Inverse-Scattering Reconstruction of Inflation from Scalar and Tensor Primordial Spectra

Technical Summary: Inverse-Scattering Reconstruction of Inflation from Scalar and Tensor Primordial Spectra

Problem Statement
The paper addresses the challenge of reconstructing the underlying inflationary dynamics and effective potentials from observed primordial power spectra. While standard approaches (e.g., Bayesian methods, flow equations, generalized slow-roll) exist, they often rely on specific parametric assumptions or deconvolution techniques. The authors propose a non-parametric framework based on inverse-scattering theory to directly infer the effective potentials governing scalar ($z''/z$) and tensor ($a''/a$) perturbations. This is particularly motivated by the potential of future experiments (e.g., CMB spectral distortions) to probe small scales ($k \sim 10^3\text{--}10^4 \text{ Mpc}^{-1}$), where transient features or deviations from canonical slow-roll evolution may leave localized imprints.

Methodology
The core methodology involves recasting the Mukhanov–Sasaki equation, which governs cosmological perturbations, as a Schrödinger-like equation on the half-line by identifying the radial coordinate $r = -\tau$ (where $\tau$ is conformal time).

1. Mapping to Scattering Theory:

   • The early-time Bunch–Davies vacuum condition (sub-horizon limit, $r \to \infty$) is identified with the asymptotic behavior of a Jost solution (outgoing plane wave).
   • The late-time freeze-out amplitude (super-horizon limit, $r \to 0$) is encoded in the small-$r$ behavior of the same Jost solution.
   • The authors distinguish between the regular solution (associated with the decaying super-horizon mode) and the Jost solution (associated with the physical, growing freeze-out mode). They argue that the physical cosmological perturbation is naturally linked to the Jost solution.

2. The Jost Function and Spectra:

   • The primordial power spectra for scalar ($P_S$) and tensor ($P_T$) perturbations are expressed in terms of the corresponding Jost functions, $F^{(S)}_{\nu_s-1/2}(k)$ and $F^{(T)}_{\nu_t-1/2}(k)$.
   • In a featureless slow-roll background, the Jost function reduces to a pure phase. Deviations from slow-roll (sharp features, oscillatory modulations) induce non-trivial $k$-dependence in the Jost function, making it a sensitive diagnostic of inflationary dynamics.
   • The tensor-to-scalar ratio $r(k)$ is derived as a ratio of the squared moduli of these Jost amplitudes.

3. Reconstruction via Born Approximation:

   • To test the framework, the authors employ the Born approximation in the large-$k$ (weak-scattering) regime. This linearizes the Marchenko integral equation, allowing the effective potential $V_s(r)$ to be reconstructed directly from the scattering kernel.
   • Two benchmark models are simulated: a smooth quadratic potential ($\phi^2$) and a step potential (which induces transient slow-roll violations and localized features).
   • Numerical solutions of the Mukhanov–Sasaki equation generate the "true" spectra, which are then used to reconstruct the effective potentials $z''/z$ and $a''/a$ via the inverse-scattering formalism.

Key Contributions

• Formalism: The paper establishes a unified inverse-scattering framework for both scalar and tensor sectors, explicitly linking the observable freeze-out amplitudes to the Jost function rather than the regular solution.
• Diagnostic Variable: It introduces a diagnostic variable $X_k = v_k/u_k$ (the ratio of tensor to scalar mode functions) and derives its evolution equation, showing that its dynamics are sourced by the difference between the scalar and tensor effective potentials ($z''/z - a''/a$).
• Tensor-to-Scalar Ratio: The tensor-to-scalar ratio is reformulated as a ratio of Jost amplitudes, recovering the standard slow-roll consistency relation ($r \approx 16\epsilon$) in the appropriate limit.
• Numerical Validation: The study provides a numerical demonstration of the reconstruction process, comparing the Born-approximation results against exact background evolutions for both smooth and feature-rich potentials.

Results

• Smooth Potentials: For the quadratic ($\phi^2$) model, the Born approximation successfully reconstructs the dominant behavior of both $z''/z$ and $a''/a$, closely matching the simulated background evolution across the domain. Minor discrepancies in the infrared limit are attributed to the sensitivity of the linearized Marchenko inversion to low-$k$ truncation.
• Step Potentials: For the step potential, the reconstruction captures the global scaling and asymptotic behavior but exhibits localized oscillatory deviations in the scalar sector ($z''/z$) near the feature region. These discrepancies arise because the sharp step induces strong scattering and non-adiabatic effects, violating the weak-scattering assumption inherent in the Born approximation.
• Scalar vs. Tensor: The tensor reconstruction ($a''/a$) remains smoother and more accurate than the scalar reconstruction for the step potential. This is attributed to the fact that tensor modes probe only the background geometry, whereas scalar modes are directly sensitive to the inflaton dynamics, which are more strongly perturbed by the step feature.

Significance and Claims
The authors claim that inverse scattering provides a "physically transparent" method for connecting features in the primordial spectra to the underlying inflationary dynamics. The Jost function is highlighted as a sensitive diagnostic for departures from canonical slow-roll evolution.

The paper modestly asserts that while the Born approximation is limited by its weak-scattering assumption (particularly for sharp features in the scalar sector), the framework remains robust for smooth slow-roll backgrounds and retains sensitivity to features even when the approximation is only partially valid. The authors position this work as a foundation for future extensions that would solve the full Marchenko equation (beyond the Born limit) and incorporate finite observational windows. They suggest that this formalism could serve as a model-independent tool for interpreting data from future high-precision observations (e.g., PIXIE, LSS surveys) targeting the small-scale primordial spectrum.