Search

We extend the state-of-the-art N-body code PKDGRAV3 with the inclusion of mesh-free gas hydrodynamics for cosmological simulations. Two new hydrodynamic solvers have been implemented, the mesh-less finite volume and mesh-less finite mass methods. The solvers manifestly conserve mass, momentum and energy, and have been validated with a wide range of standard test simulations, including cosmological simulations. We also describe improvements to PKDGRAV3 that have been implemented for performing hydrodynamic simulations. These changes have been made with efficiency and modularity in mind, and provide a solid base for the implementation of the required modules for galaxy formation and evolution physics and future porting to GPUs. The code is released in a public repository, together with the documentation and all the test simulations presented in this work., 18 pages, 14 figures; accepted for publication in MNRAS

Hydrodynamic simulations provide a powerful, but computationally expensive, approach to study the interplay of dark matter and baryons in cosmological structure formation. Here we introduce the EMulating Baryonic EnRichment (EMBER) Deep Learning framework to predict baryon fields based on dark-matter-only simulations thereby reducing computational cost. EMBER comprises two network architectures, U-Net and Wasserstein Generative Adversarial Networks (WGANs), to predict two-dimensional gas and HI densities from dark matter fields. We design the conditional WGANs as stochastic emulators, such that multiple target fields can be sampled from the same dark matter input. For training we combine cosmological volume and zoom-in hydrodynamical simulations from the Feedback in Realistic Environments (FIRE) project to represent a large range of scales. Our fiducial WGAN model reproduces the gas and HI power spectra within 10% accuracy down to ~10 kpc scales. Furthermore, we investigate the capability of EMBER to predict high resolution baryon fields from low resolution dark matter inputs through upsampling techniques. As a practical application, we use this methodology to emulate high-resolution HI maps for a dark matter simulation of a L=100 Mpc/h comoving cosmological box. The gas content of dark matter haloes and the HI column density distributions predicted by EMBER agree well with results of large volume cosmological simulations and abundance matching models. Our method provides a computationally efficient, stochastic emulator for augmenting dark matter only simulations with physically consistent maps of baryon fields., 21 pages, 12 figures, accepted by MNRAS, comments welcome

We implement EuclidEmulator (version 1), an emulator for the non-linear correction of the matter power spectrum, into the MCMC forecasting code MontePython. We compare the performance of Halofit, HMCode, and EuclidEmulator1, both at the level of power spectrum prediction and at the level of posterior probability distributions of the cosmological parameters, for different cosmological models and different galaxy power spectrum wave number cut-offs. We confirm that the choice of the power spectrum predictor has a non-negligible effect on the computed sensitivities when doing cosmological parameter forecasting, even for a conservative wave number cut-off of $0.2\,h\,{\rm Mpc}^{-1}$. We find that EuclidEmulator1 is on average up to $17\%$ more sensitive to the cosmological parameters than the other two codes, with the most significant improvements being for the Hubble parameter of up to $42\%$ and the equation of state of dark energy of up to $26\%$, depending on the case. In addition, we point out that the choice of the power spectrum predictor contributes to the risk of computing a significantly biased mean cosmology when doing parameter estimations. For the four tested scenarios we find biases, averaged over the cosmological parameters, of between 0.5 and 2$\sigma$ (from below $1\sigma$ up to $6\sigma$ for individual parameters). This paper provides a proof of concept that this risk can be mitigated by taking a well-tailored theoretical uncertainty into account as this allows to reduce the bias by a factor of 2 to 5, depending on the case under consideration, while keeping posterior credibility contours small: the standard deviations are amplified by a factor of $\leq1.4$ in all cases., Comment: 22 pages, 14 figures

The origin of Uranus and Neptune remains a challenge for planet formation models. A potential explanation is that the planets formed from a population of a few planetary embryos with masses of a few Earth masses which formed beyond Saturn's orbit and migrated inwards. These embryos can collide and merge to form Uranus and Neptune. In this work we revisit this formation scenario and study the outcomes of such collisions using 3D hydrodynamical simulations. We investigate under what conditions the perfect-merging assumption is appropriate, and infer the planets' final masses, obliquities and rotation periods, as well as the presence of proto-satellite disks. We find that the total bound mass and obliquities of the planets formed in our simulations generally agree with N-body simulations therefore validating the perfect-merging assumption. The inferred obliquities, however, are typically different from those of Uranus and Neptune, and can be roughly matched only in a few cases. In addition, we find that in most cases the planets formed in this scenario rotate faster than Uranus and Neptune, close to break-up speed, and have massive disks. We therefore conclude that forming Uranus and Neptune in this scenario is challenging, and further research is required. We suggest that future planet formation models should aim to explain the various physical properties of the planets such as their masses, compositions, obliquities, rotation rates and satellite systems., 15 pages, 10 figures, submitted to MNRAS, comments welcome

Despite many similarities, there are significant observed differences between Uranus and Neptune: while Uranus is tilted and has a regular set of satellites, suggesting their accretion from a disk, Neptune's moons are irregular and are captured objects. In addition, Neptune seems to have an internal heat source, while Uranus is in equilibrium with solar insulation. Finally, structure models based on gravity data suggest that Uranus is more centrally condensed than Neptune. We perform a large suite of high resolution SPH simulations to investigate whether these differences can be explained by giant impacts. For Uranus, we find that an oblique impact can tilt its spin axis and eject enough material to create a disk where the regular satellites are formed. Some of the disks are massive and extended enough, and consist of enough rocky material to explain the formation of Uranus' regular satellites. For Neptune, we investigate whether a head-on collision could mix the interior, and lead to an adiabatic temperature profile, which may explain its larger flux and higher moment of inertia value. We find that massive and dense projectiles can penetrate towards the centre and deposit mass and energy in the deep interior, leading to a less centrally concentrated interior for Neptune. We conclude that the dichotomy between the ice giants can be explained by violent impacts after their formation., 18 pages, 15 figures, 1 table, accepted to MNRAS

