Papers
Topics
Authors
Recent
2000 character limit reached

Primordial Curvature Power Spectrum

Updated 4 January 2026
  • The primordial curvature power spectrum is a measure of scalar curvature perturbations generated during inflation, encapsulating quantum fluctuations that seed cosmic structure.
  • It is defined via the Fourier transform of the two-point function of curvature perturbations and is tightly constrained by CMB, large-scale structure, and PBH observations.
  • The spectrum offers insights into inflationary dynamics, spatial curvature effects, and small-scale enhancements linked to primordial black hole formation and gravitational wave signals.

The primordial curvature power spectrum quantifies the two-point statistical properties of scalar curvature perturbations generated during cosmic inflation. It acts as the central observable linking quantum fluctuations of the inflaton to structure formation, cosmic microwave background (CMB) anisotropies, and a vast array of phenomenology across scales from kiloparsecs to sub-parsec domains. Its form, amplitude, and possible features are key in testing inflationary models, constraining early-Universe physics, and interpreting observational results from the CMB, large-scale structure, pulsar timing arrays (PTA), and primordial black hole (PBH) surveys.

1. Definition, Physical Origin, and Mathematical Formalism

The power spectrum of primordial curvature perturbations, denoted PR(k)\mathcal{P}_\mathcal{R}(k) or PR(k)P_\mathcal{R}(k), is defined as the Fourier-space two-point function of the gauge-invariant comoving curvature perturbation R\mathcal{R}: R(k)R(k)=(2π)3δ3(k+k)2π2k3PR(k)\langle \mathcal{R}(\mathbf{k})\mathcal{R}(\mathbf{k}') \rangle = (2\pi)^3 \delta^3(\mathbf{k}+\mathbf{k}') \, \frac{2\pi^2}{k^3} \mathcal{P}_\mathcal{R}(k) In canonical single-field slow-roll inflation, this is computed by quantizing the Mukhanov–Sasaki variable vkv_k described by

vk+[k2z/z]vk=0v_k'' + [k^2 - z''/z] v_k = 0

with z=a2ϵMPlz = a \sqrt{2\epsilon}M_\text{Pl}, aa the scale factor, and ϵ\epsilon the first slow-roll parameter. The Bunch–Davies initial vacuum is adopted, and after horizon exit the curvature perturbation Rk\mathcal{R}_k freezes out, yielding the celebrated slow-roll result: PR(k)=H28π2ϵMPl2k=aH\mathcal{P}_\mathcal{R}(k) = \frac{H^2}{8\pi^2\epsilon M_\text{Pl}^2} \bigg|_{k=aH} Observationally, this spectrum is tightly constrained near the pivot scale k=0.05Mpc1k_* = 0.05\,\mathrm{Mpc}^{-1} with amplitude As2.1×109A_s \simeq 2.1 \times 10^{-9} and spectral index ns0.965n_s \simeq 0.965 (Bastero-Gil et al., 2013, Emami et al., 2017).

2. Extraction and Features from CMB and Large-Scale Structure

On cosmological scales (104k3Mpc110^{-4}\lesssim k \lesssim 3\,\mathrm{Mpc}^{-1}), the power spectrum is reconstructed via a linear mapping from PR(k)\mathcal{P}_\mathcal{R}(k) to observables:

  • CMB temperature/polarization: CXY=4πdkkPR(k)ΔX(k)ΔY(k)C_\ell^{XY} = 4\pi \int \frac{dk}{k} \,\mathcal{P}_\mathcal{R}(k)\Delta_\ell^X(k)\Delta_\ell^Y(k)
  • Galaxy power spectra: Pgal(k,z)=b2(z)Tm2(k,z)2π2k3PR(k)P_\text{gal}(k,z) = b^2(z)T_m^2(k,z)\frac{2\pi^2}{k^3}\mathcal{P}_\mathcal{R}(k)

Regularized non-parametric inversion (Tikhonov, GCV, Mallows CpC_p) reveals robust features:

  • Infrared cutoff at k5×104Mpc1k \lesssim 5\times10^{-4}\,\mathrm{Mpc}^{-1}, corresponding to suppressed low multipoles (quadrupole, octupole) in the CMB
  • Small-scale oscillations or dips with marginal \sim2–3σ\sigma significance at k0.002k\sim0.002, $0.04$, and 0.056Mpc10.056\,\mathrm{Mpc}^{-1} (Hunt et al., 2013, Hunt et al., 2015). Below k0.1Mpc1k\sim0.1\,\mathrm{Mpc}^{-1}, the spectrum agrees with a near–scale-invariant power law (Emami et al., 2017), and no high-significance features are confirmed.

3. Theoretical Impact of Spatial Curvature

Inflation in universes with nonzero spatial curvature (K=±1K=\pm1) modifies both the background evolution and the Mukhanov–Sasaki equation: vk+[k2+Kz/z]vk=0v_k'' + [k^2+K-z''/z]v_k=0 with Laplace–Beltrami eigenvalues k2+Kk^2+K (kZk\in\mathbb{Z} for K=+1K=+1). Curvature induces:

  • Infrared cut-off in PR(k)\mathcal{P}_\mathcal{R}(k) for kKk\lesssim\sqrt{|K|}
  • Superimposed oscillatory features for k(25)Kk \sim (2–5)\sqrt{|K|}
  • Asymptotic recovery of the flat-universe power law for kKk \gg \sqrt{|K|} (Handley, 2019, Thavanesan et al., 2020) For Planck 2018 best-fit ΩK0.009\Omega_K\approx-0.009, the cut-off is at kcut0.002Mpc1k_\text{cut}\approx 0.002\,\mathrm{Mpc}^{-1}, corresponding to the observed low-\ell suppression. The curved-universe template improves χ2\chi^2 fit by Δχ26\Delta\chi^2\approx -6 to 9-9 compared to standard Λ\LambdaCDM (Handley, 2019).

