🔭 astrophysics

Bayesian Constraints on Inverse-Tangent Inflation with Constant-EOS Reheating and a Dynamical Reheating Analysis

This paper presents a Bayesian analysis of inverse-tangent inflation coupled with constant and dynamical equation-of-state reheating, demonstrating that incorporating reheating dynamics not only constrains the model's parameters and predicts high reheating temperatures but also leverages the intrinsic $n_s-H_0$ degeneracy to bridge early-universe inflation with late-time cosmological parameter inference.

Original authors: Mayur Abhisheki, Prasanta Kumar Das

Published 2026-06-11

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Mayur Abhisheki, Prasanta Kumar Das

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe as a giant, expanding balloon. A long time ago, this balloon didn't just grow; it blew up incredibly fast in a split second. This event is called Inflation. It solved many puzzles about why the universe looks the way it does today.

But here's the catch: inflation can't just stop and leave the universe cold and empty. To get the hot, dense soup of particles that eventually formed stars and galaxies, the universe had to "reheat." Think of inflation as a car speeding down a hill, and reheating as the driver slamming on the brakes and the engine revving up to generate heat.

This paper is like a detective story where the authors try to figure out exactly how that "braking and heating" happened, using clues left behind in the oldest light in the universe (the Cosmic Microwave Background, or CMB).

Here is a simple breakdown of their investigation:

1. The Mystery of the "Inverse-Tangent" Model

The authors are testing a specific theory about the shape of the "hill" the universe rolled down during inflation. They call it an Inverse-Tangent Potential.

The Analogy: Imagine a slide. Some slides are flat at the top and steep at the bottom. Others are steep at the top and flat at the bottom. This specific model is a hybrid: it's flat when you are high up (large fields) and curves like a parabola when you get near the bottom (the origin).
The Goal: They want to know: How steep was this slide? How long did the slide last?

2. The "Reheating" Connection

The big problem in cosmology is that we can't directly see the reheating phase. It's like trying to guess how long a car's engine was revving after it stopped, just by looking at the smoke coming out of the exhaust today.

The Clue: The "smoke" is the Spectral Index ( $n_s$ ). This is a number that tells us how smooth or bumpy the universe's density was.
The Link: The authors realized that the length of the slide (how long inflation lasted) and how fast the engine revved (reheating) are mathematically tied to that "bumpiness" number. If you know the bumpiness, you can work backward to guess how the reheating happened.

3. Two Ways to Look at the Engine

The team tested two different ways the "engine" (reheating) could have worked:

A. The Constant Engine (Constant EOS)

The Idea: They assumed the engine revved at a steady, unchanging rhythm.
The Result: Using data from two major telescopes (Planck and ACT), they found the universe likely reheated for a short time (3 to 36 "e-folds," which are units of expansion) and got incredibly hot (trillions of degrees).
The Twist: They found that if you assume this specific reheating history, it changes how we calculate the Hubble Constant ( $H_0$ )—the rate at which the universe is expanding today.
- The Analogy: It's like realizing that if you know exactly how hard the brakes were hit, you can recalculate the car's speed at the start of the trip. This helped the Planck telescope's data move closer to the ACT telescope's data, potentially solving a disagreement between the two about how fast the universe is expanding.

B. The Variable Engine (Dynamical EOS)

The Idea: In reality, engines don't rev at a constant speed. As the inflaton field (the "fuel") oscillates, it decays into particles at a changing rate.
The Result: They introduced a "Yukawa coupling" ( $y$ $y$ ), which is basically a dial controlling how strongly the fuel interacts with the particles it creates.
- If the dial is set high (strong interaction): The engine revs up fast, reheating happens quickly (very short duration), and the temperature is massive.
- If the dial is set low (weak interaction): The engine sputters. Reheating takes a long time, and the temperature is much lower.
The Surprise: They found that the specific shape of the inflation slide (the parameter $\kappa$ ) didn't matter much for the reheating. What mattered almost entirely was that "Yukawa dial."

4. The "Consistency Check" (The Final Filter)

The most exciting part of the paper is the "Inflation-Reheating Consistency."