Recent improvements to GPU hardware and the symplectic N-body code GENGA allow for unprecedented resolution in simulations of planet formation. In this paper, we report results from high-resolution N-body simulations of terrestrial planet formation that are mostly direct continuation of our previous 10 Myr simulations ( Woo et al., 2021a ) until 150 Myr. By assuming that Jupiter and Saturn have always maintained their current eccentric orbits (EJS), we are able to achieve a reasonably good match to the current inner solar system architecture. However, due to the strong radial mixing that occurs in the EJS scenario, it has difficulties in explaining the known isotopic differences between bodies in the inner solar system, most notably between Earth and Mars. On the other hand, assuming initially circular orbits for Jupiter and Saturn (CJS) can reproduce the observed low degree of radial mixing in the inner solar system, while failing to reproduce the current architecture of the inner solar system. These outcomes suggest a possible paradox between dynamical structure and cosmochemical data for the terrestrial planets within the classical formation scenario. In addition, we use our high-resolution simulations to study the assembly of Earth in unprecedented detail, focusing on its collisional history. We find that more than 90% of giant impacts (GIs) on Earth occur within 80 Myr in the EJS simulations, matching the possible early timing of the Moon-forming GI based on the ages recorded by various meteoritic samples. However, the CJS and EJS scenarios both result in a leftover planetesimal mass that is 1–2 orders of magnitude greater than what currently exists in the solar system. Our overall results strongly suggest the need to consider alternative initial conditions for terrestrial planet formation in the solar system. For example, it may be necessary to assume a lower initial mass in the asteroid belt in order to lower the subsequent effect of radial mixing in the EJS scenario, as well as reduce the mass of leftover planetesimals. Similarly, in the CJS scenario, including an early giant planet instability within 80 Myr could make the scenario dynamically feasible.

The formation of Uranus' regular moons has been suggested to be linked to the origin of its enormous spin axial tilt (~98o). A giant impact between proto-Uranus and a 2–3 MEarth impactor could lead to a large tilt and to the formation of an impact generated disc, where prograde and circular satellites are accreted. The most intriguing features of the current regular Uranian satellite system is that it possesses a positive trend in the mass-distance distribution and likely also in the bulk density, implying that viscous spreading of the disc after the giant impact plays a crucial role in shaping the architecture of the final system. In this paper, we investigate the formation of Uranus' satellites by combining results of SPH simulations for the giant impact, a 1D semi-analytic disc model for viscous spreading of the post-impact disc, and N-body simulations for the assembly of satellites from a disc of moonlets. Assuming the condensed rock (i.e., silicate) remains small and available to stick onto the relatively rapid growing condensed water-ice, we find that the best case in reproducing the observed mass and bulk composition of Uranus' satellite system is a pure-rocky impactor with 3 MEarth colliding with the young Uranus with an impact parameter b = 0.75. Such an oblique collision could also naturally explain Uranus' large tilt and possibly, its low internal heat flux. The giant impact scenario can naturally explain the key features of Uranus and its regular moons. We therefore suggest that the Uranian satellite system formed as a result of an impact rather than from a circumplanetary disc. Our results also suggest that objects beyond the water snow-line could be dominated by rocky objects similar to Pluto and Triton. Future missions to Uranus and its satellite system would further constrain the properties of Uranus and its moons and provide further insight on their formation processes., Icarus, 375, ISSN:0019-1035

Despite evidence for the existence of dark matter (DM) from very high and low redshifts, a moderate amount of DM particle decay remains a valid possibility. This includes both models with very long-lived yet unstable particles or mixed scenarios where only a small fraction of dark matter is allowed to decay. In this paper, we investigate how DM particles decaying into radiation affect non-linear structure formation. We look at the power spectrum and its redshift evolution, varying both the decay lifetime ($\tau$) and the fraction of decaying-to-total dark matter ($f$), and we propose a fitting function that reaches sub-percent precision below $k\sim10$ h/Mpc. Based on this fit, we perform a forecast analysis for a Euclid-like weak lensing (WL) survey, including both massive neutrino and baryonic feedback parameters. We find that with WL observations alone, it is possible to rule out decay lifetimes smaller than $\tau=75$ Gyr (at 95 percent CL) for the case that all DM is unstable. This constraint improves to $\tau=182$ Gyr if the WL data is combined with CMB priors from the Planck satellite and to $\tau=275$ Gyr if we further assume baryonic feedback to be fully constrained by upcoming Sunyaev-Zeldovich or X-ray data. The latter shows a factor of 3.2 improvement compared to constraints from CMB data alone. Regarding the scenario of a strongly decaying sub-component of dark matter with $\tau\sim 30$ Gyr or lower, it will be possible to rule out a decaying-to-total fraction of $f>0.49$, $f>0.21$, and $f>0.13$ (at the 95 percent CL) for the same three scenarios. We conclude that the upcoming stage-IV WL surveys will allow us to significantly improve current constraints on the stability of the dark matter sector., Comment: matching accepted version

Recent high precision meteoritic data improve constraints on the formation timescale and bulk composition of the terrestrial planets. High resolution N-body simulations allow direct comparison of embryo growth timescale and accretion zones to these constraints. In this paper, we present results of high resolution simulations for embryo formation from a disc of up to 41,000 fully-self gravitating planetesimals with the GPU-based N-body code GENGA. Our results indicate that the growth of embryos are highly dependent on the initial conditions. More massive initial planetesimals, a shorter gas disc decay timescale and initially eccentric Jupiter and Saturn (EJS) all lead to faster growth of embryos. Asteroid belt material can thereby be implanted into the terrestrial planet region via sweeping secular resonances. This could possibly explain the rapid growth of Mars within 10 Myr inferred from its Hf-W chronology. The sweeping secular resonance almost completely clears the asteroid belt and deposits this material in the Mercury-Venus region, altering the composition of embryos there. This could result in embryos in the Mercury-Venus region accreting an unexpectedly high mass fraction from beyond 2 AU. Changing the initial orbits of Jupiter and Saturn to more circular (CJS) or assuming embryos formed in a gas free environment removes the sweeping secular resonance effect and thus greatly decreases material accreted from beyond 2 AU for Mercury-Venus region embryos. We therefore propose that rock samples from Mercury and Venus could aid greatly in deducing the condition and lifetime of the initial protoplanetary gas disc during planetesimal and embryo formation, as well as the initial architecture of the giant planets., Accepted for publication in Icarus

Recent high-precision meteoritic data infers that Mars finished its accretion rapidly within 10 Myr of the beginning of the Solar System and had an accretion zone that did not entirely overlap with the Earth’s. Here we present a detailed study of the accretion zone of planetary embryos from high-resolution simulations of planetesimals in a disk. We found that all simulations with Jupiter and Saturn on their current eccentric orbits (EJS) result in a similar accretion zone between fast-forming Mars and Earth-region embryos. Assuming more circular orbits for Jupiter and Saturn (CJS), on the other hand, has a significantly higher chance of forming Mars with an accretion zone not entirely dominated by Earth and Venus-region embryos; however, CJS in general forms Mars slower than in EJS. By further quantifying the degree of overlap between accretion zones of embryos in different regions with the average overlap coefficient (OVL), we found that the OVL of CJS shows a better match with the OVL from a chondritic isotopic mixing model of Earth and Mars, which indicates that the giant planets are likely to have resided on more circular orbits during gas disk dissipation than they do today, matching their suggested pre-instability orbits. More samples, including those from Mercury and Venus, could potentially confirm this hypothesis.

