Scale symmetry is a fundamental symmetry of physics that seems however not to be fully realized in the universe. Here, we focus on the astronomical scales ruled by gravity, where scale symmetry holds and gives rise to a truly scale invariant distribution of matter, namely it gives rise to a fractal geometry. A suitable explanation of the features of the fractal cosmic mass distribution is provided by the nonlinear Poisson--Boltzmann--Emden equation. An alternative interpretation of this equation is connected with theories of quantum gravity. We study the fractal solutions of the equation and connect them with the statistical theory of random multiplicative cascades, which originated in the theory of fluid turbulence. The type of multifractal mass distributions so obtained agrees with results from the analysis of cosmological simulations and of observations of the galaxy distribution.

We make a statistical analysis of the angular projection of the large-scale stellar mass distribution, obtained from the Sloan Digital Sky Survey (data release 7) with the stellar masses of galaxies. We obtain a clustering length $r_0= 5.9$--$6.5\,h^{-1}$Mpc and a two-point correlation function power-law exponent $\gamma=1.83$--$1.97$, but we find that the projected stellar mass distribution is not very useful to study strong clustering and, in particular, fractal properties. Taking into account the results of the statistical analysis of volume limited samples from the same data, we conclude that a multifractal cosmic-web model with a transition to homogeneity at about $10\,h^{-1}$Mpc is adequate, but there is a considerable uncertainty in both this scale and the correlation dimension.

The cosmic web structure is studied with the concepts and methods of fractal geometry, employing the adhesion model of cosmological dynamics as a basic reference. The structures of matter clusters and cosmic voids in cosmological N-body simulations or the Sloan Digital Sky Survey are elucidated by means of multifractal geometry. A non-lacunar multifractal geometry can encompass three fundamental descriptions of the cosmic structure, namely, the web structure, hierarchical clustering, and halo distributions. Furthermore, it explains our present knowledge of cosmic voids. In this way, a unified theory of the large-scale structure of the universe seems to emerge. The multifractal spectrum that we obtain significantly differs from the one of the adhesion model and conforms better to the laws of gravity. The formation of the cosmic web is best modeled as a type of turbulent dynamics, generalizing the known methods of Burgers turbulence.

A novel fractal analysis of the cosmic web structure is carried out, employing the Sloan Digital Sky Survey, data release 7. We consider the large-scale stellar mass distribution, unlike other analyses, and determine its multifractal geometry, which is compared with the geometry of the cosmic web generated by cosmological N-body simulations. We find a good concordance, the common features being: (i) a minimum singularity strength equal to one, which corresponds to the edge of diverging gravitational energy and differs from the adhesion model prediction; (ii) a "supercluster set" of relatively high dimension where the mass concentrates; and (iii) a non-lacunar structure, like the one generated by the adhesion model.

Dark matter halos can be defined as smooth distributions of dark matter placed in a non-smooth cosmic web structure. This definition of halos demands a precise definition of smoothness and a characterization of the manner in which the transition from smooth halos to the cosmic web takes place. We introduce entropic measures of smoothness, related to measures of inequality previously used in economy and with the advantage of being connected with standard methods of multifractal analysis already used for characterizing the cosmic web structure in cold dark matter N-body simulations. These entropic measures provide us with a quantitative description of the transition from the small scales portrayed as a distribution of halos to the larger scales portrayed as a cosmic web and, therefore, allow us to assign definite sizes to halos. However, these "smoothness sizes" have no direct relation to the virial radii. Finally, we discuss the influence of N-body discreteness parameters on smoothness.

Halo models of the large scale structure of the Universe are critically examined, focusing on the definition of halos as smooth distributions of cold dark matter. This definition is essentially based on the results of cosmological N-body simulations. By a careful analysis of the standard assumptions of halo models and N-body simulations and by taking into account previous studies of self-similarity of the cosmic web structure, we conclude that N-body cosmological simulations are not fully reliable in the range of scales where halos appear. Therefore, to have a consistent definition of halos, it is necessary either to define them as entities of arbitrary size with a grainy rather than smooth structure or to define their size in terms of small-scale baryonic physics.

Large-scale structure formation can be modeled as a nonlinear process that transfers energy from the largest scales to successively smaller scales until it is dissipated, in analogy with Kolmogorov's cascade model of incompressible turbulence. However, cosmic turbulence is very compressible, and vorticity plays a secondary role in it. The simplest model of cosmic turbulence is the adhesion model, which can be studied perturbatively or adapting to it Kolmogorov's non-perturbative approach to incompressible turbulence. This approach leads to observationally testable predictions, e.g., to the power-law exponent of the matter density two-point correlation function.