The Analogy: Imagine you have a long, continuous rainbow of possible answers for how the universe started. However, when you add the rules of how the engine must have worked to match the physics of the particles, most of that rainbow disappears.
The Result: Only very specific, narrow spots on that rainbow remain valid.
- The "bumpiness" ( $n_s$ ) must be between 0.9720 and 0.9725.
- The "ripples" in the fabric of space ( $r$ ) must be between 0.026 and 0.060.
Why it matters: This proves that you can't just pick any inflation model. The model has to be perfectly compatible with the reheating physics. If the numbers don't match up, the theory is wrong.

Summary

The authors used a sophisticated statistical method (Bayesian inference) to say: "If we assume this specific shape for the inflation slide, and we assume the universe reheated in this specific way, here are the only numbers that make sense."

They found that:

The universe likely reheated at temperatures between 10 trillion and 100 quadrillion degrees.
The "reheating weight" helps align different telescope measurements of the universe's expansion speed.
The details of how the inflaton decays (the Yukawa coupling) are far more important for reheating than the exact shape of the inflation slide.
By forcing the inflation and reheating phases to agree with each other, they narrowed down the possible properties of our universe to a very small, precise set of numbers.

In short, they built a bridge between the very first split-second of the universe and the data we see today, showing that the "engine" of the early universe is tightly constrained by the laws of physics.

Technical Summary: Bayesian Constraints on Inverse-Tangent Inflation with Constant-EOS Reheating and a Dynamical Reheating Analysis

Problem Statement
Precision measurements of the Cosmic Microwave Background (CMB) by the Planck and ACT collaborations have placed stringent constraints on inflationary observables, specifically the scalar spectral index ( $n_s$ ) and the tensor-to-scalar ratio ( $r$ ). However, a critical degeneracy exists between inflationary parameters and the post-inflationary reheating phase. The number of e-folds between horizon exit and the end of inflation ( $N_k$ ), which determines theoretical predictions for $n_s$ and $r$ , depends explicitly on the reheating dynamics (duration and equation of state).

Standard analyses often treat reheating heuristically by assuming a constant equation of state ( $\omega_{RH}$ ), which may fail to capture the continuous evolution of the inflaton condensate and radiation. Furthermore, CMB data constrains $n_s$ but not reheating parameters directly. This paper addresses the need to embed a specific inflationary model—the inverse-tangent potential—into a fully Bayesian framework that connects inflationary dynamics, reheating physics (both constant and dynamical), and CMB parameter inference without assuming ad-hoc reheating scenarios.

Methodology
The authors perform a Bayesian inference analysis on an inflationary model based on the inverse-tangent potential:
$V(\phi) = V_0 \arctan^2(\kappa \phi / M_{pl})$
This potential interpolates between plateau-like behavior at large field values and quadratic behavior near the origin.

The analysis proceeds through three main methodological stages:

Bayesian Inference with Constant EOS: The authors construct an interpolated mapping from inflationary parameters $(\kappa, N_k)$ to observables $(n_s, r)$ using precomputed slow-roll solutions. They perform Bayesian inference using Planck and ACT constraints on $n_s$ and the upper bound on $r$ . Reheating parameters ( $T_{RH}, N_{RH}$ ) are not sampled independently but are derived via consistency relations assuming a constant reheating equation of state $\omega_{RH} \in \{-1/3, 0, 2/3, 1\}$ .
Hubble Parameter ( $H_0$ ) Correlation: The authors map the inflationary-reheating posterior samples onto publicly available Planck and ACT MCMC chains for $\Lambda$ CDM parameters. They calculate "reheating-weighted" $H_0$ posteriors by integrating the conditional distribution $P(H_0|n_s, D)$ with the model-induced distribution $P(n_s|M)$ . This investigates how early-universe reheating constraints propagate to late-time cosmological parameters via the intrinsic $n_s - H_0$ degeneracy in CMB likelihoods.
Dynamical Equation of State (DEOS) Analysis: Moving beyond the constant EOS approximation, the authors numerically solve coupled evolution equations for the inflaton and radiation energy densities. This includes a time-dependent decay rate $\Gamma(t)$ derived from perturbative reheating into fermion pairs, where $\Gamma$ depends on the inflaton mass and Yukawa coupling $y$ . They analyze the evolution of the effective equation of state $\omega_{RH}(t)$ , reheating duration, and temperature for different values of the potential parameter $n$ and Yukawa coupling $y$ .
Consistency Check: Finally, the authors impose a strict consistency condition where the inflationary e-folding number $N_k$ derived from slow-roll dynamics must match the $N_k$ derived from the reheating evolution for a given set of parameters.

