Superoscillatory initial states during inflation: theory, CMB constraints, and prospects for galaxy clustering

Technical Summary: Superoscillatory Initial States during Inflation

Problem Statement
Standard inflationary cosmology assumes that primordial fluctuations originate in the Bunch–Davies (BD) vacuum, yielding a smooth, nearly scale-invariant curvature spectrum. However, deviations from this vacuum—motivated by the trans-Planckian problem or excited initial states—can introduce oscillatory features into the primordial power spectrum. While generic excited-state models allow for arbitrary localized excitations, they often lack a rigorous derivation from first principles or a specific mechanism for the spectral shape. This paper addresses the need for a concrete, theoretically grounded realization of such initial states that incorporates the phenomenon of superoscillation: a mechanism where a band-limited function oscillates locally faster than its fastest Fourier component.

Methodology
The authors construct an explicit theoretical framework for Superoscillatory Initial States (SIS) using a boundary-action formalism within the effective field theory (EFT) of inflation.

1. Boundary Action Derivation:

   • The authors supplement the standard bulk action for the Mukhanov–Sasaki variable $v$ with a quadratic boundary term at the initial time surface $\tau = \tau_*$.
   • Varying the total action yields a Robin boundary condition, which determines the Bogoliubov coefficients ($\alpha_k, \beta_k$) relating the initial state to the BD vacuum.
   • By choosing a specific deformation of the boundary kernel $\kappa_k$, they derive a Bogoliubov coefficient $\beta_k$ that exhibits a superoscillatory structure. This is achieved by defining a dimensionless variable $x = (k-k_*)/\Delta k$ and utilizing the superoscillatory function $S_N(x) = [\cos(x/N) + ia \sin(x/N)]^N$ (where $a>1$).
   • The resulting $\beta_k$ is spectrally localized around a scale $k_*$ with a Gaussian envelope and a rapidly winding phase $\Phi(k) \propto a(k-k_*)/\Delta k$.

2. Primordial Spectrum Construction:

   • The primordial curvature spectrum is derived as $P_\zeta(k) = P^{BD}_\zeta(k) |1+\beta_k|^2$.
   • To leading order in the excitation amplitude $\lambda$, the spectrum features a linear oscillatory term ($\propto \lambda \cos \Phi$) and a quadratic smooth bump ($\propto \lambda^2$).
   • Consistency conditions are imposed, specifically the Gaussian dominance condition ($N > a^2 - 1$), ensuring the superoscillatory patch does not grow exponentially outside the localized region.

3. Observational Projection:

   • CMB: The authors project the primordial feature onto Cosmic Microwave Background (CMB) angular power spectra ($C_\ell$). They perform a convolution with radiation transfer functions, explicitly accounting for transfer-function smearing. They compare analytic Gaussian approximations against full Boltzmann computations using CAMB.
   • Large Scale Structure (LSS): They analyze the imprint on the galaxy power spectrum $P_g(k)$, noting that the matter transfer function $T(k)$ is smooth at the relevant scales, preserving the full oscillatory structure without the severe smearing affecting the CMB.
   • Higher-Order Statistics: Predictions for the bispectrum (non-Gaussianity) are derived using the in-in formalism, distinguishing between equilateral and folded triangle configurations.

Key Results

• Transfer-Function Smearing: The analysis reveals that the CMB transfer function significantly suppresses the oscillatory component of the SIS signal. Full CAMB calculations show that simple analytic Gaussian approximations overestimate the peak signal by a factor of approximately three.
• Observability Window: Due to smearing, there is a strict window for the superoscillatory parameter $a$ to be detectable in the CMB: $1 < a \lesssim 3\text{--}4$. For $a \gtrsim 4$, the oscillatory signal is effectively erased.
• Planck Constraints: Using Planck 2018 TT data, the authors derive an indicative matched-filter bound on the excitation amplitude: $\lambda \lesssim 0.05$ for a feature centered near the first acoustic peak ($k_* \approx 1.45 \times 10^{-2} \text{ Mpc}^{-1}$). No statistically significant detection is found; the data is consistent with the BD vacuum within cosmic variance.
• Bispectrum Predictions: The SIS model predicts a dominant non-Gaussian signal in the folded triangle configuration, with an amplitude $|f_{NL}^{fold}| \sim \lambda/(2\epsilon) \approx 2.5$ for fiducial parameters. However, current Planck constraints are too weak to detect this localized signal due to dilution over the full triangle space.
• Galaxy Clustering Superiority: The Fisher matrix analysis for the DESI survey indicates that galaxy clustering is a qualitatively superior probe. Unlike the CMB, the 3D power spectrum preserves the full oscillatory structure. The preliminary forecast suggests a sensitivity of $\sigma(\lambda) \approx 0.016$ (before accounting for survey window functions), offering a potential to probe below the current Planck bound.
• Structural Constraints: The paper identifies five specific constraints that distinguish SIS from generic excited-state models:
  1. Local oscillation frequency exceeding the global bandwidth.
  2. A specific ratio between the widths of the smooth and oscillatory components.
  3. A minimum order $N$ for the superoscillatory function ($N \geq a^2$).
  4. A binomial harmonic spectrum for the Fourier coefficients.
  5. Exponential suppression of bin-averaged signals ($e^{-a^2/2}$).

Significance and Claims
The paper claims to provide the first explicit boundary-action derivation linking superoscillatory quantum interference to inflationary initial-state modifications. Its primary significance lies in transforming the concept of superoscillation from a mathematical curiosity into a testable, tightly constrained phenomenological framework.

• Theoretical Economy: The framework demonstrates that a single quadratic boundary term can generate a rich, structured spectrum (oscillations, smooth bumps, specific consistency relations) without ad-hoc parametrization.
• Overdetermined Consistency: The authors argue that any future detection of a localized oscillatory feature would face a specific set of "overdetermined" consistency checks (e.g., the relationship between $a$, $N$, and the width ratio). This allows for the potential falsification of a superoscillatory origin if these structural constraints are violated.
• Observational Strategy: The work highlights a shift in observational strategy, arguing that galaxy clustering (specifically with DESI) offers a unique channel free from the transfer-function smearing that limits CMB sensitivity to superoscillatory features.
• Modesty: The authors explicitly state that their implementation is phenomenological; they do not claim to derive the boundary kernel from a specific UV completion (e.g., string theory or brane models) nor do they claim to have derived the preferred scale $k_*$ from first principles. The central scale $k_*$ is treated as a scan parameter. They acknowledge that whether DESI can robustly probe below the Planck bound depends on future analyses incorporating full survey window functions and systematic effects.

In summary, the paper establishes a concrete bridge between the foundational quantum mechanics of superoscillations (Aharonov, Berry) and precision cosmology, offering a falsifiable hypothesis for the initial state of the universe that is currently constrained by Planck but potentially testable by next-generation galaxy surveys.