The Kolmogorov approach to turbulence is applied to the Burgers turbulence in the stochastic adhesion model of large-scale structure formation. As the perturbative approach to this model is unreliable, here is proposed a new, non-perturbative approach, based on a suitable formulation of Kolmogorov's scaling laws. This approach suggests that the power-law exponent of the matter density two-point correlation function is in the range 1--1.33, but it also suggests that the adhesion model neglects important aspects of the gravitational dynamics.

We develop a method of multifractal analysis of N-body cosmological simulations that improves on the customary counts-in-cells method by taking special care of the effects of discreteness and large scale homogeneity. The analysis of the Mare-Nostrum simulation with our method provides strong evidence of self-similar multifractal distributions of dark matter and gas, with a halo mass function that is of Press-Schechter type but has a power-law exponent -2, as corresponds to a multifractal. Furthermore, our analysis shows that the dark matter and gas distributions are indistinguishable as multifractals. To determine if there is any gas biasing, we calculate the cross-correlation coefficient, with negative but inconclusive results. Hence, we develop an effective Bayesian analysis connected with information theory, which clearly demonstrates that the gas is biased in a long range of scales, up to the scale of homogeneity. However, entropic measures related to the Bayesian analysis show that this gas bias is small (in a precise sense) and is such that the fractal singularities of both distributions coincide and are identical. We conclude that this common multifractal cosmic web structure is determined by the dynamics and is independent of the initial conditions.

We introduce new statistical methods for the study of cosmic voids, focusing on the statistics of largest size voids. We distinguish three different types of distributions of voids, namely, Poisson-like, lognormal-like and Pareto-like distributions. The last two distributions are connected with two types of fractal geometry of the matter distribution. Scaling voids with Pareto distribution appear in fractal distributions with box-counting dimension smaller than three (its maximum value), whereas the lognormal void distribution corresponds to multifractals with box-counting dimension equal to three. Moreover, voids of the former type persist in the continuum limit, namely, as the number density of observable objects grows, giving rise to lacunar fractals, whereas voids of the latter type disappear in the continuum limit, giving rise to non-lacunar (multi)fractals. We propose both lacunar and non-lacunar multifractal models of the cosmic web structure of the Universe. A non-lacunar multifractal model is supported by current galaxy surveys as well as cosmological $N$-body simulations. This model suggests, in particular, that small dark matter halos and, arguably, faint galaxies are present in cosmic voids.

Voids are a prominent feature of fractal point distributions but there is no precise definition of what is a void (except in one dimension). Here we propose a definition of voids that uses methods of discrete stochastic geometry, in particular, Delaunay and Voronoi tessellations, and we construct a new algorithm to search for voids in a point set. We find and rank-order the voids of suitable examples of fractal point sets in one and two dimensions to test whether Zipf's power-law holds. We conclude affirmatively and, furthermore, that the rank-ordering of voids conveys similar information to the number-radius function, as regards the scaling regime and the transition to homogeneity. So it is an alternative tool in the analysis of fractal point distributions with crossover to homogeneity and, in particular, of the distribution of galaxies.

CONTEXT: Cosmic voids are observed in the distribution of galaxies and, to some extent, in the dark matter distribution. If these distributions have fractal geometry, it must be reflected in the geometry of voids; in particular, we expect scaling sizes of voids. However, this scaling is not well demonstrated in galaxy surveys yet. AIMS: Our objective is to understand the geometry of cosmic voids in relation to a fractal structure of matter. We intend to distinguish monofractal voids from multifractal voids, regarding their scaling properties. We plan to analyse voids in the distributions of mass concentrations (halos) in a multifractal and their relation to galaxy voids. METHODS: We make a statistical analysis of point distributions based on the void probability function and correlation functions. We assume that voids are spherical and devise a simple spherical void finder. For continuous mass distributions, we employ the methods of fractal geometry. We confirm the analytical predictions with numerical simulations. Smoothed mass distributions are suitable for the method of excursion sets. RESULTS: Voids are very nonlinear and non-perturbative structures. Voids reflect the fractal geometry of the matter distribution but not always directly: scaling sizes of voids imply fractal geometry, but fractal voids may have a complicated geometry and may not have scaling sizes. Proper multifractal voids are of this type. A natural multifractal biasing model implies that the voids in the galaxy distribution inherit the same complicated geometry. CONCLUSIONS: Current galaxy surveys as well as cosmological N-body simulations indicate that cosmic voids are proper multifractal voids. This implies the presence in the voids of galaxies or, at least, small dark matter halos.

