Cosmic Skewness in Cosmology

• Cosmic skewness is defined as the reduced third moment of the density contrast, quantifying asymmetry in the cosmic field distribution.
• It serves as a crucial probe of nonlinear gravitational dynamics, dark energy effects, and BAO signatures through its scale and redshift dependence.
• Measurement techniques such as counts-in-cells and the skew spectrum offer robust tools for refining cosmological parameter constraints.

Cosmic skewness is the third reduced moment of cosmic fields, most notably the one-point probability distribution of the matter density contrast. It is a fundamental statistic in cosmology, encoding leading-order departures from Gaussianity in the evolving cosmic web. Its scale, redshift, tracer, and application dependencies make it a central probe for nonlinear structure formation, gravitational dynamics, and fundamental physics across multiple domains of astrophysics and cosmology.

1. Definitions and Theoretical Formulation

Cosmic skewness, typically denoted $S_3$, quantifies the normalized third central moment of the one-point distribution of a cosmological field. For the matter density contrast $\delta = \rho / \langle \rho \rangle - 1$, the key definitions are:

• Variance: $$\sigma^2 = \langle (\delta - \langle \delta \rangle)^2 \rangle = \langle \delta^2 \rangle$$
• Skewness (mathematical): $$S = \langle (\delta - \langle \delta \rangle)^3 \rangle / \sigma^3 = \langle \delta^3 \rangle / \sigma^3$$
• Kurtosis (mathematical): $$K = \langle (\delta - \langle \delta \rangle)^4 \rangle/\sigma^4 - 3$$
• Cosmological/reduced skewness: $$S_3 \equiv \langle \delta^3 \rangle / \langle \delta^2 \rangle^2 = S / \sigma$$

These definitions extend naturally to smoothed fields $\delta_R$ via convolution with a window $W_R(|\mathbf{x}|)$, where the smoothing scale $R$ is a critical parameter.

In Gaussian initial conditions, $S_3 = 0$. Gravitational instability rapidly generates nonzero skewness, making $S_3$ a leading probe of cosmic non-Gaussianity (Einasto et al., 2020).

Numerical and perturbative approaches consistently show $S_3$ as a function of validation choices: real-space smoothing, Fourier domain filtering, and chosen moments (variance, skewness) (Dai et al., 2020, Einasto et al., 2020). For analytic models, in leading-order perturbation theory for an Einstein–de Sitter universe, $S_3^{\mathrm{EdS}} = 34/7 \approx 4.857$ (Velten et al., 2019, Fazolo et al., 2022, Kitaura, 2010).

2. Evolution, Scale Dependence, and Non-Universality

The functional dependence of cosmic skewness $S_3$ on scale, epoch, and smoothing filter is highly nontrivial and departs significantly from the universality seen in other cosmological statistics.

• Evolutionary tracks: At fixed smoothing scale $R_t$, $S_3(\sigma)$ increases steeply with $\sigma$ up to $\sim 2$, peaks, and then gradually declines at higher $\sigma$. Conversely, at fixed redshift, $S_3(\sigma)$ grows sharply for $\sigma \lesssim 1$ and plateaus for $\sigma \gtrsim 2$. The early-time ($z \to 30$ or $\sigma \to 0$) asymptote is $S_3 \to 3 \pm 0.5$, with corresponding kurtosis $S_4 \to 15 \pm 3$ (Einasto et al., 2020).
• Analytic fits: The $S_3$ evolution is captured by a four-parameter phenomenological form:

$$S_3(\sigma, z) = S_{\min}(z) + \frac{C_1(z) \sigma^{5/2}}{1 + C_2(z) \sigma^{5/2}}$$

with power-law redshift evolution for the amplitudes ($S_3,1,0 = 4.9$, $S_3,\max,0 = 9.4$ for $\Omega_m=0.307, h=0.7, \sigma_8=0.828$), and the formula reproduces $S_3(\sigma, z)$ to $\sim 5\%$ over $0.05 \lesssim \sigma \lesssim 5$, $0 \lesssim z \lesssim 20$ (Einasto et al., 2020).

• Non-universality: Unlike other cosmic statistics (halo mass function, concentration-mass), $S_3(\sigma)$ is not universal—its value depends independently on smoothing length and redshift, due to the strongly nonlinear and non-Gaussian gravitational dynamics. The collapse to a single-parameter function of $\sigma$ fails for $S_3$ (Einasto et al., 2020).

Perturbative and lognormal approximations (e.g., $S_3 = 34/7 + \gamma$ or $S_3 = 3 + \sigma^2$) are accurate solely in the linear regime ($\sigma \lesssim 0.1$), but break down at mildly nonlinear and nonlinear scales (Einasto et al., 2020).

