🔭 astrophysics

Cosmic CORALS: Timing the Universe with high-z star clusters

This paper demonstrates that measuring the ages of high-redshift star clusters observed by JWST provides a cosmology-independent method to constrain the Hubble constant and matter density, with forecasts indicating that a future sample of ~300 such clusters could achieve competitive 4% precision on $H_0$ .

Original authors: Elena Tomasetti, Martin Millon, Licia Verde, Frédéric Courbin, Raul Jimenez, Michele Moresco, Carmela Lardo, Andrea Cimatti

Published 2026-07-23

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Elena Tomasetti, Martin Millon, Licia Verde, Frédéric Courbin, Raul Jimenez, Michele Moresco, Carmela Lardo, Andrea Cimatti

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, ticking clock that started running about 13.8 billion years ago. For a long time, astronomers have been trying to figure out exactly how fast that clock is ticking and how much "stuff" (like dark matter and dark energy) is inside it to keep the gears turning. This is the field of cosmology, and the two biggest knobs on the machine are the Hubble constant ( $H_0$ ), which tells us how fast the universe is expanding today, and the matter density ( $\Omega_m$ ), which tells us how much heavy stuff is in the mix. The problem is, when astronomers measure these knobs using different methods, they get slightly different answers, leading to a confusing "tension" in the scientific community. To solve this, we need a new way to check the time that doesn't rely on the old, potentially biased methods. We need a cosmic stopwatch that started ticking right at the beginning of time.

This is where star clusters come in. Think of a globular cluster as a giant, ancient family reunion where all the stars were born at roughly the same time. Because we know how fast stars age, measuring how old these clusters are gives us a direct "time stamp" of when the universe was young. If we can find these ancient families in the very distant past (high redshift), we can see how the universe's age relates to its expansion at that specific moment. It's like checking the time on a watch that was made yesterday versus one made a thousand years ago to see if the watch has been running fast or slow over history.

In this paper, the authors launch a new project called CORALS (Clusters of stars as Observational Records of cosmic Aging at Lookback timeS) to do exactly this. They focus on a spectacular cosmic event called the "Cosmic Gems," a distant galaxy arc magnified by a massive gravitational lens (a natural telescope made of gravity) located at a redshift of $z = 9.625$ . This is incredibly far away, meaning we are seeing it as it existed when the universe was only a tiny fraction of its current age. Using the powerful James Webb Space Telescope (JWST), the team peered into this arc and used a special image-processing technique called "deconvolution" to separate the blurry light of the background galaxy from the sharp points of light belonging to individual star clusters. They successfully identified 20 distinct point sources (which correspond to 10 unique lensed clusters) hidden within the arc.

The team then acted like cosmic detectives, analyzing the light from these clusters to determine their ages. Crucially, they did this without assuming any specific model of the universe first, ensuring their measurements were unbiased. They tested three different theories about how these stars formed (instantly, slowly fading away, or delayed) and found that, regardless of the theory, the results were consistent: the oldest half of these clusters formed about $102 \pm 63$ million years after the Big Bang.

With these "time stamps" in hand, the authors combined their new high-redshift data with existing, very precise age measurements of local globular clusters (the "old" clocks nearby). They then performed a joint statistical fit to see how the age of the universe changes with distance. The result is a new, independent way to measure the universe's expansion knobs. Currently, with this small sample, the constraints are still a bit loose, yielding a Hubble constant of $H_0 = 70^{+27}_{-16}$ km s $^{-1}$ Mpc $^{-1}$ and a matter density of $\Omega_m = 0.33^{+0.37}_{-0.21}$ . While these numbers have large error bars, the study proves a vital concept: by looking at the universe at different ages, the relationship between $H_0$ and $\Omega_m$ changes, effectively "rotating" the uncertainty in a way that helps break the deadlock between different measurement methods.

The paper doesn't stop at what they found today; it looks forward to what could be possible. The authors ran simulations to forecast what would happen if we could find a much larger sample of these ancient clusters—about 300 lensed proto-globular clusters spread out from the local universe up to $z \approx 10$ . Their simulations suggest that with such a sample, we could tighten the precision on the Hubble constant to 4% and on the matter density to 11%. This would make the "Cosmic CORALS" method competitive with, and completely independent from, the most advanced methods currently in use. The authors emphasize that this isn't just a dream; with the James Webb Space Telescope, the upcoming Euclid mission, the Nancy Grace Roman Space Telescope, and the future Extremely Large Telescope, we are on the verge of routinely discovering these cosmic timekeepers, turning this simulation into a real, powerful tool for understanding our universe.

Technical Summary: Cosmic CORALS: Timing the Universe with high-z star clusters

Problem Statement
The current tension in the measurement of the Hubble constant ( $H_0$ ) introduces a discrepancy of over 1 Gyr in the absolute age of the Universe at $z=0$ for a fixed matter density parameter ( $\Omega_m$ ). While local globular clusters (GCs) provide robust lower limits on the age of the Universe, they primarily constrain the $H_0 - \Omega_m$ degeneracy along a specific direction in parameter space, requiring external priors to break the degeneracy. The authors propose that anchoring the age-redshift relation ( $t(z)$ ) at high redshifts using the oldest star clusters could provide an independent, self-consistent cosmological probe. The James Webb Space Telescope (JWST), combined with gravitational lensing, now enables the detection and resolution of these high-redshift progenitors of local GCs, offering a potential avenue to measure cosmic time at epochs where the Universe was matter-dominated, thereby altering the nature of the cosmological parameter degeneracy.