On the one hand, the large scale structure of matter is arguably scale invariant, and, on the other hand, halos and voids are recognized as prominent features of that structure. To unify both approaches, we propose to model the dark matter distribution as a set of fractal distributions of halos of different kinds. This model relies on the concept of multifractal as the most general scaling distribution and on a plausible notion of halo as a singular mass concentration in a multifractal. Voids arise as complementary to halos, namely, as formed by regular mass depletions. To provide halos with definite size and masses, we coarse-grain the dark matter distribution, using the length given by the lower cutoff to scaling. This allows us to relate the halo mass function to the multifractal spectrum. Hence, we find that a log-normal model of the mass distribution nicely fits in this picture and, moreover, the Press-Schechter mass function can be recovered as a bifractal limit. To support our model of fractal distributions of halos, we perform a numerical study of the distribution produced in cosmological N-body simulations. In the Virgo L-CDM GIF2 simulation, we indeed find fractal distributions of halos with various dimensions and a halo mass function of bifractal type. However, this mass function is not of Press-Schechter's type, and we interprete it instead as caused by the undersampling of the distribution at the scale of halos, due to discretization.

There is evidence of a scale-invariant matter distribution up to scales over 10 Megaparsecs. We review scaling (fractal or multifractal) models of large scale structure and their observational evidence. We conclude that the dynamics of cosmological structure formation seems to be driven to a multifractal attractor. This supports previous studies, which we review, of structure formation by means of the renormalization group within a hydrodynamic formulation.

"Cut-out sets" are fractals that can be obtained by removing a sequence of disjoint regions from an initial region of d-dimensional euclidean space. Conversely, a description of some fractals in terms of their void complementary set is possible. The essential property of a sequence of fractal voids is that their sizes decrease as a power law, that is, they follow Zipf's law. We prove the relation between the box dimension of the fractal set (in d <= 3) and the exponent of the Zipf law for convex voids; namely, if the Zipf law exponent e is such that 1 < e < d/(d-1) and, in addition, we forbid the appearance of degenerate void shapes, we prove that the corresponding cut-out set has box dimension d/e (d-1 < d/e < d). We explore the application of this result to the large scale distribution of matter in cosmology, in connection with ``cosmic foam'' models.

Spherical accretion flows are simple enough for analytical study, by solution of the corresponding fluid dynamic equations. The solutions of stationary spherical flow are due to Bondi. The questions of the choice of a physical solution and of stability have been widely discussed. The answer to these questions is very dependent on the problem of boundary conditions, which vary according to whether the accretor is a compact object or a black hole. We introduce a particular, simple form of stationary spherical flow, namely, self-similar Bondi flow, as a case with physical interest in which analytic solutions for perturbations can be found. With suitable no matter-flux-perturbation boundary conditions, we will show that acoustic modes are stable in time and have no spatial instability at r=0. Furthermore, their evolution eventually becomes ergodic-like and shows no trace of instability or of acquiring any remarkable pattern.

The question of stability of steady spherical accretion has been studied for many years and, recently, the concept of spatial instability has been introduced. Here we study perturbations of steady spherical accretion flows (Bondi solutions), restricting ourselves to the case of self-similar flow, as a case amenable to analytic treatment and with physical interest. We further restrict ourselves to its acoustic perturbations. The radial perturbation equation can be solved in terms of Bessel functions. We study the formulation of adequate boundary conditions and decide for no matter-flux-perturbation conditions (at the Bondi radius and at r=0). We also consider the problem of initial conditions and time evolution, in particular, of radial perturbations. No spatial instability at r=0 is found. The time evolution is such that perturbations eventually become ergodic-like and show no trace of instability or of acquiring any remarkable pattern.

Voids are a prominent feature of the galaxy distribution but their quantitative study is hindered by the lack of a precise definition of what constitutes a void. Here we propose a definition of voids in point distributions that uses methods of discrete stochastic geometry, in particular, Delaunay and Voronoi tessellations, and we construct a new void-finder. We then apply the void-finder to scaling point distributions. First, we find the voids of pure fractals with a transition to homogeneity and show that the rank ordering of the voids also scales (Zipf's law) and, in addition, shows the transition to homogeneity. However, a pure fractal is arguably not a good model of the galaxy distribution, so we construct from a cosmological $N$-body simulation a bifractal mock galaxy sample representing two galaxy populations, which we identify as "wall" and "field" galaxies. The wall galaxy distribution fits a pure fractal with a transition to homogeneity and, furthermore, the rank ordering of its voids shows a scaling range with the right slope plus a transition to homogeneity.

We examine the proposal that a model of the large-scale matter distribution consisting of randomly placed haloes with power-law profile, as opposed to a fractal model, can account for the observed power-law galaxy-galaxy correlations. We conclude that such model, which can actually be considered as a degenerate multifractal model, is not realistic but suggests a new picture of multifractal models, namely, as sets of fractal distributions of haloes. We analyse, according to this picture, the properties of the matter distribution produced in cosmological N-body simulations, with affirmative results; namely, haloes of similar mass have a fractal distribution with a given dimension, which grows as the mass diminishes.