We use a recently introduced statistic called {\em Integrated Bispectrum} (IB) to probe the gravity-induced non-Gaussianity at the level of the bispectrum from weak lensing convergence or $\kappa$ maps. We generalize the concept of the IB to spherical coordinates. This result is next connected to the response function approach. Finally, we use the Euclid Flagship simulations to compute the IB as a function of redshift and wave number. We also outline how the IB can be computed using a variety of analytical approaches including the ones based on Effective Field Theory (EFT), {\em Halo models} and models based on the {\em Separate Universe approach} in projection or two-dimension (2D). Comparing these results against simulations we find that the existing theoretical models tend to over-predict the numerical value of the IB. We emphasize the role of the finite volume effect in the numerical estimation of the IB. We introduced the concept of squeezed and collapsed tripsectrum for 2D $\kappa$ maps. We derive the IB for many parameterized theories of modified gravity including the Horndeskii and beyond-Horndeskii theories specifically for the non-degenerate scenarios that are also known as the Gleyzes-Langlois-Piazza-Venizzi or GPLV theories. In addition, the cosmological models with clustering quintessence and models involving massive neutrinos are also derived., Comment: 49 pages, 8 figures, To appear in JCAP

In the late stages of terrestrial planet formation, pairwise collisions between planetary-sized bodies act as the fundamental agent of planet growth. These collisions can lead to either growth or disruption of the bodies involved and are largely responsible for shaping the final characteristics of the planets. Despite their critical role in planet formation, an accurate treatment of collisions has yet to be realized. While semi-analytic methods have been proposed, they remain limited to a narrow set of post-impact properties and have only achieved relatively low accuracies. However, the rise of machine learning and access to increased computing power have enabled novel data-driven approaches. In this work, we show that data-driven emulation techniques are capable of classifying and predicting the outcome of collisions with high accuracy and are generalizable to any quantifiable post-impact quantity. In particular, we focus on the dataset requirements, training pipeline, and classification and regression performance for four distinct data-driven techniques from machine learning (ensemble methods and neural networks) and uncertainty quantification (Gaussian processes and polynomial chaos expansion). We compare these methods to existing analytic and semi-analytic methods. Such data-driven emulators are poised to replace the methods currently used in N-body simulations, while avoiding the cost of direct simulation. This work is based on a new set of 14,856 SPH simulations of pairwise collisions between rotating, differentiated bodies at all possible mutual orientations., Computational Astrophysics and Cosmology, 7 (1), ISSN:2197-7909

We present a new, updated version of the EuclidEmulator (called EuclidEmulator2), a fast and accurate predictor for the nonlinear correction of the matter power spectrum. Percent-level accurate emulation is now supported in the eight-dimensional parameter space of $w_0w_a$CDM$+\sum m_\nu$models between redshift $z=0$ and $z=3$ for spatial scales within the range 0.01 $h$/Mpc $\leq k \leq$ 10 $h$/Mpc. In order to achieve this level of accuracy, we have had to improve the quality of the underlying N-body simulations used as training data: (1) we use self-consistent linear evolution of non-dark matter species such as massive neutrinos, photons, dark energy and the metric field, (2) we perform the simulations in the so-called N-body gauge, which allows one to interpret the results in the framework of general relativity, (3) we run over 250 high-resolution simulations with $3000^3$ particles in boxes of 1 (Gpc/$h$)${}^3$ volumes based on paired-and-fixed initial conditions and (4) we provide a resolution correction that can be applied to emulated results as a post-processing step in order to drastically reduce systematic biases on small scales due to residual resolution effects in the simulations. We find that the inclusion of the dynamical dark energy parameter $w_a$ significantly increases the complexity and expense of creating the emulator. The high fidelity of EuclidEmulator2 is tested in various comparisons against N-body simulations as well as alternative fast predictors like Halofit, HMCode and CosmicEmu. A blind test is successfully performed against the Euclid Flagship v2.0 simulation. Nonlinear correction factors emulated with EuclidEmulator2 are accurate at the level of 1% or better for 0.01 $h$/Mpc $\leq k \leq$ 10 $h$/Mpc and $z\leq3$ compared to high-resolution dark matter only simulations. EuclidEmulator2 is publicly available at https://github.com/miknab/EuclidEmulator2 ., Comment: 28 pages, 19 figures, submitted to MNRAS

An accurate modelling of baryonic feedback effects is required to exploit the full potential of future weak-lensing surveys such as Euclid or LSST. In this second paper in a series of two, we combine Euclid-like mock data of the cosmic shear power spectrum with an eROSITA X-ray mock of the cluster gas fraction to run a combined likelihood analysis including both cosmological and baryonic parameters. Following the first paper of this series, the baryonic effects (based on the baryonic correction model of Schneider et al. 2019) are included in both the tomographic power spectrum and the covariance matrix. However, this time we assume the more realistic case of a $\Lambda$CDM cosmology with massive neutrinos, and we consider several extensions of the currently favoured cosmological model. For the standard $\Lambda$CDM case, we show that including X-ray data reduces the uncertainties on the sum of the neutrino mass by $\sim30$ percent, while there is only a mild improvement on other parameters such as $\Omega_m$ and $\sigma_8$. As extensions of $\Lambda$CDM, we consider the cases of a dynamical dark energy model (wCDM), a $f(R)$ gravity model (fRCDM), and a mixed dark matter model ($\Lambda$MDM) with both a cold and a warm/hot dark matter component. We find that combining weak lensing with X-ray data only leads to a mild improvement of the constraints on the additional parameters of wCDM, while the improvement is more substantial for both fRCDM and $\Lambda$MDM. Ignoring baryonic effects in the analysis pipeline leads to significant false-detections of either phantom dark energy or a light subdominant dark matter component. Overall we conclude that for all cosmologies considered, a general parametrisation of baryonic effects is both necessary and sufficient to obtain tight constraints on cosmological parameters., Comment: Accepted version (JCAP)

