Beyond21: A Global Framework for Cosmic Dawn and Reionization Within and Beyond the Standard Model

The paper introduces Beyond21, a fast, open-source Python package that provides a unified framework for modeling the Cosmic Dawn and Epoch of Reionization to predict global 21-cm signals and other observables, while enabling flexible exploration of both standard astrophysics and beyond-Standard-Model scenarios like millicharged dark matter.

The Gist

Imagine the early universe as a giant, dark stage before the curtain rises on the first act of the cosmic play. For hundreds of millions of years after the Big Bang, this stage was pitch black, filled only with a cold, expanding gas. Then, the "Cosmic Dawn" arrived: the moment the first stars and galaxies flickered to life, turning on the lights, heating the air, and changing the chemistry of the entire universe.

This paper introduces a new, super-fast digital tool called Beyond21. Think of it as a universal "Cosmic Weather Simulator" that helps scientists understand exactly how that first light changed the universe, and whether there are hidden forces (like new types of dark matter) playing behind the scenes.

Here is a breakdown of what the paper does, using simple analogies:

1. The Problem: Too Many Puzzles, Not Enough Time

Scientists have many different ways to look at this early universe:

• The 21-cm Signal: Imagine the universe is a giant radio. Neutral hydrogen gas (the most common stuff in space) hums a specific radio note. When the first stars turn on, they change the pitch of that hum.
• Galaxy Counts: We can count how many bright stars (galaxies) existed at different times using telescopes like Hubble and JWST.
• X-ray Background: We can measure the faint "glow" of high-energy X-rays left over from the first black holes and stars.

The problem is that these pieces of the puzzle are all connected. If you change how stars form, it changes the radio hum, the galaxy count, and the X-ray glow. Previous computer programs were slow or only looked at one piece at a time.

2. The Solution: Beyond21 (The "All-in-One" Simulator)

Beyond21 is a new computer program that acts like a master chef. Instead of cooking just the soup (the radio signal) or just the salad (the galaxies), it cooks the entire meal at once.

• It's Lightning Fast: Most cosmic simulations take days or weeks to run. Beyond21 runs in 0.1 seconds (about the time it takes to blink). This means scientists can run it thousands of times to test different theories instantly.
• It's Modular: Think of the code like a set of LEGO bricks. If a scientist wants to test a new theory about how stars form, they can just swap out one brick without rebuilding the whole castle. If they want to test a new type of physics, they can snap that in too.

3. What It Simulates: The Cosmic Chain Reaction

The simulator tracks a chain reaction, like a line of falling dominoes:

1. Stars are Born: It calculates how the first stars (Pop II and Pop III) form in clouds of gas.
2. Light is Emitted: These stars shoot out UV light, X-rays, and other radiation.
3. The Gas Reacts: This radiation hits the surrounding gas.
   • UV Light acts like a heater, warming up the gas.
   • X-rays act like a microwave, penetrating deep and heating the gas from the inside out.
   • Lyman-alpha Light acts like a tuner, changing the "spin" of the hydrogen atoms, which changes the radio signal we hear.
4. The Result: The simulator predicts what we should see today: the radio signal, the number of galaxies, and the X-ray background.

4. The "Secret Ingredient": Testing New Physics

The coolest part of the paper is how Beyond21 tests New Physics. The authors used it to test a theory called "Millicharged Dark Matter."

• The Analogy: Imagine the universe is a room full of people (normal matter) and invisible ghosts (dark matter). Usually, the ghosts don't interact with the people. But in this theory, the ghosts have a tiny, tiny electric charge (a "millicharge").
• The Effect: Because they have a charge, the ghosts can bump into the people and steal their heat. This makes the gas in the early universe colder than we expected.
• The Result: If the gas is colder, the "radio hum" (the 21-cm signal) should be much deeper and louder.
• The Test: The authors ran Beyond21 with this "millicharged ghost" theory. They found that it could explain a deeper signal, but only if the X-ray heating from stars was weak. If the stars were too hot (too many X-rays), they would override the cooling effect of the ghosts.

5. Why This Matters

Before this tool, trying to figure out if "ghosts" (new physics) exist was like trying to solve a Rubik's cube while wearing blindfolded gloves. You had to guess one piece, check it, then guess another, and it took forever.

