Marius Cautun

Evolution of the cosmic web

The cosmic web is the largest scale manifestation of the anisotropic gravitational collapse of matter. It represents the transitional stage between linear and non-linear structures and contains easily accessible information about the early phases of structure formation processes. Here we investigate the characteristics and the time evolution of morphological components. Our analysis involves the application of the NEXUS Multiscale Morphology Filter technique, predominantly its NEXUS+ version, to high resolution and large volume cosmological simulations. We quantify the cosmic web components in terms of their mass and volume content, their density distribution and halo populations. We employ new analysis techniques to determine the spatial extent of filaments and sheets, like their total length and local width. This analysis identifies clusters and filaments as the most prominent components of the web. In contrast, while voids and sheets take most of the volume, they correspond to underdense environments and are devoid of group-sized and more massive haloes. At early times the cosmos is dominated by tenuous filaments and sheets, which, during subsequent evolution, merge together, such that the present-day web is dominated by fewer, but much more massive, structures. The analysis of the mass transport between environments clearly shows how matter flows from voids into walls, and then via filaments into cluster regions, which form the nodes of the cosmic web. We also study the properties of individual filamentary branches, to find long, almost straight, filaments extending to distances larger than 100 h-1 Mpc. These constitute the bridges between massive clusters, which seem to form along approximatively straight lines.

NEXUS : tracing the cosmic web connection.

We introduce the NEXUS algorithm for the identification of cosmic web environments: clusters, filaments, walls and voids. This is a multiscale and automatic morphological analysis tool that identifies all the cosmic structures in a scale free way, without preference for a certain size or shape. We develop the NEXUS method to incorporate the density, tidal field, velocity divergence and velocity shear as tracers of the cosmic web. We also present the NEXUS+ procedure which, taking advantage of a novel filtering of the density in logarithmic space, is very successful at identifying the filament and wall environments in a robust and natural way. To assess the algorithms we apply them to an N-body simulation. We find that all methods correctly identify the most prominent filaments and walls, while there are differences in the detection of the more tenuous structures. In general, the structures traced by the density and tidal fields are clumpier and more rugged than those present in the velocity divergence and velocity shear fields. We find that the NEXUS+ method captures much better the filamentary and wall networks and is successful in detecting even the fainter structures. We also confirm the efficiency of our methods by examining the dark matter particle and halo distributions.

The Copernicus Complexio: a high-resolution view of the small-scale Universe

We introduce Copernicus Complexio (COCO), a high-resolution cosmological N-body simulation of structure formation in the ΛCDM model. COCO follows an approximately spherical region of radius ∼17.4 h−1 Mpc embedded in a much larger periodic cube that is followed at lower resolution. The high-resolution volume has a particle mass of 1.135 × 105 h−1 M⊙ (60 times higher than the Millennium-II simulation). COCO gives the dark matter halo mass function over eight orders of magnitude in halo mass; it forms ∼60 haloes of galactic size, each resolved with about 10 million particles. We confirm the power-law character of the subhalo mass function, , down to a reduced subhalo mass Msub/M200 ≡ μ = 10−6, with a best-fitting power-law index, s = 0.94, for hosts of mass 〈M200〉 = 1012 h−1 M⊙. The concentration–mass relation of COCO haloes deviates from a single power law for masses M200 < afew × 108 h−1 M⊙, where it flattens, in agreement with results by Sanchez-Conde et al. The host mass invariance of the reduced maximum circular velocity function of subhaloes, ν ≡ Vmax/V200, hinted at in previous simulations, is clearly demonstrated over five orders of magnitude in host mass. Similarly, we find that the average, normalized radial distribution of subhaloes is approximately universal (i.e. independent of subhalo mass), as previously suggested by the Aquarius simulations of individual haloes. Finally, we find that at fixed physical subhalo size, subhaloes in lower mass hosts typically have lower central densities than those in higher mass hosts.