Baryonic feedback effects lead to a suppression of the weak lensing angular power spectrum on small scales. The poorly constrained shape and amplitude of this suppression is an important source of uncertainties for upcoming cosmological weak lensing surveys such as Euclid or LSST. In this first paper in a series of two, we use simulations to build a Euclid-like tomographic mock data-set for the cosmic shear power spectrum and the corresponding covariance matrix, which are both corrected for baryonic effects following the baryonification method of Schneider et al. (2019). In addition, we develop an emulator to obtain fast predictions of the baryonic power suppression, allowing us to perform a likelihood inference analysis for a standard $\Lambda$CDM cosmology with both cosmological and astrophysical parameters. Our main findings are the following: (i) ignoring baryonic effects leads to a greater than 5$\sigma$ bias on the cosmological parameters $\Omega_m$ and $\sigma_8$; (ii) restricting the analysis to the largest scales, that are mostly unaffected by baryons, makes the bias disappear, but results in a blow-up of the $\Omega_m$-$\sigma_8$ contour area by more than a factor of 10; (iii) ignoring baryonic effects on the covariance matrix does not significantly affect cosmological parameter estimates; (iv) while the baryonic suppression is mildly cosmology dependent, this effect does not noticeably modify the posterior contours. Overall, we conclude that including baryonic uncertainties in terms of nuisance parameters results in unbiased and surprisingly tight constraints on cosmology., Comment: Accepted version (JCAP)

Feedback processes from baryons are expected to strongly affect weak-lensing observables of current and future cosmological surveys. In this paper we present a new parametrisation of halo profiles based on gas, stellar, and dark matter density components. This parametrisation is used to modify outputs of gravity-only $N$-body simulations (following the prescription of Schneider and Teyssier [1]) in order to mimic baryonic effects on the matter density field. The resulting baryonic correction model relies on a few well motivated physical parameters and is able to reproduce the redshift zero clustering signal of hydrodynamical simulations at two percent accuracy below $k\sim10$ h/Mpc. A detailed study of the baryon suppression effects on the matter power spectrum and the weak lensing shear correlation reveals that the signal is dominated by two parameters describing the slope of the gas profile in haloes and the maximum radius of gas ejection. We show that these parameters can be constrained with the observed gas fraction of galaxy groups and clusters from X-ray data. Based on these observations we predict a beyond percent effect on the power spectrum above $k=0.2-1.0$ h/Mpc with a maximum suppression of 15-25 percent around $k\sim 10$ h/Mpc. As a result, the weak lensing angular shear power spectrum is suppressed by 15-25 percent at scales beyond $\ell\sim 100-600$ and the shear correlations $\xi_{+}$ and $\xi_{-}$ are affected at the 10-25 percent level below 5 and 50 arc-minutes, respectively. The relatively large uncertainties of these predictions are a result of the poorly known hydrostatic mass bias of current X-ray observations as well as the generic difficulty to observe the low density gas outside of haloes., Comment: Corrected typo in labelling of Fig 3, 13, and 14

We perform $N$-body simulations of the early phases of open cluster evolution including a large population of planetesimals, initially arranged in Kuiper-belt like discs around each star. Using a new, 4th-order and time-reversible $N$-body code on Graphics Processing Units (GPUs), we evolve the whole system under the stellar gravity, i.e. treating planetesimals as test particles, and consider two types of initial cluster models, similar to IC348 and the Hyades, respectively. In both cases, planetesimals can be dynamically excited, transferred between stars or liberated to become free-floating (such as A/2017 U1 or 'Oumuamua) during the early cluster evolution. We find that planetesimals captured from another star are not necessarily dynamically distinct from those native to a star. After an encounter both native and captured planetesimals can exhibit aligned periastrons, qualitatively similar to that seen in the Solar system and commonly thought to be the signature of Planet 9. We discuss the implications of our results for both our Solar system and exoplanetary systems., 15 pages, 7 figures, 4 tables, accepted for publication in MNRAS

We present a new power spectrum emulator named EuclidEmulator that estimates the nonlinear correction to the linear dark matter power spectrum. It is based on a spectral decomposition method called polynomial chaos expansion. All steps in the construction of the emulator have been tested and optimized: the large high-resolution N-body simulations carried out with PKDGRAV3 were validated using a simulation from the Euclid Flagship campaign and demonstrated to have converged up to wavenumbers $k\approx 5\,h\,{\rm Mpc}^{-1}$ for redshifts $z\leq 5$. The emulator is constructed using the uncertainty quantification software UQLab and it has been optimized first by creating mock emulators based on Takahashi's HALOFIT. We show that it is possible to successfully predict the performance of the final emulator in this way prior to performing any N-body simulations. We provide a C-code to calculate the nonlinear correction at a relative accuracy of $\sim0.3\%$ with respect to N-body simulations within 50 ms. The absolute accuracy of the final nonlinear power spectrum is comparable to one obtained with N-body simulations, i.e. $\sim 1\%$ for $k\lesssim 1\,h\,{\rm Mpc}^{-1}$ and $z\lesssim 3.5$. This enables efficient forward modeling in the nonlinear regime allowing for maximum likelihood estimation of cosmological parameters. EuclidEmulator has been compared to HALOFIT and CosmicEmu, an alternative emulator based on the Mira-Titan Universe, and shown to be more accurate than these other approaches. This work paves a new way for optimal construction of future emulators that also consider other cosmological observables, use higher resolution input simulations and investigate higher dimensional cosmological parameter spaces., 19 pages, 13 figures

The giant impact hypothesis for Moon formation successfully explains the dynamic properties of the Earth-Moon system but remains challenged by the similarity of isotopic fingerprints of the terrestrial and lunar mantles. Moreover, recent geochemical evidence suggests that the Earth's mantle preserves ancient (or "primordial") heterogeneity that predates the Moon-forming giant impact. Using a new hydrodynamical method, we here show that Moon-forming giant impacts lead to a stratified starting condition for the evolution of the terrestrial mantle. The upper layer of the Earth is compositionally similar to the disk, out of which the Moon evolves, whereas the lower layer preserves proto-Earth characteristics. As long as this predicted compositional stratification can at least partially be preserved over the subsequent billions of years of Earth mantle convection, the compositional similarity between the Moon and the accessible Earth's mantle is a natural outcome of realistic and high-probability Moon-forming impact scenarios. The preservation of primordial heterogeneity in the modern Earth not only reconciles geochemical constraints but is also consistent with recent geophysical observations. Furthermore, for significant preservation of a proto-Earth reservoir, the bulk composition of the Earth-Moon system may be systematically shifted towards chondritic values., Comments are welcome

Using cosmological simulations, we show that the cosmic web of dwarf galaxies in a warm dark matter (WDM) universe, wherein low mass halo formation is heavily suppressed, is nearly indistinguishable to that of a cold dark matter (CDM) universe whose low mass halos are not seen because galaxy formation is suppressed below some threshold mass. Low mass warm dark matter halos are suppressed nearly equally in all environments. For example, WDM voids in the galaxy distribution are neither larger nor emptier than CDM voids, once normalized to the same total galaxy number density and assuming galaxy luminosity scales with halo mass. It is thus a challenge to find hints about the dark matter particle in the cosmic web of galaxies. However, if the scatter between dwarf galaxy luminosity and halo properties is large, low mass CDM halos would sometimes host bright galaxies thereby populating voids that would be empty in WDM. Future surveys that will capture the small scale clustering in the local volume could thus help determine whether the CDM problem of the over-abundance of small halos with respect to the number density of observed dwarf galaxies has a cosmological solution or an astrophysical solution.