Methodology
The paper introduces the CORALS (Clusters of stars as Observational Records of cosmic Aging at Lookback timeS) project, utilizing the "Cosmic Gems" arc at $z = 9.625$ as a primary case study.

Data Reduction and Photometry:
- The authors utilized multi-band JWST/NIRCam imaging (8 bands from F090W to F444W).
- They applied the STARRED deconvolution pipeline to separate compact point sources from the extended background emission and lensed caustics.
- This process identified 20 distinct point sources (10 unique lensed clusters, each split into two counter-images) along the arc, including a newly identified pair (Source F) and fainter tail sources (G–J).
- Sources with unreliable detections in multiple bands were discarded, leaving 10 robust sources for analysis.
Age Determination (Cosmology-Independent):
- Stellar ages were derived using the BAGPIPES spectral energy distribution (SED) fitting code with Bruzual & Charlot (2003) SPS models and a Kroupa IMF.
- Crucially, the fitting process explicitly removed cosmological priors on the age of the Universe to ensure the resulting ages are unbiased cosmological anchors.
- Three Star Formation History (SFH) scenarios were tested: single burst, exponentially declining, and delayed exponentially declining. The duration of star formation was limited to 100 Myr.
- The authors adopted the standard stellar age ( $t_\star$ , time since onset of star formation) rather than the mass-weighted age for cosmological analysis. The exponentially declining SFH was selected as the fiducial model as a robust physical middle ground.
Cosmological Fitting:
- The derived ages for the Cosmic Gems ( $z \approx 9.6$ ) were combined with state-of-the-art local GC ages ( $z=0$ ) from Valcin et al. (2026).
- A joint Bayesian fit was performed on the age-redshift relation within a flat $\Lambda$ CDM model using MCMC (emcee).
- The formation redshift ( $z_f$ ) for both populations was treated as a free parameter rather than a fixed prior, allowing the data to drive the constraints.

Key Results

Stellar Ages: The analysis yielded a combined age for the oldest half of the Cosmic Gems sample of $t_\star = 102 \pm 63$ Myr. The results were consistent across all three tested SFH models within $1\sigma$ , confirming the robustness of the age determination despite photometric degeneracies.
Cosmological Constraints (Current Data): Combining the high- $z$ $z$ Gems with local GCs resulted in the following constraints:
- $H_0 = 70^{+27}_{-16}$ km s $^{-1}$ Mpc $^{-1}$
- $\Omega_m = 0.33^{+0.37}_{-0.21}$
- While the statistical precision is currently loose, the joint fit successfully breaks the $H_0 - \Omega_m$ degeneracy. The authors demonstrate that high- $z$ measurements penalize low $\Omega_m$ values, while local measurements penalize high $\Omega_m$ , effectively rotating the degeneracy ellipse.
Theoretical Degeneracy Rotation: The paper analytically derives that the $H_0 - \Omega_m$ degeneracy direction shifts from $\Omega_m^{0.28} H_0 \approx \text{const}$ at $z \sim 0$ to $\Omega_m^{0.5} H_0 \approx \text{const}$ at $z \gg 1$ . This geometric rotation is the mechanism allowing high- $z$ clocks to constrain parameters independently.
Forecasts: Simulations of a future sample of $\sim 300$ $\sim 300$ lensed proto-GCs distributed up to $z \approx 10$ $z \approx 10$ (accounting for realistic scatter in formation redshifts and observational errors) predict:
- Statistical precision of 4% on $H_0$ .
- Statistical precision of 11% on $\Omega_m$ .
- These forecasts suggest the method could become competitive with multi-probe frameworks (e.g., BAO + SNIa) without requiring external calibrations.

Significance and Claims
The paper claims to represent a "new avenue in cosmology" by establishing high-redshift star clusters as independent cosmic clocks. Its primary significance lies in:

Independence: Providing a method to constrain cosmological parameters that is independent of the cosmic distance ladder and the Cosmic Microwave Background (CMB), thereby offering a direct cross-check for the $H_0$ tension.
Geometric Leverage: Demonstrating that the rotation of the $H_0 - \Omega_m$ degeneracy at high redshift allows for self-consistent constraints when combining local and high- $z$ data.
Timeliness: The work is positioned at a critical juncture where JWST routinely observes high- $z$ lensed clusters, and upcoming facilities (Euclid, Roman, ELT) will drastically increase the available sample size, making the transition from pilot studies to statistically significant cosmological probes imminent.

The authors remain modest regarding current constraints, acknowledging that the present sample size yields loose limits and that systematic effects (e.g., stellar models, binary evolution) must be rigorously quantified as larger samples become available. However, they assert that the methodology is robust and the potential for future precision is high.

More like this