⚛️ phenomenology

Probing Cosmic Curvature with Fast Radio Bursts and DESI DR2

This study utilizes a sample of 120 localized Fast Radio Bursts combined with DESI DR2 Baryon Acoustic Oscillation data and artificial neural networks to constrain the cosmic curvature parameter $\Omega_k$ in a model-independent manner, finding results consistent with a spatially flat Universe while demonstrating the growing potential of FRBs as a precision cosmological probe.

Original authors: Jéferson A. S. Fortunato, Wiliam S. Hipólito-Ricaldi, Gustavo E. Romero

Published 2026-01-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jéferson A. S. Fortunato, Wiliam S. Hipólito-Ricaldi, Gustavo E. Romero

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. For decades, cosmologists have been trying to figure out the exact shape of that balloon. Is it perfectly flat like a sheet of paper? Is it curved like a sphere (closed)? Or is it curved like a saddle (open)? This shape is defined by a number called cosmic curvature ( $\Omega_k$ ).

This paper is like a team of detectives using a brand-new type of "cosmic flashlight" to solve this mystery without relying on old, potentially biased maps.

Here is the story of how they did it, broken down into simple parts:

1. The New Flashlight: Fast Radio Bursts (FRBs)

For a long time, astronomers used things like exploding stars (supernovae) to measure distances. But this team decided to use Fast Radio Bursts (FRBs).

What are they? Think of FRBs as incredibly bright, millisecond-long flashes of radio waves coming from deep space.
The "Sugar" Analogy: As these radio flashes travel through the universe, they pass through a fog of invisible electrons. This fog acts like sugar dissolving in coffee: the more sugar (electrons) the light passes through, the more the "flavor" (the signal) gets spread out or "dispersed."
The Clue: By measuring exactly how much the signal is spread out (called the Dispersion Measure), the team can calculate how much "fog" the light traveled through. Since the fog is spread out across the universe, the amount of fog tells them how far away the flash came from.

2. The Problem: The "Model" Trap

Usually, to turn these measurements into a map of the universe, scientists have to assume a specific story about how the universe works (a "cosmological model"). It's like trying to measure a room's size while assuming you already know the exact size of your ruler. If your assumption about the ruler is wrong, your measurement of the room is wrong.

The authors wanted to avoid this trap. They wanted to measure the shape of the universe without assuming a specific story about how it expands.

3. The Solution: Two Different Paths to the Same Destination

To solve this without a pre-set ruler, they used two different methods to calculate the distance to these radio flashes and compared the results.

Path A: The FRB-Only Map (The "Direct" Route)
They used a super-smart computer program (an Artificial Neural Network) to learn the relationship between the radio flashes and their distance, purely based on the data they collected. This program acted like a translator, turning the "sugar spread" (dispersion) directly into a distance map. This method does depend on the shape of the universe (curvature), so it gave them a distance estimate that changes depending on whether the universe is flat, open, or closed.
Path B: The FRB + BAO Map (The "Cross-Check" Route)
They combined their FRB data with data from BAO (Baryon Acoustic Oscillations). Think of BAO as "fossil ripples" left over from the Big Bang that act as a standard-sized ruler across the universe. By mixing the FRB data with these fossil rulers, they created a second distance estimate. Crucially, this second method is mathematically designed to be independent of the universe's shape.

4. The Detective Work: Comparing the Maps

Now, they had two maps:

One that changes based on the shape of the universe.
One that doesn't care about the shape.

They compared the two. If the universe were perfectly flat, the two maps would match up perfectly. If the universe were curved, the maps would drift apart. By tweaking the "curvature number" ( $\Omega_k$ ) until the two maps lined up, they could find the true shape of the universe.

5. The Results: A Flat Universe (Mostly)

After crunching the numbers on 120 of these radio flashes and combining them with the latest BAO data (from the DESI survey), they found:

The Verdict: The universe appears to be flat (like a sheet of paper).
The Numbers: Their best guess for the curvature is very close to zero.
- When they carefully accounted for all the messy connections between their data points (using a "full covariance" method), they got a result of -0.31 ± 0.57.
- When they used a simpler method, they got -0.13 ± 0.46.
The "Mild" Hint: While both results are consistent with a perfectly flat universe (zero), there is a tiny, "mild" hint that the universe might be slightly curved inward (negative curvature), like a sphere. However, the "error bars" are still wide enough that we can't say for sure yet.

Why This Matters

The authors emphasize that this is a model-independent discovery. They didn't have to assume the universe follows a specific set of rules to get this result. They just let the data speak.

They also found that being very careful about how they handled the data's uncertainties (the "covariance" method) made their error bars wider, which is actually more honest. It prevents them from being too confident in a result that might be shaky.

In short: By using fast radio flashes as cosmic flashlights and comparing them against ancient fossil rulers, this team confirmed that our universe is likely flat, while proving that this new method is a powerful, independent tool for mapping the cosmos.

Technical Summary: Probing Cosmic Curvature with Fast Radio Bursts and DESI DR2

Problem Statement
Determining the spatial geometry of the Universe, parameterized by the curvature density parameter $\Omega_k$ , remains a fundamental challenge in cosmology. Current constraints are largely consistent with a flat universe ( $\Omega_k = 0$ ), yet significant tensions exist between different analyses. Some studies utilizing Cosmic Microwave Background (CMB) data alone or combined with Baryon Acoustic Oscillations (BAO) suggest a preference for a closed universe ( $\Omega_k < 0$ ), while others favor flatness or open geometries. A critical limitation of existing constraints is their inherent dependence on specific cosmological models (typically non-flat $\Lambda$ CDM) and assumptions regarding the early universe (e.g., the sound horizon $r_d$ ). Furthermore, the intrinsic degeneracy between cosmic curvature and the Hubble constant ( $H_0$ ) complicates these measurements. There is a pressing need for model-independent probes that can constrain $\Omega_k$ without relying on early-universe calibration or specific expansion history priors.