Giant impacts (GIs) are common in the late stage of planet formation. The Smoothed Particle Hydrodynamics (SPH) method is widely used for simulating the outcome of such violent collisions, one prominent example being the formation of the Moon. However, a decade of numerical studies in various areas of computational astrophysics has shown that the standard formulation of SPH suffers from several shortcomings such as artificial surface tension and its tendency to promptly damp turbulent motions on scales much larger than the physical dissipation scale, both resulting in the suppression of mixing. In order to quantify how severe these limitations are when modeling GIs we carried out a comparison of simulations with identical initial conditions performed with the standard SPH as well as with the novel Lagrangian Meshless Finite Mass (MFM) method in the GIZMO code. We confirm the lack of mixing between the impactor and target when SPH is employed, while MFM is capable of driving vigorous sub-sonic turbulence and leads to significant mixing between the two bodies. Modern SPH variants with artificial conductivity, a different formulation of the hydro force or reduced artificial viscosity, do not improve mixing as significantly. Angular momentum is conserved similarly well in both methods, but MFM does not suffer from spurious transport induced by artificial viscosity, resulting in a slightly higher angular momentum of the proto-lunar disk. Furthermore, SPH initial conditions exhibit an unphysical density discontinuity at the core-mantle boundary which is easily removed in MFM., 7 pages 5 figures

Monthly Notices of the Royal Astronomical Society, 369 (3), ISSN:0035-8711, ISSN:1365-2966, ISSN:1365-8711

We present a statistical study of substructure within a sample of LCDM clusters and galaxies simulated with up to 25 million particles. With thousands of subhalos per object we can accurately measure their spatial clustering and velocity distribution functions and compare these with observational data. The substructure properties of galactic halos closely resembles those of galaxy clusters with a small scatter in the mass and circular velocity functions. The velocity distribution function is non-Maxwellian and flat topped with a negative kurtosis of about -0.7. Within the virial radius the velocity bias $b=\sigma_{\rm sub}/\sigma_{\rm DM}\sim 1.12 \pm 0.04$, increasing to b > 1.3 within the halo centers. Slow subhalos are much less common, due to physical disruption by gravitational tides early in the merging history. This leads to a spatially anti-biased subhalo distribution that is well fitted by a cored isothermal. Observations of cluster galaxies do not show such biases which we interpret as a limitation of pure dark matter simulations - we estimate that we are missing half of the halo population which has been destroyed by physical overmerging. High resolution hydrodynamical simulations are required to study these issues further. If CDM is correct then the cluster galaxies must survive the tidal field, perhaps due to baryonic inflow during elliptical galaxy formation. Spirals can never exist near the cluster centers and the elliptical galaxies there will have little remaining dark matter. This implies that the morphology-density relation is set {\it before} the cluster forms, rather than a subsequent transformation of disks to S0's by virtue of the cluster environment.

Recent observations suggest that Carina and other nearby dwarf spheroidal galaxies are surrounded by unbound stars tidally stripped by the Milky Way. We run high-resolution N-Body simulations of dwarf galaxies orbiting within the Milky Way halo to determine if such observations can be explained with dark matter potentials as those implied by current structure formation models. We show that tidal forces acting on dwarfs with constant density cores or with cuspy profiles having a low concentration parameter ($c < 5$) lead to flat outer stellar density profiles like that of Carina for a variety of orbital configurations. On the contrary, it is more difficult to remove stars from cuspy dark matter halos with concentrations as high as predicted by CDM models at the mass scale of dwarf galaxies ($c \simgt 10$) and the data can only be reproduced assuming nearly radial orbits. Our simulations show that Carina is losing mass at a fractional rate $< 0.1$ Gyr$^{-1}$ and its mass-to-light ratio could be inflated by at most a factor of 2 due to unbound stars projected along the line of sight. We follow the evolution of the tidal debris within a triaxial clumpy cold dark matter Milky Way halo which causes differential precession and small scale heating of the stellar streams. This renders their use as a dynamical tracer of the Galactic potential practically useless, but does provide a novel test of the nature of the dark matter. Models with warm dark matter (WDM) or fluid dark matter (FDM) produce dwarf halos with lower central densities than CDM and would be consistent with the observed tidal tails even for orbits with eccentricities as low as indicated by current data on nearby dwarf spheroidals. Galactic halos in FDM are smooth and spherical and would be favored by the detection of coherent streams., 10 pages, 8 figures, submitted to MNRAS

Monthly Notices of the Royal Astronomical Society, 364 (2), ISSN:0035-8711, ISSN:1365-2966, ISSN:1365-8711

In this paper we present solutions to three short comings of Smoothed Particles Hydrodynamics (SPH) encountered in previous work when applying it to Giant Impacts. First we introduce a novel method to obtain accurate SPH representations of a planet's equilibrium initial conditions based on equal area tessellations of the sphere. This allows one to imprint an arbitrary density and internal energy profile with very low noise which substantially reduces computation because these models require no relaxation prior to use. As a consequence one can significantly increase the resolution and more flexibly change the initial bodies to explore larger parts of the impact parameter space in simulations. The second issue addressed is the proper treatment of the matter/vacuum boundary at a planet's surface with a modified SPH density estimator that properly calculates the density stabilizing the models and avoiding an artificially low density atmosphere prior to impact. Further we present a novel SPH scheme that simultaneously conserves both energy and entropy for an arbitrary equation of state. This prevents loss of entropy during the simulation and further assures that the material does not evolve into unphysical states. Application of these modifications to impact simulations for different resolutions up to $6.4 \cdot 10^6$ particles show a general agreement with prior result. However, we observe resolution dependent differences in the evolution and composition of post collision ejecta. This strongly suggests that the use of more sophisticated equations of state also demands a large number of particles in such simulations.