4. Constraints from Primordial Black Hole Production and Evaporation

PBH non-detections and Hawking evaporation yields direct constraints on PR(k)\mathcal{P}_\mathcal{R}(k) at very small scales (k1015Mpc1k \gtrsim 10^{15}\,\mathrm{Mpc}^{-1}). The collapse fraction β(M)\beta(M) is exponentially sensitive to the mass variance σ2(M)\sigma^2(M), related (in radiation domination) by

σ2(M)=dkkW2(kR)1681(kR)4PR(k)\sigma^2(M) = \int \frac{dk}{k} W^2(kR) \frac{16}{81}(kR)^4 \mathcal{P}_\mathcal{R}(k)

Non-detection of sub-solar-mass PBHs in BBN and CMB imposes bounds:

  • k3.7×1017k0.05k\sim 3.7\times10^{17}\,k_{0.05}: PR(k)3×103\mathcal{P}_\mathcal{R}(k) \lesssim 3\times10^{-3}
  • k7.4×1018k0.05k\sim7.4\times10^{18}\,k_{0.05}: PR(k)4×103\mathcal{P}_\mathcal{R}(k) \lesssim 4\times10^{-3} Evaporating PBH relic and neutrino/γ-ray backgrounds further constrain PR(k)2×102\mathcal{P}_\mathcal{R}(k) \lesssim 2\times10^{-2} for 4.5×1018k1.8×1021Mpc14.5\times10^{18}\lesssim k\lesssim 1.8\times10^{21}\,\mathrm{Mpc}^{-1} (Dalianis, 2018, Yang, 2024, Sato-Polito et al., 2019). These bounds fill gaps in previously unconstrained windows and eliminate broad blue-tilted enhancements or features on ultra-small scales.

5. Model-Independent Reconstruction in PTA and GW Observations

PTA collaborations (NANOGrav, PPTA, EPTA, CPTA) have detected a common-spectrum stochastic GW signal, interpretable as scalar-induced GWs sourced by enhanced PR(k)\mathcal{P}_\mathcal{R}(k) at k107 ⁣ ⁣109Mpc1k\sim 10^7\!-\!10^9\,\mathrm{Mpc}^{-1}. Bayesian model-independent reconstruction approaches (linear interpolation over free nodes, polynomial, lognormal, broken power-law) consistently favor:

  • Broad or narrow peaks in PR(k)\mathcal{P}_\mathcal{R}(k) centered at kp106 ⁣ ⁣109Mpc1k_p \sim 10^6\!-\!10^9\,\mathrm{Mpc}^{-1}
  • Peak amplitudes PR(kp)2×102\mathcal{P}_\mathcal{R}(k_p)\sim 2\times10^{-2}, seven orders above CMB normalization
  • No current discrimination between single-peak, double-peak, broken power law, lognormal, or box shapes (Bayes factors <1<1 between models) Best-fit reconstructions are consistent with BBN, μ-distortion, and PBH constraints, requiring the peak to be narrow (Δlnk<1\Delta\ln k < 1) to avoid overproduction of PBHs (Fei, 2023, Yi et al., 2023, You et al., 2023, Fan et al., 2024).
Source kk-range [Mpc1\mathrm{Mpc}^{-1}] Bound or Detection (PR(k)\mathcal{P}_\mathcal{R}(k))
CMB (Planck) 10410^{-4}–$3$ 2.1×1092.1\times10^{-9}
PTA (NANOGrav) 10610^610910^9 102\sim10^{-2} (peak)
PBH evaporation 101510^{15}102010^{20} 2×102\lesssim 2\times10^{-2}

6. Theoretical Mechanisms for Small-Scale Enhancements

Transient features in the inflationary potential (inflection points, ultra-slow-roll phases) or nontrivial sound speed evolution can produce large peaks or steep growth in PR(k)\mathcal{P}_\mathcal{R}(k):

  • Ultra-slow-roll inflation (η6\eta \simeq -6): k4k^4 growth
  • Sudden drop in sound speed: k2k^2 rise, further steepening with simultaneous η\eta and csc_s changes, realizing k6k^6 envelope (Zhai et al., 2023)
  • All viable PBH-producing models must generate a narrow enhancement with PR(kp)102\mathcal{P}_\mathcal{R}(k_p)\lesssim 10^{-2}, width Δ(lnk)1\Delta(\ln k)\lesssim1, and slopes α,β4\alpha,\beta\lesssim 4 in broken power-law templates (Yi et al., 2022, Zheng et al., 2022).

7. Implications, Open Questions, and Outlook

The aggregate constraints demand near–scale invariance at cosmological scales, with only tightly localized peaks at small scales. Any inflationary scenario with broad or persistent blue tilts, slow-roll violation, or non-Gaussian statistics is subject to strong bounds from PBHs and spectral distortions. Detection of a cutoff or oscillatory features from spatial curvature would point toward “just-enough” inflation and nonflat initial conditions (Handley, 2019, Thavanesan et al., 2020). Current and future GW/PTA experiments promise to decisively probe the relevant PR(k)\mathcal{P}_\mathcal{R}(k) window, test the origin of BBH mergers, and further elucidate inflationary microphysics on scales previously inaccessible to cosmology.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (16)

Whiteboard

Topic to Video (Beta)

Follow Topic

Get notified by email when new papers are published related to Primordial Curvature Power Spectrum.