Planes of satellite galaxies: when exceptions are the rule

The detection of planar structures within the satellite systems of both the Milky Way (MW) and Andromeda (M31) has been reported as being in stark contradiction to the predictions of the standard cosmological model (Λ cold dark matter – ΛCDM). Given the ambiguity in defining a planar configuration, it is unclear how to interpret the low incidence of the MW and M31 planes in ΛCDM. We investigate the prevalence of satellite planes around galactic mass haloes identified in high-resolution cosmological simulations. We find that planar structures are very common, and that ∼10 per cent of ΛCDM haloes have even more prominent planes than those present in the Local Group. While ubiquitous, the planes of satellite galaxies show a large diversity in their properties. This precludes using one or two systems as small-scale probes of cosmology, since a large sample of satellite systems is needed to obtain a good measure of the object-to-object variation. This very diversity has been misinterpreted as a discrepancy between the satellite planes observed in the Local Group and ΛCDM predictions. In fact, ∼10 per cent of ΛCDM galactic haloes have planes of satellites that are as infrequent as the MW and M31 planes. The look-elsewhere effect plays an important role in assessing the detection significance of satellite planes and accounting for it leads to overestimating the significance level by a factor of 30 and 100 for the MW and M31 systems, respectively.

Weak lensing by voids in modified lensing potentials

We study lensing by voids in Cubic Galileon and Nonlocal gravity cosmologies, which are examples of theories of gravity that modify the lensing potential. We find voids in the dark matter and halo density fields of N-body simulations and compute their lensing signal analytically from the void density profiles, which we show are well fit by a simple analytical formula. In the Cubic Galileon model, the modifications to gravity inside voids are not screened and they approximately double the size of the lensing effects compared to GR. The difference is largely determined by the direct effects of the fifth force on lensing and less so by the modified density profiles. For this model, we also discuss the subtle impact on the force and lensing calculations caused by the screening effects of haloes that exist in and around voids. In the Nonlocal model, the impact of the modified density profiles and the direct modifications to lensing are comparable, but they boost the lensing signal by only ≈ 10%, compared with that of GR. Overall, our results suggest that lensing by voids is a promising tool to test models of gravity that modify lensing.

Assembly of filamentary void galaxy configurations

We study the formation and evolution of filamentary configurations of dark matter haloes in voids. Our investigation uses the high-resolution Lambda cold dark matter simulation CosmoGrid to look for void systems resembling the VGS_31 elongated system of three interacting galaxies that was recently discovered by the Void Galaxy Survey inside a large void in the Sloan Digital Sky Survey galaxy redshift survey. H I data revealed these galaxies to be embedded in a common elongated envelope, possibly embedded in intravoid filament. In the CosmoGrid simulation we look for systems similar to VGS_31 in mass, size and environment. We find a total of eight such systems. For these systems, we study the distribution of neighbour haloes, the assembly and evolution of the main haloes and the dynamical evolution of the haloes, as well as the evolution of the large-scale structure in which the systems are embedded. The spatial distribution of the haloes follows that of the dark matter environment. We find that VGS_31-like systems have a large variation in formation time, having formed between 10 Gyr ago and the present epoch. However, the environments in which the systems are embedded evolved to resemble each other substantially. Each of the VGS_31-like systems is embedded in an intravoid wall, that no later than z = 0.5 became the only prominent feature in its environment. While part of the void walls retain a rather featureless character, we find that around half of them are marked by a pronounced and rapidly evolving substructure. Five haloes find themselves in a tenuous filament of a fewh-1 Mpc long inside the intravoid wall. Finally, we compare the results to observed data from VGS_31. Our study implies that the VGS_31 galaxies formed in the same (proto) filament, and did not meet just recently. The diversity amongst the simulated halo systems indicates that VGS_31 may not be typical for groups of galaxies in voids.