Terrestrial planets are thought to be the result of a vast number of gravitational interactions and collisions between smaller bodies. We use numerical simulations to show that practically identical initial conditions result in a wide array of final planetary configurations. This is a result of the chaotic evolution of trajectories which are highly sensitive to minuscule displacements. We determine that differences between systems evolved from virtually identical initial conditions can be larger than the differences between systems evolved from very different initial conditions. This implies that individual simulations lack predictive power. For example, there is not a reproducible mapping between the initial and final surface density profiles. However, some key global properties can still be extracted if the statistical spread across many simulations is considered. Based on these spreads, we explore the collisional growth and orbital properties of terrestrial planets which assemble from different initial conditions (we vary the initial planetesimal distribution, planetesimal masses, and giant planet orbits). Confirming past work, we find that the resulting planetary systems are sculpted by sweeping secular resonances. Configurations with giant planets on eccentric orbits produce fewer and more massive terrestrial planets on tighter orbits than those with giants on circular orbits. This is further enhanced if the initial mass distribution is biased to the inner regions. In all cases, the outer edge of the system is set by the final location of the $\nu_6$ resonance and we find that the mass distribution peaks at the $\nu_5$ resonance. Using existing observations, we find that extrasolar systems follow similar trends. Although differences between our numerical modelling and exoplanetary systems remain, we suggest that CoRoT-7, HD 20003, and HD 20781 may host undetected giant planets., Comment: replaced to match published version, 20 pages, 11 figures, published in MNRAS, simulation outputs available at https://cheleb.net/astro/sp15/

We report on the successful completion of a 2 trillion particle cosmological simulation to z=0 run on the Piz Daint supercomputer (CSCS, Switzerland), using 4000+ GPU nodes for a little less than 80h of wall-clock time or 350,000 node hours. Using multiple benchmarks and performance measurements on the US Oak Ridge National Laboratory Titan supercomputer, we demonstrate that our code PKDGRAV3, delivers, to our knowledge, the fastest time-to-solution for large-scale cosmological N-body simulations. This was made possible by using the Fast Multipole Method in conjunction with individual and adaptive particle time steps, both deployed efficiently (and for the first time) on supercomputers with GPU-accelerated nodes. The very low memory footprint of PKDGRAV3 allowed us to run the first ever benchmark with 8 trillion particles on Titan, and to achieve perfect scaling up to 18000 nodes and a peak performance of 10 Pflops., Comment: 13 pages, 7 figures, submitted to Computational Astrophysics and Cosmology

Future galaxy surveys require one percent precision in the theoretical knowledge of the power spectrum over a large range including very nonlinear scales. While this level of accuracy is easily obtained in the linear regime with perturbation theory, it represents a serious challenge for small scales where numerical simulations are required. In this paper we quantify the precision of present-day $N$-body methods, identifying main potential error sources from the set-up of initial conditions to the measurement of the final power spectrum. We directly compare three widely used $N$-body codes, Ramses, Pkdgrav3, and Gadget3 which represent three main discretisation techniques: the particle-mesh method, the tree method, and a hybrid combination of the two. For standard run parameters, the codes agree to within one percent at $k\leq1$ $h\,\rm Mpc^{-1}$ and to within three percent at $k\leq10$ $h\,\rm Mpc^{-1}$. We also consider the bispectrum and show that the reduced bispectra agree at the sub-percent level for $k\leq 2$ $h\,\rm Mpc^{-1}$. In a second step, we quantify potential errors due to initial conditions, box size, and resolution using an extended suite of simulations performed with our fastest code Pkdgrav3. We demonstrate that the simulation box size should not be smaller than $L=0.5$ $h^{-1}\rm Gpc$ to avoid systematic finite-volume effects (while much larger boxes are required to beat down the statistical sample variance). Furthermore, a maximum particle mass of $M_{\rm p}=10^{9}$ $h^{-1}\rm M_{\odot}$ is required to conservatively obtain one percent precision of the matter power spectrum. As a consequence, numerical simulations covering large survey volumes of upcoming missions such as DES, LSST, and Euclid will need more than a trillion particles to reproduce clustering properties at the targeted accuracy., matches published version

[abridged] We present a detailed comparison of fundamental dark matter halo properties retrieved by a substantial number of different halo finders. These codes span a wide range of techniques including friends-of-friends (FOF), spherical-overdensity (SO) and phase-space based algorithms. We further introduce a robust (and publicly available) suite of test scenarios that allows halo finder developers to compare the performance of their codes against those presented here. This set includes mock haloes containing various levels and distributions of substructure at a range of resolutions as well as a cosmological simulation of the large-scale structure of the universe. All the halo finding codes tested could successfully recover the spatial location of our mock haloes. They further returned lists of particles (potentially) belonging to the object that led to coinciding values for the maximum of the circular velocity profile and the radius where it is reached. All the finders based in configuration space struggled to recover substructure that was located close to the centre of the host halo and the radial dependence of the mass recovered varies from finder to finder. Those finders based in phase space could resolve central substructure although they found difficulties in accurately recovering its properties. Via a resolution study we found that most of the finders could not reliably recover substructure containing fewer than 30-40 particles. However, also here the phase space finders excelled by resolving substructure down to 10-20 particles. By comparing the halo finders using a high resolution cosmological volume we found that they agree remarkably well on fundamental properties of astrophysical significance (e.g. mass, position, velocity, and peak of the rotation curve).

We present results from a suite of N-body simulations that follow the accretion history of the terrestrial planets using a new parallel treecode that we have developed. We initially place 2000 equal size planetesimals between 0.5--4.0 AU and the collisional growth is followed until the completion of planetary accretion (> 100 Myr). All the important effect of gas in laminar disks are taken into account: the aerodynamic gas drag, the disk-planet interaction including Type I migration, and the global disk potential which causes inward migration of secular resonances as the gas dissipates. We vary the initial total mass and spatial distribution of the planetesimals, the time scale of dissipation of nebular gas, and orbits of Jupiter and Saturn. We end up with one to five planets in the terrestrial region. In order to maintain sufficient mass in this region in the presence of Type I migration, the time scale of gas dissipation needs to be 1-2 Myr. The final configurations and collisional histories strongly depend on the orbital eccentricity of Jupiter. If today's eccentricity of Jupiter is used, then most of bodies in the asteroidal region are swept up within the terrestrial region owing to the inward migration of the secular resonance, and giant impacts between protoplanets occur most commonly around 10 Myr. If the orbital eccentricity of Jupiter is close to zero, as suggested in the Nice model, the effect of the secular resonance is negligible and a large amount of mass stays for a long period of time in the asteroidal region. With a circular orbit for Jupiter, giant impacts usually occur around 100 Myr, consistent with the accretion time scale indicated from isotope records. However, we inevitably have an Earth size planet at around 2 AU in this case. It is very difficult to obtain spatially concentrated terrestrial planets together with very late giant impacts., 51 pages, 19 figures, 2 tables, published in Icarus