Key Results

Inflationary Parameters: For a range of constant $\omega_{RH}$ $ω_{R H}$ , both Planck and ACT data favor $\kappa \simeq 0.5 - 0.6$ $κ ≃ 0.5 - 0.6$ and $N_k \simeq 40 - 60$ $N_{k} ≃ 40 - 60$ .
- Planck prefers $N_k \approx 45-47$ for low $\omega_{RH}$ , while ACT systematically favors higher values ( $N_k \approx 52-53$ ), reflecting differences in spectral index constraints.
- As $\omega_{RH}$ increases to $2/3$ and $1$, the preferred $N_k$ shifts to $\approx 58$ , with uncertainties shrinking to the percent level due to reduced degeneracy.
Reheating Characteristics (Constant EOS):
- Low $\omega_{RH}$ ( $-1/3, 0$ ) corresponds to longer reheating durations ( $N_{RH} \sim 20-36$ e-folds) and lower temperatures ( $T_{RH} \sim 10^{10} - 10^{14}$ GeV).
- High $\omega_{RH}$ ( $2/3, 1$ ) leads to shorter durations ( $N_{RH} \sim 3-6$ e-folds) and higher temperatures ( $T_{RH} \sim 10^{12} - 10^{14}$ GeV).
Impact on $H_0$ : Reheating-weighted posteriors shift the Planck-inferred $H_0$ toward higher values, moving it closer to the ACT preferred region. This shift arises purely from the $n_s - H_0$ degeneracy in the CMB likelihood as reheating constraints narrow the allowed range of $n_s$ . No modifications to late-time cosmology are required; the shift is a consequence of early-universe physics filtering the CMB posterior.
Dynamical Reheating (DEOS):
- For a constant decay rate, reheating is a gradual process. For $n=1$ , equality occurs after $\sim 7-8$ e-folds ( $T_{RH} \sim 10^{12}-10^{13}$ GeV); for $n=3$ , it occurs faster ( $\sim 4-5$ e-folds) with higher $T_{RH}$ .
- For a time-dependent decay rate, the reheating history is largely insensitive to the inflationary parameter $\kappa$ but highly sensitive to the Yukawa coupling $y$ .
- Varying $y$ from $1$ to $10^{-12}$ changes $N_{RH}$ from $\mathcal{O}(1)$ to $\mathcal{O}(30)$ e-folds and $T_{RH}$ from $\sim 10^{14}$ GeV down to $\sim 10^{-2}$ GeV. The lowest couplings ( $y \sim 10^{-12}$ ) yield temperatures near the BBN limit ( $\sim 4.6 - 11.5$ MeV).
Inflation-Reheating Consistency: Imposing the condition that inflationary and reheating $N_k$ must match restricts the viable parameter space significantly. Instead of a continuous band in the $(n_s, r)$ plane, only isolated points satisfy consistency. For $y \in [10^{-12}, 10^{-3}]$ , the consistent solutions cluster in a narrow region: $n_s \simeq 0.9720 - 0.9725$ and $r \simeq 0.026 - 0.060$ . No consistent solutions were found for $y=1$ .

Significance and Claims
The paper claims to demonstrate that reheating dynamics provide a "nontrivial bridge" between early-universe inflation and late-time cosmological parameter inference. Specifically:

Parameter Constraints: Reheating physics, when treated consistently within a Bayesian framework, significantly restricts the viable parameter space of the inverse-tangent inflation model, reducing it to a narrow region in the $(n_s, r)$ plane.
$H_0$ Tension: The analysis suggests that early-universe reheating constraints can influence the inference of the Hubble constant ( $H_0$ ) by shifting the Planck posterior toward the ACT preferred region. This occurs through the geometry of the CMB likelihood ( $n_s - H_0$ degeneracy) without invoking new late-time degrees of freedom.
DEOS Realism: The dynamical equation of state framework reveals that reheating is an emergent phenomenon governed by microphysical decay rates (specifically the Yukawa coupling) rather than just the shape of the inflationary potential. This offers a more self-consistent description of perturbative reheating compared to constant EOS approximations.
Model Independence: The shift in $H_0$ is presented as a model-independent consequence of propagating early-universe constraints through existing CMB data structures, offering a potential mechanism to partially alleviate tensions between CMB datasets.