Methodology
This work proposes a novel, cosmological model-independent framework to constrain $\Omega_k$ by leveraging the synergy between localized Fast Radio Bursts (FRBs) and BAO measurements from the Dark Energy Spectroscopic Instrument (DESI) Data Release 2 (DR2). The methodology proceeds through the following steps:

Data Selection: The authors utilize a sample of 120 localized FRBs with known redshifts and dispersion measures (DM). Strict selection criteria are applied to ensure physical consistency, specifically requiring that the intergalactic medium contribution ( $DM_{IGM}$ ) remains positive after subtracting Milky Way, halo, and host galaxy contributions. Seven events are excluded due to proximity (negligible IGM contribution) or anomalous dispersion measures.
Reconstruction of $H(z)$ and $D_C(z)$ :
- The observed dispersion measure is decomposed to isolate the IGM contribution, $DM_{IGM}$ .
- An Artificial Neural Network (ANN), specifically a Multilayer Perceptron (MLP), is trained to reconstruct the relationship between $DM_{IGM}$ and redshift $z$ in a data-driven manner.
- Using the Macquart relation, the derivative of the mean $DM_{IGM}$ with respect to redshift is used to reconstruct the Hubble parameter $H(z)$ and, by integration, the comoving distance $D_C(z)$ .
- The Hubble constant $H_0$ is derived internally by extrapolating the reconstructed $H(z)$ to $z=0$ , avoiding external priors.
Dual Estimation of Angular Diameter Distance ( $D_A$ ):
- FRB-only approach ( $D_A^{FRB}$ ): The comoving distance $D_C(z)$ is mapped to the angular diameter distance $D_A(z)$ using the FLRW metric relations, which explicitly depend on $\Omega_k$ .
- FRB+BAO approach ( $D_A^{BAO+FRB}$ ): The reconstructed $H(z)$ from FRBs is combined with dimensionless BAO observables ( $D_M/r_d$ and $D_H/r_d$ ) from DESI DR2. In this formulation, the sound horizon $r_d$ cancels out algebraically, yielding an estimate of $D_A(z)$ that is independent of $\Omega_k$ and early-universe physics.
Likelihood Analysis: The curvature parameter $\Omega_k$ $Ω_{k}$ is constrained by comparing the curvature-dependent $D_A^{FRB}(z)$ $D_{A}^{F R B} (z)$ with the curvature-independent $D_A^{BAO+FRB}(z)$ $D_{A}^{B A O + F R B} (z)$ . Two statistical treatments are employed:
- Covariance-based Likelihood: A full covariance matrix is constructed to account for statistical uncertainties and, crucially, the correlated errors induced by the shared $H(z)$ reconstruction in both distance estimators.
- Gaussian $\chi^2$ Approach: A simplified treatment assuming independent, Gaussian-distributed data points, neglecting off-diagonal correlations.

Key Contributions

Model-Independent Curvature Probe: The paper demonstrates a method to constrain $\Omega_k$ without assuming a fiducial cosmological model or relying on CMB-based priors for the sound horizon.
FRB as a Primary Tracer: It establishes localized FRBs as a viable, independent tracer for reconstructing the Hubble parameter and cosmic distances, complementing traditional probes like supernovae and cosmic chronometers.
Covariance Handling: The study explicitly highlights the impact of accounting for correlated uncertainties between the FRB-derived distances. It shows that neglecting these correlations (Gaussian treatment) leads to tighter but potentially underestimated error bars, whereas the full covariance approach provides a more conservative and robust uncertainty assessment.
Integration with DESI DR2: The work utilizes the latest BAO data from DESI DR2, combining it with FRB data to test the large-scale geometry of the Universe.

Results
The analysis yields the following constraints on the spatial curvature parameter $\Omega_k$ :

Full Covariance Treatment: $\Omega_k = -0.31 \pm 0.57$ .
Gaussian (Diagonal) Treatment: $\Omega_k = -0.13 \pm 0.46$ .

Both estimations are consistent with a spatially flat universe ( $\Omega_k = 0$ ) at the $1\sigma$ confidence level. However, both methods exhibit a mild preference for negative curvature (a closed universe). The inclusion of the full covariance matrix broadens the confidence intervals, preventing the underestimation of uncertainties, but does not qualitatively alter the central trend or the conclusion of consistency with flatness. The reconstructed $H_0$ from the FRB data is $68.74 \pm 6.11$ km s $^{-1}$ Mpc $^{-1}$ .

Significance and Claims
The authors claim that this work reinforces the growing relevance of FRBs in precision cosmology. By providing a complementary, model-independent perspective on cosmic curvature, the study offers a powerful cross-check against CMB-based constraints. The results suggest that while current data supports a flat universe, the persistent mild preference for negative curvature warrants further investigation as data quality and quantity improve. The paper emphasizes that the synergy between FRBs and BAO measurements offers a robust pathway to understanding the large-scale geometry of the Universe without relying on specific early-universe assumptions. The authors conclude that future increases in the number of localized FRBs and improved modeling of host galaxy contributions will be decisive in tightening these constraints.