betaFeedback

WorksPhysics and Astronomy

Cosmological Backreaction and the Separate Universe Conjecture: Dynamical volume partitioning of Newtonian cosmologies

Abstract

We provide a description of general Newtonian cosmologies for collisionless matter in terms of an exact and dynamical volume partitioning of global realizations of Newtonian models. Employing the formalism for the description of cosmological backreaction, an interaction between the inhomogeneities of the formation of structures in the Universe and their average properties can be analyzed. Our results can be regarded as a tool for many potential applications in the context of Newtonian simulations, but they also provide an exact way to assume analytical properties of the universe models on regional scales in conformity with global integral constraints on their average properties. We discuss approaches in relation to `separate Universe conjectures' in the literature and examine their consistency with the exact partitioning framework. In particular, using the Average Expansion Rate Approximation (AvERA) as an example, we show that identifying locally assigned Friedmannian rates with physical regional expansion rates is constrained by the global Newtonian 3-torus condition. Alternatively, when the simulation cells are interpreted as non-material control volumes, Reynolds transport introduces mass flux and flux-backreaction terms, so that their dynamics is not in general that of independent Friedmann domains. We also discuss the corresponding consistency conditions for stable clustering prescriptions.

Introduction

The cosmological backreaction connects the formation of structures and the expansion history of the Universe [7]. In order to quantitatively examine this connection, we investigate an exact Lagrangian volume partitioning of the universe model into spatial regions at different scales, e.g. for the two-scale problem into overdense (’virialized’) and underdense (void) regions. In general relativity the former are characterized by positive spatial scalar curvature and the latter by negative spatial scalar curvature. Due to the volume dominance of the latter regions, starting at the onset of structure formation, together with the dynamical non-conservation of spatial curvature [11], the emergence of negative scalar curvature on average is expected leading to a dark energy-like effect [14].

In this work we concentrate on inhomogeneous Newtonian cosmologies with the motivation to establish a controlled analytical way of making assumptions on some regional scale of the problem, and with the perspective of providing tools for a corresponding general-relativistic treatment. The main reason to start within Newtonian cosmology is that we can globally control cosmological backreaction: vanishing spatial curvature together with the global topological architecture of a 3–torus ensures vanishing of the kinematical backreaction term that is necessary and sufficient for vanishing averages of the deviations from a global background cosmology [13]. Backreaction might be globally non-vanishing only for non-Euclidean background spacetimes and non-Euclidean global topologies.

In Newtonian cosmology, essentially two ingredients are needed to render the equations of Newtonian gravity unique: (i) introducing a global background and (ii) subjecting deviations thereof to periodic boundary conditions. This agrees with the common architecture of a Newtonian simulation in cosmology. Exact average properties can then be subjected to a global constraint, investigated in the fluid approximation for a ’dust matter model’ in the mean-field approximation of self-gravitating collisionless matter in [13] (see also [22]). The Vlasov-Poisson dynamics of the mean-field approximation is briefly presented in appendix A, and the Buchert-Ehlers theorem is reviewed in appendix B. We will then consider a multi-scale partitioning of the global Newtonian model, based on previous work in the framework of general relativity [11] and [39].

The result (B.22) reviewed in appendix B is valid for the average of the matter model ‘dust’ on all compact domains from an infinitesimal domain up to any scale, e.g., up to the boundary of a compact and boundary-free universe model. We will in what follows include velocity dispersion that has been shown to not violate this theorem in [10] (see also the discussion by [27]). We so extend the results to a general collisionless matter model. Deviations from mean-field gravity are not considered in this paper but would furnish an interesting generalization possibility along the lines of the explicit coarse-graining investigated in [12], or along the lines of the BBGKY hierarchy advanced in [37], or by including noise [15, 28]. Our results are presented for scalar variables; an extension to vector- and tensor-valued variables is straightforward within Newtonian cosmology, based on the seminal paper on the tensor virial conjecture by [19], since averaging of tensors is well-defined in Euclidean space.