3. Measurement Methodologies and Skew-Spectrum Formalism

Direct measurement of the cosmic skewness can be performed via moments of the one-point distribution, counts-in-cells, or via a range of higher-order Fourier statistics.

• Counts-in-cells: Traditional approaches utilize spatial binning and direct measurement of $S_3$ via volume or pixel averages (Einasto et al., 2020, Ben-David et al., 2015).
• Skew spectrum: A powerful alternative is the skew spectrum, the cross-power spectrum of the density and its squared or filtered counterpart:

$$P_\mathrm{skew}(k) = \langle \delta(\mathbf{k}) [\delta_R^2](\mathbf{k'}) \rangle$$

which compresses bispectrum information, providing nearly full three-point constraining power with significantly reduced computational cost and tractable covariance estimation (Dai et al., 2020, Hou et al., 2022, Hou et al., 2024). The skew spectrum is maximally sensitive to primordial non-Gaussianity (probed via $f_{\rm NL}^{\rm loc}$), breaking degeneracies in parameter inference.

• Simulation constraints: Large suites of $N$-body simulations (e.g., Quijote, GLAM) allow high-precision construction of $S_3(\sigma, z)$ and skew spectra, with full joint covariance with the power spectrum enabling robust cosmological inference (Einasto et al., 2020, Hou et al., 2022, Hou et al., 2024).
• CMB analysis: Cosmic skewness is also measured in CMB data, e.g., via skewness power spectra (two-to-one cumulant correlators), with direct ties to theoretical bispectra and dedicated filtering to separate contributions from primordial, point-source, and lensing-secondary sources (0907.4051, Wilson et al., 2012).

4. Physical Origins and Astrophysical Applications

Cosmic skewness is generated by fundamentally nonlinear gravitational dynamics and is further shaped by baryonic, relativistic, and exotic physical processes:

• Structure formation: Asymmetry between over- and under-densities (gravitational collapse vs. void development) produces persistent skewness, even at early epochs ($z \sim 30$), with high-$\sigma$ (small-scale) modes departing first from initial Gaussianity (Einasto et al., 2020).
• Dark energy: Skewness is a sensitive probe of dark-energy clustering. In smooth dark energy models, $S_3$ remains near the $\Lambda$CDM value ($S_3 \sim 5$ for Planck-preferred parameters), while clustering dark energy or low sound-speed models can raise $S_3$ to $15-20$—a factor $\sim 3$ enhancement that is robust against other source uncertainties, providing a “smoking gun” for new physics (Velten et al., 2019, Fazolo et al., 2022).
• Baryon Acoustic Oscillations (BAO): The scale dependence of $S_3$ encodes detectable BAO signatures; the BAO “wiggle” in $S_3(R)$ is $\sim 3\%$ in amplitude, at $R_\mathrm{peak} \sim 82\,h^{-1}\,\mathrm{Mpc}$ for WMAP7 cosmology (Juszkiewicz et al., 2012).
• CMB and high-order statistics: Skewness in the CMB and foreground maps provides a stringent probe of primordial non-Gaussianity, foreground structure, and astrophysical contaminants. Skewness-based statistics help isolate B-mode polarization regions and enable “local” assessments of non-Gaussian point-source contamination (Ben-David et al., 2015, 0907.4051, Wilson et al., 2012).
• Relativistic and projection effects: In distance-redshift measurements, lensing-induced fluctuations lead to a negative skewness in the Hubble diagram ($S_3 \sim -0.4$ to $-1.2$ for $z \sim 0.5-1$), directly probing the late-time matter bispectrum (Schiavone et al., 2023).
• Other regimes: Skewness plays fundamental roles in non-Gaussianity of 21-cm fields during the Epoch of Reionization, cosmic shear three-point functions and aperture mass statistics, non-Gaussianity induced by cosmic strings, early universe bubbles, and stochastic gravitational wave background analyses (Ma et al., 2023, Gomes et al., 14 Jan 2026, Yamauchi et al., 2010, Sugimura et al., 2012, Fujimoto et al., 1 Feb 2026).

