900 K. The addition of up to 12 at% of carbon (C) to Fe-S liquid alloys does not significantly modify the structure, which is largely controlled by the perturbation to the Fe-Fe network induced by S atoms. The liquid density determined from diffraction and/or absorption techniques allows us to build a non-ideal ternary mixing model as a function of pressure, temperature, and composition in terms of the content of alloying light elements. The composition of the Moon's core is addressed based on this thermodynamic model. Under the assumption of a homogeneous liquid core proposed by two recent Moon models, the sulfur content would be 27-36 wt% or 12-23 wt%, respectively, while the carbon content is mainly limited by the Fe-C-S miscibility gap, with an upper bound of 4.3 wt%. On the other hand, if the core is partially molten, the core temperature is necessarily lower than 1850 K estimated in the text, and the composition of both the inner and outer core would be controlled by aspects of the Fe-C-S phase diagram not yet sufficiently constrained.

Knowledge of the interior structure and atmosphere of Mars is essential to understanding how the planet has formed and evolved. A major obstacle to investigations of planetary interiors, however, is that they are not directly accessible. Most of the geophysical data provide global information that cannot be separated into contributions from the core, the mantle and the crust. The NASA InSight mission changed this situation by providing high-quality seismic and lander radio science data 1 , 2 . Here we use the InSight’s radio science data to determine fundamental properties of the core, mantle and atmosphere of Mars. By precisely measuring the rotation of the planet, we detected a resonance with a normal mode that allowed us to characterize the core and mantle separately. For an entirely solid mantle, we found that the liquid core has a radius of 1,835 ± 55 km and a mean density of 5,955–6,290 kg m −3 , and that the increase in density at the core–mantle boundary is 1,690–2,110 kg m −3 . Our analysis of InSight’s radio tracking data argues against the existence of a solid inner core and reveals the shape of the core, indicating that there are internal mass anomalies deep within the mantle. We also find evidence of a slow acceleration in the Martian rotation rate, which could be the result of a long-term trend either in the internal dynamics of Mars or in its atmosphere and ice caps.

We present the first observations of seismic waves propagating through the core of Mars. These observations, made using seismic data collected by the InSight geophysical mission, have allowed us to construct the first seismically constrained models for the elastic properties of Mars' core. We observe core-transiting seismic phase SKS from two farside seismic events detected on Mars and measure the travel times of SKS relative to mantle traversing body waves. SKS travels through the core as a compressional wave, providing information about bulk modulus and density. We perform probabilistic inversions using the core-sensitive relative travel times together with gross geophysical data and travel times from other, more proximal, seismic events to seek the equation of state parameters that best describe the liquid iron-alloy core. Our inversions provide constraints on the velocities in Mars' core and are used to develop the first seismically based estimates of its composition. We show that models informed by our SKS data favor a somewhat smaller (median core radius = 1,780 to 1,810 km) and denser (core density = 6.2 to 6.3 g/cm3) core compared to previous estimates, with a P-wave velocity of 4.9 to 5.0 km/s at the core-mantle boundary, with the composition and structure of the mantle as a dominant source of uncertainty. We infer from our models that Mars' core contains a median of 20 to 22 wt% light alloying elements when we consider sulfur, oxygen, carbon, and hydrogen. These data can be used to inform models of planetary accretion, composition, and evolution.

Magnetic field observations from the MGS, MAVEN, and InSight missions reveal that a dynamo was active in Mars’s early history. One unique feature of Mars’s magnetic crustal field is its hemispheric dichotomy, where magnetic fields in the southern hemisphere are much stronger than those in the northern hemisphere. Here we use numerical dynamo simulations to investigate the potential hemispheric nature of Mars’s ancient dynamo. Previous studies show that a hemispheric heat flux perturbation at the core–mantle boundary could result in either a stable hemispherical magnetic field or a constantly reversing field, depending on choices of parameters used in those models. These two scenarios lead to different implications for the origin of crustal fields. Here we test the dynamo sensitivity to varying hemispheric heat flux perturbations at the core–mantle boundary in a broader parameter regime to understand whether a hemispheric dynamo is likely for early Mars. We find that features of the dynamo change from stable, hemispheric magnetic fields to reversing, hemispheric fields, with increasing hemispheric heat flux perturbations at the core–mantle boundary. We also find that magnetic fields powered by bottom heating are more stable and transition from a nonreversing, hemispheric magnetic field to a multipolar field at higher hemispheric heat flux perturbations, while the transition happens at a much lower heat flux perturbation for magnetic fields powered by internal heating.

FeO represents an important end-member for planetary interiors mineralogy. However, its properties in the liquid state under high pressure are poorly constrained. Here, in situ high-pressure and high-temperature X-ray diffraction experiments, ab initio simulations, and thermodynamic calculations are combined to study the local structure and density evolution of liquid FeO under extreme conditions. Our results highlight a strong shortening of the Fe-Fe distance, particularly pronounced between ambient pressure and similar to 40 GPa, possibly related with the insulator to metal transition occurring in solid FeO over a similar pressure range. Liquid density is smoothly evolving between 60 and 150 GPa from values calculated for magnetic liquid to those calculated for non-magnetic liquid, compatibly with a continuous spin crossover in liquid FeO. The present findings support the potential decorrelation between insulator/metal transition and the high-spin to low-spin continuous transition, and relate the changes in the microscopic structure with macroscopic properties, such as the closure of the Fe-FeO miscibility gap. Finally, these results are used to construct a parameterized thermal equation of state for liquid FeO providing densities up to pressure and temperature conditions expected at the Earth's core-mantle boundary.