In section II, in order to provide a self-contained presentation, we briefly lay down the formalism of self-gravitating collisionless systems to establish the foundations on which this paper is based. Readers familiar with these foundations may skip this section including appendix A and move to section III, where we partition the system to obtain an exact description of underdense and overdense regions and their nonlinear dynamical interaction, and we consider physical conditions to close this family of interacting systems of partitioning equations. We then discuss relations to ‘separate Universe conjectures’ in the literature in section IV, and we end with a discussion in section V. Appendices are devoted to velocity moments of Vlasov’s equation and to known theorems and their generalization to include velocity dispersion.

The Euler-Newton-Jeans system and spatial averages of its trace parts

We consider a Galilei-Newton spacetime, foliated into flat spatial hypersurfaces t=const.⁡t = \operatorname{const.}, labeled by the Newtonian absolute time coordinate tt. We also consider a non-rotational, inertial coordinate system in this spacetime with Euclidean geometry, realized by Cartesian coordinates xix_i, i=1,2,3i = 1, 2, 3.

Self-gravitating collisionless systems in Newtonian gravity

In this section, we consider a continuous self-gravitating and collisionless system within Newtonian gravity. The governing equations are re-derived in appendix A. We start by presenting the Euler-Newton-Jeans system, discuss its trace parts and the equation of state we obtain.

The Euler-Newton-Jeans system

In Appendix A, we develop the calculations to obtain the first members of a hierarchy of velocity moment equations of Vlasov’s equation that governs the dynamics of a self-gravitating collisionless continuum in Newtonian gravity in the mean-field approximation. These first members of the hierarchy are composed of the continuity equation for the mass density field ϱ(xi,t)\varrho(x_i,t), the Euler-Jeans equation for the mean velocity field components vˉi(xi,t)\bar{v}_i(x_i,t), the evolution equation for the reduced second moment tensor components Πij(xi,t)\Pi_{ij}(x_i,t), and the gravitational field equations for the gravitational field strength components in the mean-field approximation gˉi(xi,t)\bar{g}_i(x_i,t):4

ddtϱ=−ϱvˉj,j ;\frac{\mathrm{d}}{\mathrm{d}t}\varrho=-\varrho\bar{v}_{j,j}\,;
ddtvˉi=gˉi−1ϱΠij,j ;\frac{\mathrm{d}}{\mathrm{d}t}\bar{v}_i=\bar{g}_i-\frac{1}{\varrho}\Pi_{ij,j}\,;
ddtΠij=−Lijk,k−[vˉi,kΠjk+vˉj,kΠki+vˉk,kΠij] ;\frac{\mathrm{d}}{\mathrm{d}t}\Pi_{ij}=-L_{ijk,k}-\left[\bar{v}_{i,k}\Pi_{jk}+\bar{v}_{j,k}\Pi_{ki}+\bar{v}_{k,k}\Pi_{ij}\right]\,;
gˉj,i−gˉi,j=0 ;\bar{g}_{j,i}-\bar{g}_{i,j}=0\,;
gˉk,k=Λ−4πGϱ.(1)\bar{g}_{k,k}=\Lambda-4\pi G\varrho. \tag*{(1)}

In this system, d/dt≡∂/∂t+vˉ⋅∇\mathrm{d}/\mathrm{d}t \equiv\partial/\partial t+\bar{\mathbf{v}}\cdot\nabla denotes the Lagrangian derivative operator. The field equations can be rewritten in terms of the gravitational potential Φ\Phi defined by gˉi≔−Φ,i\bar{g}_i \coloneqq-\Phi_{,i}, obeying the Poisson equation:

ΔΦ=4πGϱ−Λ.(2)\Delta\Phi=4\pi G\varrho-\Lambda. \tag*{(2)}

This system of equations is not closed unless the divergence of the third reduced velocity moment tensor LijkL_{ijk} is neglected, or its relation to other variables is given. We will call the above system the Euler-Newton-Jeans system when dropping the third reduced velocity moment.