5. Impact on Cosmological Inference and Survey Design

Inclusion of cosmic skewness and three-point information (via $S_3$, the bispectrum, or the skew spectrum) significantly strengthens cosmological parameter inference:

• Parameter constraints: Addition of the skew spectrum to power spectrum analyses sharpens marginalized error contours by $31$–$71\%$ for key parameters ($\Omega_m$, $\Omega_b$, $h$, $n_s$, $f_{\rm NL}^{\rm loc}$, $M_\nu$). The combined approach is equivalent to a factor $\sim 3$ gain in survey volume for primordial non-Gaussianity (Dai et al., 2020, Hou et al., 2022, Hou et al., 2024).
• Degeneracy breaking: Skewness-based statistics efficiently break degeneracies between amplitude and bias, and, when combined with two-point functions, self-calibrate higher-order bias terms—crucial for next-generation (DESI, Euclid, LSST, SPHEREx) surveys (Gomes et al., 14 Jan 2026, Hou et al., 2022, Hou et al., 2024).
• Non-Gaussian likelihoods: Realistic cosmological likelihood analyses must account for the non-Gaussianity of the power spectrum estimator itself at nonlinear scales, which arises from trispectrum and pentaspectrum contributions, manifesting as excess skewness beyond the Gaussian case. Analytical expressions for the resulting skewness enable robust likelihood construction for Euclid and comparable surveys (Collaboration et al., 11 Nov 2025).
• Systematics and foregrounds: Skewness is less sensitive to linear bias and redshift-space distortions than two-point statistics, though baryonic processes and nonlinear biasing must be carefully modeled at smaller scales (Juszkiewicz et al., 2012, Velten et al., 2019).
• Observational requirements: Detecting BAO or dark energy signatures in $S_3$ at few-percent levels demands survey volumes $\gtrsim 1$–$5\,(\mathrm{Gpc}/h)^3$ and galaxy number densities $\gtrsim 10^{-4}\,h^3\,\mathrm{Mpc}^{-3}$, with shot noise and window function corrections under control (Juszkiewicz et al., 2012).

6. Limitations of Theoretical Models and Current Challenges

Standard analytic models (Eulerian perturbation theory, lognormal models) fail to capture the redshift and scale dependence of cosmic skewness at nonlinear scales:

• Breakdown of perturbation theory: Perturbation theory is valid only for $\sigma \lesssim 0.1$; beyond this the neglect of shell crossing, higher-order and nonlocal contributions leads to qualitative mispredictions (Einasto et al., 2020, Kitaura, 2010).
• Lognormal model inadequacy: Lognormal-based predictions lack explicit redshift dependence and fail to model turnover and decline of $S_3$ with increasing $\sigma$ at fixed $R_t$ (Einasto et al., 2020).
• Necessity for simulation-based inference: Accurate quantitative characterizations of skewness at all relevant scales require extensive $N$-body or hydrodynamic simulations, often necessitating simulation-based inference frameworks (e.g., normalizing flows as in SimBIG) (Hou et al., 2024).
• Astrophysical complexity: Baryonic feedback, galaxy formation processes, and multi-tracer bias impact measured skewness, demanding self-consistent joint modeling with power spectrum and higher moments (Velten et al., 2019, Gomes et al., 14 Jan 2026, Collaboration et al., 11 Nov 2025).

7. Broader Applications and Future Prospects

Cosmic skewness plays a central role in emerging frontiers:

• Epoch of Reionization: Skewness in 21-cm fields (smoothed skewness $\Gamma(k)$, skew spectrum $S(k)$) encodes sign-changing features corresponding to Ly$\alpha$ pumping, X-ray heating, and ionization bubble growth, with forecast S/N ratios greatly exceeding those for the bispectrum in experiments like SKA1-low (Ma et al., 2023).
• CMB and primordial physics: Local inhomogeneous or direction-dependent skewness can signal exotic scenarios (e.g., non-Gaussian “bubbles”) and distinguish them from standard local-$f_{\rm NL}$ configurations (Sugimura et al., 2012).
• Stochastic gravitational-wave backgrounds: The skewness of the PTA Hellings–Downs curve quantifies intrinsic non-Gaussianity due to finite source populations and enables direct inference of source discreteness (Fujimoto et al., 1 Feb 2026).
• Cosmic shear and weak lensing: Mass aperture skewness ($\langle M_{\rm ap}^3 \rangle$), its sensitivity to intrinsic alignments, and its joint use with two-point functions provide robust self-calibration avenues for lensing surveys (Gomes et al., 14 Jan 2026).

The convergence of computational advances, high-precision survey data, and increasingly accurate theoretical and simulation-based modeling positions cosmic skewness as a critical probe in the cosmological parameter space, structure formation, and tests of fundamental physics. However, the complexity of its redshift-, scale-, and tracer-dependence, and sensitivity to nonlinear and non-Gaussian dynamics, present ongoing challenges for both modelers and observers.