Beyond21 takes the blindfold off and gives you a robot arm. It allows scientists to:

• Connect the dots: See how a change in star formation affects the radio signal and the X-ray background simultaneously.
• Rule out theories: Quickly see if a new physics idea breaks the rules of the universe we observe.
• Prepare for the future: As new telescopes (like the Square Kilometre Array) come online, this tool will be ready to instantly interpret the massive amount of data they will send back.

In a Nutshell

Beyond21 is a fast, flexible, and all-encompassing simulator that lets scientists play "What If?" with the early universe. It helps them understand how the first stars lit up the dark, and whether the invisible "dark matter" that makes up most of the universe is doing something sneaky we haven't noticed yet. It turns a complex, multi-year research project into a 0.1-second calculation, opening the door to discovering the hidden laws of our cosmos.

Technical Summary

Here is a detailed technical summary of the paper "Beyond21: A Global Framework for Cosmic Dawn and Reionization Within and Beyond the Standard Model" by Omer Zvi Katz.

1. Problem Statement

The Cosmic Dawn (CD) and the Epoch of Reionization (EoR) represent critical phases in the early universe where the first stars and galaxies formed, driving the heating, ionization, and chemical enrichment of the intergalactic medium (IGM). While these eras are probed by diverse observables—such as the global 21-cm signal, high-redshift galaxy UV luminosity functions (UVLFs), the cosmic X-ray background (CXB), and CMB optical depth—interpreting these data is challenging.

• Complexity: These observables depend on the coupled evolution of stellar populations, radiation backgrounds (UV, Ly-$\alpha$, X-ray), and the thermal/ionization state of the IGM.
• New Physics Sensitivity: These signals are also sensitive to Beyond the Standard Model (BSM) physics, such as dark matter annihilation, decay, or scattering with baryons.
• Limitations of Existing Tools: Current codes often focus on a single observable (e.g., only 21-cm) or rely on simplified semi-analytical approaches that lack the flexibility to simultaneously model multiple observables or easily incorporate complex BSM physics. Furthermore, many existing tools are computationally expensive, hindering high-dimensional parameter inference.

2. Methodology: The Beyond21 Framework

The authors introduce Beyond21, a fully open-source Python package designed to provide a self-consistent, modular, and computationally efficient framework for modeling global evolution from $z \sim 1800$ down to $z \sim 6$.

Core Architecture

• Modularity: The code is structured into distinct modules (CosmoHMF, IonSFR, Xrays, NoIonUV, EvolveIGM) allowing users to modify specific astrophysical prescriptions (e.g., star formation efficiency, escape fractions) or physics (e.g., BSM heat exchange) without rewriting the entire pipeline.
• Computational Efficiency: A full global evolution run executes in approximately 0.1 seconds on a single CPU core. This speed enables high-dimensional Markov Chain Monte Carlo (MCMC) inference and broad parameter space exploration.
• Initialization: The code can initialize at very high redshifts ($z \sim 1800$), capturing the evolution from before hydrogen recombination through the end of reionization, which is essential for BSM scenarios involving early dark matter interactions.

Physical Components

1. Star Formation (SFRD):

   • Tracks two stellar populations: Population III (Pop-III, metal-free) and Population II (Pop-II, metal-enriched).
   • Uses the Sheth-Tormen Halo Mass Function (HMF).
   • Implements a halo-mass filter ($f_{gal}$) that accounts for cooling thresholds (atomic vs. molecular), Lyman-Werner (LW) feedback, and baryon-dark matter streaming velocities.
   • Includes photo-heating feedback suppression for Pop-III stars in ionized bubbles.

2. Radiation Fields:

   • UV (Ionizing & Non-Ionizing): Calculates the production of ionizing photons (driving reionization) and non-ionizing UV (Ly-$\alpha$ and LW). The Ly-$\alpha$ flux is computed by summing contributions from all Lyman-series transitions, accounting for radiative cascades.
   • X-rays: Models High-Mass X-ray Binaries (HMXBs) as the dominant source. Uses a double power-law spectrum and calculates the global X-ray intensity, accounting for attenuation in host galaxies and the IGM.
   • Secondary Effects: Computes X-ray induced Ly-$\alpha$ emission and heating/ionization rates via secondary electrons.