Discussion of the trace parts and thoughts on the equation of state

The Euler-Jeans equation can be rewritten as:

∂∂tvˉi+vˉk∂∂xkvˉi=gˉi−1ϱ∂∂xjΠij.(3)\frac{\partial}{\partial t}\bar{v}_i+\bar{v}_k\frac{\partial}{\partial x_k}\bar{v}_i=\bar{g}_i-\frac{1}{\varrho}\frac{\partial}{\partial x_j}\Pi_{ij}. \tag*{(3)}

We introduce a short-hand notation for the divergence of the symmetric tensor Π\boldsymbol{\Pi}:

Ψi≡∂∂xjΠij.(4)\Psi_i\equiv\frac{\partial}{\partial x_j}\Pi_{ij}. \tag*{(4)}

We perform a spatial derivative with respect to xjx_j and we take the trace of the equation to obtain:

ddt∂vˉi∂xi+∂vˉk∂xi∂vˉi∂xk=∂gˉi∂xi−∂∂xi(1ϱΨi).(5)\frac{\mathrm{d}}{\mathrm{d}t}\frac{\partial\bar{v}_i}{\partial x_i}+\frac{\partial\bar{v}_k}{\partial x_i}\frac{\partial\bar{v}_i}{\partial x_k}=\frac{\partial\bar{g}_i}{\partial x_i}-\frac{\partial}{\partial x_i}\left(\frac{1}{\varrho}\Psi_i\right). \tag*{(5)}

We then rewrite this equation in terms of kinematical invariants, splitting the velocity gradient into a symmetric and an antisymmetric part:

∂vˉi∂xj=12(∂vˉi∂xj+∂vˉj∂xi)+12(∂vˉi∂xj−∂vˉj∂xi)=13θδij+σij+ωij,(6)\frac{\partial\bar{v}_i}{\partial x_j}=\frac{1}{2}\left(\frac{\partial\bar{v}_i}{\partial x_j}+\frac{\partial\bar{v}_j}{\partial x_i}\right)+\frac{1}{2}\left(\frac{\partial\bar{v}_i}{\partial x_j}-\frac{\partial\bar{v}_j}{\partial x_i}\right)=\frac{1}{3}\theta\delta_{ij}+\sigma_{ij}+\omega_{ij}, \tag*{(6)}

where in the second equality we introduced θ\theta representing the trace (the rate of expansion), σij\sigma_{ij} a symmetric traceless tensor (the shear), and ωij\omega_{ij} an antisymmetric tensor (the vorticity). We furthermore define the rate of shear σ\sigma and the rate of vorticity ω\omega as follows:

σ2≡12σijσij ;ω2≡12ωijωij.(7)\sigma^2\equiv\frac{1}{2}\sigma_{ij}\sigma_{ij}\ ;\qquad\omega^2\equiv\frac{1}{2}\omega_{ij}\omega_{ij}. \tag*{(7)}

We then define the two of the scalar principal invariants of the mean velocity gradient:

I=∇⋅vˉ=∂vˉi∂xi=θ ;(8)\mathrm{I}=\nabla\cdot\bar{\mathbf{v}}=\frac{\partial\bar{v}_i}{\partial x_i}=\theta\ ; \tag*{(8)}
II=12∇⋅(vˉ(∇⋅vˉ)−(vˉ⋅∇)vˉ)=12(∂vˉi∂xi∂vˉj∂xj−∂vˉi∂xj∂vˉj∂xi)=13θ2+ω2−σ2.(9)\begin{aligned} \mathrm{II}&=\frac{1}{2}\nabla\cdot\left(\bar{\mathbf{v}}(\nabla\cdot\bar{\mathbf{v}})-(\bar{\mathbf{v}}\cdot\nabla)\bar{\mathbf{v}}\right)\\ &=\frac{1}{2}\left(\frac{\partial\bar{v}_i}{\partial x_i}\frac{\partial\bar{v}_j}{\partial x_j}-\frac{\partial\bar{v}_i}{\partial x_j}\frac{\partial\bar{v}_j}{\partial x_i}\right)=\frac{1}{3}\theta^2+\omega^2-\sigma^2 . \tag*{(9)} \end{aligned}

Equation (5) then takes the form of Raychaudhuri’s equation:

dθdt=Λ−4πGϱ−13θ2+2(ω2−σ2)−∂∂xi(1ϱΨi).(10)\frac{\mathrm{d}\theta}{\mathrm{d}t}=\Lambda-4\pi G\varrho-\frac{1}{3}\theta^2+2(\omega^2-\sigma^2)-\frac{\partial}{\partial x_i}\left(\frac{1}{\varrho}\Psi_i\right). \tag*{(10)}

If we assume that typical velocities of the system stay in the vicinity of the mean velocity, ∣v−vˉ∣≃ϵvˉ|\boldsymbol{v}-\bar{\boldsymbol{v}}|\simeq\epsilon\bar{v}, we can estimate the moments of the hierarchy:

Πij∼ϱϵ2∣vˉ∣2;(11)\Pi_{ij}\sim\varrho\epsilon^2|\bar{\boldsymbol{v}}|^2; \tag*{(11)}
Lijk∼ϱϵ3∣vˉ∣3,(12)L_{ijk}\sim\varrho\epsilon^3|\bar{\boldsymbol{v}}|^3, \tag*{(12)}

References

  1. [1]Al Roumi F., Buchert T., Wiegand A., 2017: Lagrangian theory of structure formation in relativistic cosmology. IV. Lagrangian approach to gravitational waves. Phys. Rev. D 96 123538 (arXiv:1711.01597)
  2. [2]Bernardeau F., 1994: The Nonlinear Evolution of Rare Events Astrophys. J. 427 51 (arXiv:astro-ph/9311066)DOI
  3. [3]Binney J., Tremaine S., 2008: Galactic Dynamics, 2nd edition, Princeton Series in Astrophysics, Princeton Univ. Press
  4. [4]Bolejko K., 2017: Cosmological backreaction within the Szekeres model and emergence of spatial curvature JCAP 06 025 (arXiv:1704.02810)DOI
  5. [5]Bolejko K., 2018: Relativistic numerical cosmology with Silent Universes Class. Quant. Grav. 35 024003 (arXiv:1708.09143)DOI
  6. [6]Buchert T., 2000: On average properties of inhomogeneous fluids in General Relativity I: Dust Cosmologies Gen. Rel. Grav. 32 105 (arXiv:gr-qc/9906015)DOI
  7. [7]Buchert T., 2008: Dark Energy from structure: a status report Gen. Rel. Grav. 40 467 (arXiv:0707.2153)DOI
  8. [8]Buchert T., 2011: Toward physical cosmology: focus on inhomogeneous geometry and its non-perturbative effects Class. Quant. Grav. 28 164007 (arXiv:1103.2016)DOI
  9. [9]Buchert T., 2016: Lecture Notes Master 2: Cosmology and Gravitational Systems, École Normale Supérieure, Lyon
  10. [10]Buchert T., 2018: On Backreaction in Newtonian cosmology MNRAS 473 L46 (arXiv:1704.00703)DOI
  11. [11]Buchert T., Carfora M., 2008: On the curvature of the present-day universe Class. Quant. Grav. 25 195001 (arXiv:0803.1401)DOI
  12. [12]Buchert T., Domínguez A., 2005: Adhesive Gravitational Clustering A&A 438 443 (arXiv:astro-ph/0502318)
  13. [13]Buchert T., Ehlers J., 1997: Averaging inhomogeneous Newtonian cosmologies A&A 320 1 (arXiv:astro-ph/9510056)
  14. [14]Buchert T., Rä sän en S., 2012: Backreaction in late-time cosmology Ann. Rev. Nucl. Part. Sci. 62 57 (arXiv:1112.5335)
  15. [15]Buchert T., Domínguez A., Pérez-Mercader J., 1999: Extending the scope of models for large-scale structure formation in the Universe A&A 349 43 (arXiv:astro-ph/9709218)
  16. [16]Buchert T., Delgado Gaspar I., Ostrowski J.J., 2022: On general-relativistic Lagrangian perturbation theory and its non-perturbative generalization Universe 8 583 (arXiv:2209.13417)
  17. [17]Buchert T., Kerscher M., Sicka C., 2000: Backreaction of inhomogeneities on the expansion of cosmological parameters Phys. Rev. D 62 043525 (arXiv:astro-ph/9912347)DOI
  18. [18]Carr B.J., Harada T., 2015: Separate universe problem: 40 years on Phys. Rev. D 91 084048 (arXiv:1405.3624)DOI
  19. [19]Chandrasekhar S., Lee E.P., 1968: A Tensor Virial-Equation for Stellar Dynamics MNRAS 139 135DOI
  20. [20]Dai L., Pajer E., Schmidt F., 2015: On Separate Universes JCAP 10 059 (arXiv:1504.00351)DOI
  21. [21]Dai L., Pajer E., Schmidt F., 2015: Conformal Fermi Coordinates JCAP 11 043 (arXiv:1502.02011)DOI
  22. [22]Ehlers J., Buchert T., 1997: Newtonian Cosmology in Lagrangian Formulation: Foundations and Perturbation Theory Gen. Rel. Grav. 29 733 (arXiv:astro-ph/9609036)DOI