The final inspiral and coalescence of a black hole binary can produce highly beamed gravitational wave radiation. To conserve linear momentum, the black hole remnant can recoil with "kick" velocity as high as 4000 km/s. We present two sets of full N-body simulations of recoiling massive black holes (MBH) in high-resolution, non-axisymmetric potentials. The host to the first set of simulations is the main halo of the Via Lactea I simulation (Diemand et al. 2007). The nature of the resulting orbits is investigated through a numerical model where orbits are integrated assuming an evolving, triaxial NFW potential, and dynamical friction is calculated directly from the velocity dispersion along the major axes of the main halo of Via Lactea I. By comparing the triaxial case to a spherical model, we find that the wandering time spent by the MBH is significantly increased due to the asphericity of the halo. For kicks larger than 200 km/s, the remnant MBH does not return to the inner 200 pc within 1 Gyr, a timescale an order of magnitude larger than the upper limit of the estimated QSO lifetime. The second set of simulations is run using the outcome of a high-resolution gas-rich merger (Mayer et al. 2007) as host potential. In this case, a recoil velocity of 500 km/s cannot remove the MBH from the nuclear region., 4 pages, 4 figures. Proceedings of the conference Galactic & Stellar Dynamics In the Era of High Resolution Surveys

The effects of particle discreteness in N-body simulations of Lambda Cold Dark Matter (LambdaCDM) are still an intensively debated issue. In this paper we explore such effects, taking into account the scatter caused by the randomness of the initial conditions, and focusing on the statistical properties of the cosmological density field. For this purpose, we run large sets of LambdaCDM simulations and analyse them using a large variety of diagnostics, including new and powerful wavelet statistics. Among other facts, we point out (1) that dynamical evolution does not propagate discreteness noise up from the small scales at which it is introduced, and (2) that one should aim to satisfy the condition epsilon ~ 2d, where "epsilon" is the force resolution and "d" is the interparticle distance. We clarify what such a condition means, and how to implement it in modern cosmological codes.

Remnant planetesimals might have played an important role in reducing the orbital eccentricities of the terrestrial planets after their formation via giant impacts. However, the population and the size distribution of remnant planetesimals during and after the giant impact stage are unknown, because simulations of planetary accretion in the runaway growth and giant impact stages have been conducted independently. Here we report results of direct N-body simulations of the formation of terrestrial planets beginning with a compact planetesimal disk. The initial planetesimal disk has a total mass and angular momentum as observed for the terrestrial planets, and we vary the width (0.3 and 0.5AU) and the number of planetesimals (1000-5000). This initial configuration generally gives rise to three final planets of similar size, and sometimes a fourth small planet forms near the location of Mars. Since a sufficient number of planetesimals remains, even after the giant impact phase, the final orbital eccentricities are as small as those of the Earth and Venus., 36 pages, 9 figures, 1 table, Accepted in ApJ

In cold dark matter cosmological models, structures form and grow by merging of smaller units. Numerical simulations have shown that such merging is incomplete; the inner cores of halos survive and orbit as "subhalos" within their hosts. Here we report a simulation that resolves such substructure even in the very inner regions of the Galactic halo. We find hundreds of very concentrated dark matter clumps surviving near the solar circle, as well as numerous cold streams. The simulation reveals the fractal nature of dark matter clustering: Isolated halos and subhalos contain the same relative amount of substructure and both have cuspy inner density profiles. The inner mass and phase-space densities of subhalos match those of recently discovered faint, dark matter-dominated dwarf satellite galaxies and the overall amount of substructure can explain the anomalous flux ratios seen in strong gravitational lenses. Subhalos boost gamma-ray production from dark matter annihilation, by factors of 4-15, relative to smooth galactic models. Local cosmic ray production is also enhanced, typically by a factor 1.4, but by more than a factor of ten in one percent of locations lying sufficiently close to a large subhalo. These estimates assume that gravitational effects of baryons on dark matter substructure are small., Comment: 14 pages, 5 figures, to appear in Nature, includes supplementary information. Full version of Figure 1 available at http://www.ucolick.org/~diemand/vl2/fig1.png

We have carried out a hydrodynamical code comparison study of interacting multiphase fluids. The two commonly used techniques of grid and smoothed particle hydrodynamics (SPH) show striking differences in their ability to model processes that are fundamentally important across many areas of astrophysics. Whilst Eulerian grid based methods are able to resolve and treat important dynamical instabilities, such as Kelvin-Helmholtz or Rayleigh-Taylor, these processes are poorly or not at all resolved by existing SPH techniques. We show that the reason for this is that SPH, at least in its standard implementation, introduces spurious pressure forces on particles in regions where there are steep density gradients. This results in a boundary gap of the size of the SPH smoothing kernel over which information is not transferred.

Monthly Notices of the Royal Astronomical Society, 376 (1), ISSN:0035-8711, ISSN:1365-2966, ISSN:1365-8711

If the dark matter particle is a neutralino then the first structures to form are cuspy cold dark matter (CDM) haloes collapsing after redshifts z ~ 100 in the mass range 10^{-6} - 10^{-3} Msun. We carry out a detailed study of the survival of these micro-haloes in the Galaxy as they experience tidal encounters with stars, molecular clouds, and other dark matter substructures. We test the validity of analytic impulsive heating calculations using high resolution N-body simulations. A major limitation of analytic estimates is that mean energy inputs are compared to mean binding energies, instead of the actual mass lost from the system. This energy criterion leads to an overestimate of the stripped mass and underestimate of the disruption timescale since CDM haloes are strongly bound in their inner parts. We show that a significant fraction of material from CDM micro-haloes can be unbound by encounters with Galactic substructure and stars, however the cuspy central regions remain relatively intact. Furthermore, the micro-haloes near the solar radius are those which collapse significantly earlier than average and will suffer very little mass loss. Thus we expect a fraction of surviving bound micro-haloes, a smooth component with narrow features in phase space, which may be uncovered by direct detection experiments, as well as numerous surviving cuspy cores with proper motions of arc-minutes per year, which can be detected indirectly via their annihilation into gamma-rays.