3. IGM Evolution:

   • Solves coupled differential equations for the neutral hydrogen fraction ($x_{HI}$) and gas kinetic temperature ($T_k$).
   • Ionization: Distinguishes between UV-driven ionization (localized bubbles) and X-ray driven ionization (homogeneous background). Uses a global volume-averaged approach for the ionized fraction ($Q_{HII}$).
   • Thermal History: Includes adiabatic cooling, Compton scattering, X-ray heating, and Ly-$\alpha$/21-cm heating/cooling.
   • 21-cm Signal: Calculates the global brightness temperature ($T_{21}$) by solving for the spin temperature ($T_s$), which is coupled to $T_k$ via the Wouthuysen-Field effect and to the CMB temperature ($T_\gamma$).

3. Key Contributions

• Unified Pipeline: Beyond21 is the first framework to simultaneously predict the global 21-cm signal, UVLFs, the unresolved CXB, and the ionization history ($\tau_e$) within a single, self-consistent model.
• BSM Integration: The modular design allows for the straightforward implementation of new physics. The authors demonstrate this by integrating the Two-Component Dark Matter (2cDM) model, where a fraction of dark matter is millicharged and scatters with baryons.
• High-Speed Inference: The $\sim 0.1$ s runtime makes it uniquely suited for statistical inference studies, allowing for the exploration of complex parameter spaces that were previously computationally prohibitive.
• Validation: The code's default outputs (fiducial model) show excellent agreement with the widely used semi-numerical code 21cmFAST for 21-cm signals and ionization histories, while also matching HST and JWST UVLF data.

4. Results

• Fiducial Model: The authors calibrated a fiducial model using HST UVLF data, Chandra CXB upper limits, and Planck CMB optical depth constraints. The model successfully reproduces the observed UVLFs at $z \sim 6-10$ and the global 21-cm signal evolution.
• Extended Model (JWST): By introducing a Gaussian scatter in the halo-mass to UV-magnitude relation, the model was extended to fit JWST data at higher redshifts ($z > 10$), finding a best-fit scatter of $\sigma_{UV} \approx 1.7$ mag.
• BSM Case Study (Millicharged DM):
  • The authors implemented a 2cDM scenario where a small fraction of dark matter ($f_m \lesssim 3 \times 10^{-4}$) is millicharged.
  • Mechanism: Millicharged particles scatter with baryons, transferring heat. In the specific parameter space chosen, this interaction cools the baryonic gas relative to the standard $\Lambda$CDM scenario during the Cosmic Dawn.
  • Impact: The reduced baryon temperature ($T_k$) leads to a significantly deeper global 21-cm absorption trough (enhanced by a factor of $\sim 3$ compared to $\Lambda$CDM).
  • Degeneracy Breaking: The study highlights that the amplitude of the 21-cm signal in BSM scenarios is highly degenerate with astrophysical parameters (e.g., X-ray heating strength and UV intensity). Beyond21's ability to jointly constrain these parameters is crucial for distinguishing new physics from astrophysical uncertainties.

5. Significance

• Synergy of Probes: Beyond21 provides a critical tool for the next generation of cosmology, enabling the joint analysis of 21-cm experiments (like EDGES, SARAS, and future SKA/HERA), JWST galaxy surveys, and X-ray observatories (Chandra/Athena).
• New Physics Constraints: By offering a fast and flexible framework, Beyond21 allows researchers to rigorously test BSM theories against a broad set of observational constraints, helping to break degeneracies between astrophysical uncertainties and new physics signatures.
• Community Resource: As a fully open-source, modular Python package, it lowers the barrier to entry for researchers to model complex early-universe scenarios and facilitates the development of future extensions (e.g., stochastic star formation, dust attenuation).

In summary, Beyond21 represents a significant advancement in computational cosmology, bridging the gap between detailed astrophysical modeling and efficient statistical inference, thereby empowering the community to extract robust constraints on both the nature of the first galaxies and potential new physics from the dawn of the universe.