A new spin on discs of satellite galaxies

We investigate the angular and kinematic distributions of satellite galaxies around a large sample of bright isolated primaries in the spectroscopic and photometric catalogues of the Sloan Digital Sky Survey (SDSS). We detect significant anisotropy in the spatial distribution of satellites. To test whether this anisotropy could be related to the rotating discs of satellites recently found by Ibata et al. in a sample of SDSS galaxies, we repeat and extend their analysis. Ibata et al. found an excess of satellites on opposite sides of their primaries having anticorrelated radial velocities. We find that this excess is sensitive to small changes in the sample selection criteria which can greatly reduce its significance. In addition, we find no evidence for correspondingly correlated velocities for satellites observed on the same side of their primaries, which would be expected for rotating discs of satellites. We conclude that the detection of rotating planes of satellites in the observational sample of Ibata et al. is not robust to changes in the sample selection criteria. We compare our data to the Λ cold dark matter Millennium simulations populated with galaxies according to the semi-analytic model of Guo et al. We find excellent agreement with the spatial distribution of satellites in the SDSS data and the lack of a strong signal from coherent rotation.

Subhalo statistics of galactic haloes: beyond the resolution limit

We study the substructure population of Milky Way (MW)-mass haloes in the Λ cold dark matter (ΛCDM) cosmology using a novel procedure to extrapolate subhalo number statistics beyond the resolution limit of N-body simulations. The technique recovers the mean and the variance of the subhalo abundance, but not its spatial distribution. It extends the dynamic range over which precise statistical predictions can be made by the equivalent of performing a simulation with 50 times higher resolution, at no additional computational cost. We apply this technique to MW-mass haloes, but it can easily be applied to haloes of any mass. We find up to 20 per cent more substructures in MW-mass haloes than found in previous studies. Our analysis lowers the mass of the MW halo required to accommodate the observation that the MW has only three satellites with a maximum circular velocity Vmax ≥ 30 km s- 1 in the ΛCDM cosmology. The probability of having a subhalo population similar to that in the MW is 20 per cent for a virial mass, M200 = 1 × 1012 M⊙ and practically zero for haloes more massive than M200 = 2 × 1012 M⊙.

The view from the boundary: a new void stacking method

We introduce a new method for stacking voids and deriving their profile that greatly increases the potential of voids as a tool for precision cosmology. Given that voids are distinctly non-spherical and have most of their mass at their edge, voids are better described relative to their boundary rather than relative to their centre, as in the conventional spherical stacking approach. The boundary profile is obtained by computing the distance of each volume element from the void boundary. Voids can then be stacked and their profiles computed as a function of this boundary distance. This approach enhances the weak lensing signal of voids, both shear and convergence, by a factor of 2 when compared to the spherical stacking method. It also results in steeper void density profiles that are characterized by a very slow rise inside the void and a pronounced density ridge at the void boundary. The resulting boundary density profile is self-similar when rescaled by the thickness of the density ridge, implying that the average rescaled profile is independent of void size. The boundary velocity profile is characterized by outflows in the inner regions whose amplitude scales with void size, and by a strong inflow into the filaments and walls delimiting the void. This new picture enables a straightforward discrimination between collapsing and expanding voids both for individual objects as well as for stacked samples.

Tracing the cosmic web

The cosmic web is one of the most striking features of the distribution of galaxies and dark matter on the largest scales in the Universe. It is composed of dense regions packed full of galaxies, long filamentary bridges, flattened sheets and vast low-density voids. The study of the cosmic web has focused primarily on the identification of such features, and on understanding the environmental effects on galaxy formation and halo assembly. As such, a variety of different methods have been devised to classify the cosmic web - depending on the data at hand, be it numerical simulations, large sky surveys or other. In this paper, we bring 12 of these methods together and apply them to the same data set in order to understand how they compare. In general, these cosmic-web classifiers have been designed with different cosmological goals in mind, and to study different questions. Therefore, one would not a priori expect agreement between different techniques; however, many of these methods do converge on the identification of specific features. In this paper, we study the agreements and disparities of the different methods. For example, each method finds that knots inhabit higher density regions than filaments, etc. and that voids have the lowest densities. For a given web environment, we find a substantial overlap in the density range assigned by each web classification scheme. We also compare classifications on a halo-by-halo basis; for example, we find that 9 of 12 methods classify around a third of group-mass haloes (i.e. M-halo similar to 10(13.5) h(-1) M-circle dot) as being in filaments. Lastly, so that any future cosmic-web classification scheme can be compared to the 12 methods used here, we have made all the data used in this paper public.