Monthly Notices of the Royal Astronomical Society, 375 (1), ISSN:0035-8711, ISSN:1365-2966, ISSN:1365-8711

We study the formation of disks via the cooling flow of gas within galactic haloes using smoothed particle hydrodynamics simulations. These simulations resolve mass scales of a few thousand solar masses in the gas component for the first time. Thermal instabilities result in the formation of numerous warm clouds that are pressure confined by the hot ambient halo gas. The clouds fall slowly onto the disk through non-spherical accretion from material flowing preferentially down the angular momentum axis. The rotational velocity of the infalling cold gas decreases as a function of height above the disk, closely resembling that of the extra-planar gas recently observed around the spiral galaxy NGC 891.

Monthly Notices of the Royal Astronomical Society, 368 (3), ISSN:0035-8711, ISSN:1365-2966, ISSN:1365-8711

The recent discovery of hidden non-axysimmetric and disk-like structures in bright Virgo dwarf elliptical and lenticular galaxies (dE/dSph/dS0) indicates that they may have late-type progenitors. Using N-body simulations we follow the evolution of disk galaxies within aCDM cluster simulated with 10 7 particles, where the hierarchical growth and galaxy harassment are modeled self-consistently. Most of the galaxies undergo significant morphological transformation, even at the outskirts of the cluster, and move through the Hubble sequence from late type disks to dwarf spheroidals. None of the disks are completely destroyed therefore they can not be the progenitors of ultra compact dwarf galaxies (UCDs). The time evolution of the simulated galaxies is compared with unsharp-masked images obtained from VLT data and the projected kinematics of our models with the latest high resolution spectroscopic studies from the Keck and Palomar telescopes.

We use high resolution 3D SPH simulations to study the evolution of self-gravitating binary protoplanetary disks. Heating by shocks and cooling are included. We consider different orbital separations and masses of the disks and central stars. Isolated massive disks ($M \sim 0.1 M_{{\o}dot}$) fragment into protoplanets as a result of gravitational instability for cooling times comparable to the orbital time. Fragmentation does not occur in binary systems with a separation of about 60 AU. This is because efficient heating owing to strong tidally induced spiral shocks damps any overdensity. The resulting temperatures, above 200 K, would vaporize water ice in the outer disk, posing a problem even for the other model of giant planet formation, core-accretion. Light disks ($M \sim 0.01 M_{\odot}$) do not fragment but remain cold because their low self-gravity inhibits strong shocks. Core accretion would not be hampered in the latter. At separations of about 120 AU the efficiency of fragmentation by disk instability rises and approaches that in isolated systems. If disk instability is the main formation mechanism for giant planets, on going surveys targeting binary systems should find considerably fewer planets in systems with separations below 100 AU., Comment: 5 pages, 3 figures, submitted to MNRAS Letters

We use numerical simulations in a ACDM cosmology to model density profiles in a set of 16 dark matter haloes with resolutions of up to seven million particles within the virial radius. These simulations allow us to follow robustly the formation and evolution of the central cusp over a large mass range of 10 11 -10 14 M○., down to approximately 0.5 per cent of the virial radius, and from redshift of 5 to the present, covering a larger range in parameter space than previous works. We confirm that the cusp of the density profile is set at redshifts of 2 or greater and remains remarkably stable to the present time, when considered in non-comoving coordinates. Motivated by the diversity and evolution of halo profile shapes, we fit our haloes to the two-parameter profile, p oc 1/((c γ r/r vir ) γ [1 + (c γ r/r vir )] 3-γ }, where the steepness of the cusp is given by the asymptotic inner slope parameter, y, and its radial extent is described by the concentration parameter, c γ (with c γ defined as the virial radius divided by the concentration radius). In our simulations, we find γ ≃ 1.4 - 0.08 log 10 (M/M*) for haloes of 0.01 M*-1000 M*, with a large scatter of Δγ ∼ ±0.3, where M* is the redshift dependent characteristic mass of collapsing haloes; and c γ ≃ 8.(M/M*) -0.15 , with a large M/M* dependent scatter roughly equal to ± cy. Our redshift zero haloes have inner slope parameters ranging approximately from r -1 (Navarro, Frank and White) to r -1.5 (Moore et al.), with a median of roughly r- 1.3 . This two-parameter profile fit works well for all types of haloes in our simulations, whether or not they show evidence of a steep asymptotic cusp. We also model a cluster in power-law cosmologies of P k n , with n = (0, -1, -2, -2.7). Here we find that the concentration radius and the inner cusp slope are both a function of n, with larger concentration radii and shallower cusps for steeper power spectra. We have completed a thorough resolution study and find that the minimum resolved radius is well described by the mean interparticle separation over a range of masses and redshifts. The trend of steeper and more concentrated cusps for smaller M/M* haloes clearly shows that dwarf-sized ACDM haloes have, on average, significantly steeper density profiles within the inner few per cent of the virial radius than inferred from recent observations. Code to reproduce this profile can be downloaded from http://www.icc.dur.ac.uk/ ∼reed/profile.html.

We use a set of high-resolution cosmological N-body simulations to investigate the inner mass profile of galaxy-sized cold dark matter (CDM) halos. These simulations extend the thorough numerical convergence study presented in Paper I of this series (Power et al. 2003), and demonstrate that the mass profile of CDM halos can be robustly estimated beyond a minimum converged radius of order r_conv ~ 1 kpc/h in our highest resolution runs. The density profiles of simulated halos become progressively shallow from the virial radius inwards, and show no sign of approaching a well-defined power-law behaviour near the centre. At r_conv, the logarithmic slope of the density profile is steeper than the asymptotic \rho \propto r^-1 expected from the formula proposed by Navarro, Frenk, and White (1996), but significantly shallower than the steeply divergent \rho \propto r^-1.5 cusp proposed by Moore et al. (1999). We perform a direct comparison of the spherically-averaged dark matter circular velocity (V_c) profiles with rotation curves of low surface brightness (LSB) galaxies from the samples of de Blok et al. (2001), de Blok and Bosma (2002), and Swaters et al. (2003). Most (about two-thirds) LSB galaxies in this dataset are roughly consistent with CDM halo V_c profiles. However, about one third of LSBs in these samples feature a sharp transition between the rising and flat part of the rotation curve that is not seen in the V_c profiles of CDM halos. This discrepancy has been interpreted as excluding the presence of cusps, but we argue that it might simply reflect the difference between circular velocity and gas rotation speed likely to arise in gaseous disks embedded within realistic, triaxial CDM halos.