The effective stress tensor of a homogeneous turbulent rotating fluid is anisotropic. This leads us to consider the most general axisymmetric four-rank ``viscosity tensor'' for a Newtonian fluid and the new terms in the turbulent effective force on large scales that arise from it, in addition to the microscopic viscous force. Some of these terms involve couplings to vorticity and others are angular momentum non conserving (in the rotating frame). Furthermore, we explore the constraints on the response function and the two-point velocity correlation due to axisymmetry. Finally, we compare our viscosity tensor with other four-rank tensors defined in current approaches to non-rotating anisotropic turbulence.

The process of downfall of initially homogeneous gas onto a solid ball due to the ball's gravity (relevant in astrophysical situations) is studied with a combination of analytic and numerical methods. The initial explicit solution soon becomes discontinuous and gives rise to a shock wave. Afterwards, there is a crossover between two intermediate asymptotic similarity regimes, where the shock wave propagates outwards according to two self-similar laws, initially accelerating and eventually decelerating and vanishing, leading to a static state. The numerical study allows one to investigate in detail this dynamical problem and its time evolution, verifying and complementing the analytic results on the initial solution, intermediate self-similar laws and static long-term solution.

The process of gravitational accretion of initially homogeneous gas onto a solid ball is studied with methods of fluid dynamics. The fluid partial differential equations for polytropic flow can be solved exactly in an early stage, but this solution soon becomes discontinuous and gives rise to a shock wave. Afterwards, there is a crossover between two intermediate asymptotic self-similar regimes, where the shock wave propagates outwards according to two similarity laws, initially accelerating, then decelerating (and eventually vanishing). Lastly, we study the final static state. Our purpose is to attain a global picture of the process.

Large scale features of a randomly isotropically forced incompressible and unbounded rotating fluid are examined in perturbation theory. At first order in both the random force amplitude and the angular velocity we find two types of modifications to the fluid equation of motion. The first correction transforms the molecular shear viscosity into a (rotation independent) effective viscosity. The second perturbative correction leads to a new large scale non-dissipative force proportional to the fluid angular velocity in the slow rotation regime. This effective force does no net work and alters the dispersion relation of inertial waves propagating in the fluid. Both dynamically generated corrections can be identified with certain components of the most general axisymmetric ``viscosity tensor'' for a Newtonian fluid.

We study here, from first principles, what properties of voids are to be expected in a fractal point distribution and how the void distribution is related to its morphology. We show this relation in various examples and apply our results to the distribution of galaxies. If the distribution of galaxies forms a fractal set, then this property results in a number of scaling laws to be fulfilled by voids. Consider a fractal set of dimension $D$ and its set of voids. If voids are ordered according to decreasing sizes (largest void has rank R=1, second largest R=2 and so on), then a relation between size $\Lambda$ and rank of the form $\Lambda (R) \propto R^{-z}$ must hold, with $z = d/D$, and where $d$ is the euclidean dimension of the space where the fractal is embedded. The physical restriction $D < d$ means that $z > 1$ in a fractal set. The average size $\bar \Lambda$ of voids depends on the upper ($\Lambda_u$) and the lower ($\Lambda_l$) cut-off as ${\bar \Lambda} \propto \Lambda_u^{1-D/d} \Lambda_l^{D/d}$. Current analysis of void sizes in the galaxy distribution do not show evidence of a fractal distribution, but are insufficient to rule it out. We identify possible shortcomings of current void searching algorithms, such as changes of shape in voids at different scales or merging of voids, and propose modifications useful to test fractality in the galaxy distribution.

The effect of the large-scale cosmological expansion on small systems is studied in the light of modern cosmological models of large-scale structure. We identify certain assumptions of earlier works which render them unrealistic regarding these cosmological models. The question is reanalyzed by dropping these assumptions to conclude that a given small system can experience either an expansion or a contraction of cosmological origin.

The coarse-graining operation in hydrodynamics is equivalent to a change of scale which can be formalized as a renormalization group transformation. In particular, its application to the probability distribution of a self-gravitating fluid yields an "exact renormalization group equation" of Fokker-Planck type. Since the time evolution of that distribution can also be described by a Fokker-Planck equation, we propose a connection between both equations, that is, a connection between scale and time evolution. We finally remark on the essentially non-perturbative nature of astrophysical problems, which suggests that the exact renormalization group is the adequate tool for them.

There is an ongoing debate in cosmology about the value of the length scale at which ``homogeneity'' in the matter distribution is reached or even if such a scale exists. In the wake of this debate, we intend in this letter to clarify the meaning of the statement transition to homogeneity and of the concept of correlation length. We show that there are two scales each associated to the fractal and to the homogeneous regimes of the matter distribution, respectively. The distinction between both scales has deep consequences, for there can be a regime which despite having small fluctuations around the average density exhibits large clusters of galaxies.