  23. [23]Ehlers J., Rienstra W., 1969: The locally isotropic solutions of the Liouville and Poisson equations Astrophys. J. 155 105
  24. [24]Fardeau N., Buchert T., Al Roumi F., Felgery F., 2023: Non-perturbative collapse models for collisionless self-gravitating flows Phys. Rev. D 108 083502 (arXiv:2301.09154)DOI
  25. [25]Flanders H., 1973: Differentiation under the Integral Sign American Mathematical Monthly, vol. 80 (June-July 1973), p. 615-627DOI
  26. [26]Galoppo M., Buchert T., Mourier P., 2026: Backreaction and the role of spatial curvature in the cosmic neighborhood Astrophys. J. 1002 L39 (arXiv:2605.05454)DOI
  27. [27]Kaiser N., 2017: Why there is no Newtonian backreaction MNRAS 469 744 (arXiv:1703.08809)DOI
  28. [28]Lapi A., Haridasu B.S., Boco L., Cueli M.M., Baccigalupi C., Danese L., 2025: Little ado about everything II: an ‘emergent’ dark energy from structure formation to rule cosmic tensions JCAP 04 015 (arXiv:2502.05823)
  29. [29]Mecke K.R., Buchert T., Wagner H., 1994: Robust morphological measures for large-scale structure in the Universe A&A 288 697 (arXiv:astro-ph/9312028)DOI
  30. [30]Oestreicher A., Koksbang S. M., 2024: Backreaction in Numerical Relativity: Averaging on Newtonian gauge-like hypersurfaces in Einstein Toolkit cosmological simulations Open J. of Astrophysics 7 92 (arXiv:2408.03049)DOI
  31. [31]Olson D.W., Sachs R.K., 1973: The Production of Vorticity in an Expanding, Self-Gravitating Fluid Astrophys. J. 185 91DOI
  32. [32]Peebles P.J.E., 1980: The Large-scale Structure of the Universe, Princeton Univ. Press
  33. [33]Rácz G., et al., 2017: Concordance cosmology without dark energy MNRAS 469 L1 (arXiv:1607.08797)DOI
  34. [34]Rácz G., et al., 2018: Compactified cosmological simulations of the infinite universe MNRAS 477 1949 (arXiv:1711.04959)DOI
  35. [35]Reynolds O., 1903: Papers on Mechanical and Physical Subjects. Vol. 3, The Sub-Mechanics of the Universe Cambridge: Cambridge University Press. pp. 12-13
  36. [36]Ruchaud B.F., 2018: Replacing dark energy by silent virialisation A&A 610 A51 (arXiv:1706.06179)
  37. [37]Tremblin P., Chabrier G., Padioleau T., Daley-Yates S., 2022: Non-ideal self-gravity and cosmology: the importance of correlations in the dynamics of the large-scale structures of the Universe A&A 659 A108 (arXiv:2109.09087)DOI
  38. [38]Vigneron Q., Buchert T., 2019: Dark Matter from Back-reaction? Collapse models on galaxy cluster scales Class. Quant. Grav. 36 175006 (arXiv:1902.08441)DOI
  39. [39]Wiegand A., Buchert T., 2010: Multiscale cosmology and structure-emerging Dark Energy: A plausibility analysis Phys. Rev. D 82 023523 (arXiv:1002.3912)DOI
  40. [40]Wiltshire D.L., 2007: Exact Solution to the Averaging Problem in Cosmology Phys. Rev. Lett. 99 251101 (arXiv:0709.0732)DOI
  41. [41]Wiltshire D.L., 2014: Cosmic structure, averaging and dark energy In: Novello, M., Perez Bergliaffa, S.E., eds., Cosmology and Gravitation: Proc. XXth Brazilian School, Cambridge Scientific Publishers, (arXiv:1311.3787)arxiv.org/abs/1311.3787

Paper details

vlnc:W_22cfxgct
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 7

downloads · 2

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Claude Opus 5 (bare)

View statement
Looking for related papers…
Contents