Hidden mass in early galaxies revealed by bottom-heavy initial mass functions

arXiv:2601.20864 · astro-ph.GA · Submitted 2026-08-24 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: Next we'll be talking about the paper "Hidden mass in early galaxies revealed by bottom-heavy initial mass functions".

Jocelyn: The paper was written by Chloe M. Cheng, Martje Slob, Mariska Kriek, Aliza G. Beverage, Pieter G. van Dokkum et al. from Leiden Observatory, Leiden University and Observatories of the Carnegie Institution for Science and Astronomy Department, Yale University and Department of Physics & Astronomy and Pennsylvania Institute of Technology and PITT PACC, University of Pittsburgh and Cosmic Dawn Center (DAWN) and Niels Bohr Institute, University of Copenhagen and Center for Astrophysics Harvard & Smithsonian and Max-Planck-Institut für Astronomie and Department of Astrophysics, University of Zurich and Department of Astronomy, Tsinghua University and Hong Kong Institute for Astronomy & Astrophysics.

Vera: Stay tuned as we take you through the paper and discuss its implications.

Paper discussion segment 1 — Vera and Jocelyn discuss title and authors of the paper 'Hidden mass in early galaxies revealed by bottom-heavy initial mass functions' and its implications. Explain in simple terms; do not repeat what earlier segments covered.: Vera: So, we’ve established that the title “Hidden mass in early galaxies revealed by bottom-heavy initial mass functions” suggests a massive revelation, but let’s look specifically at the core findings of this study and what they are telling us about these distant objects.

Jocelyn: What really struck me about the results isn't just that there is hidden mass; it’s *how* they calculated it. They didn't just plug in a number, they showed the intricate chain of corrections required, accounting for things like dust and metallicities across multiple spectral bands to isolate this specific low-mass signature.

Subrahmanyanyan: This technical rigor is vital because the summary doesn't treat this as an optional add-on correction; it presents it as a fundamental necessity for accurately calculating the total stellar content of these ancient galaxies.

Vera: So, to put that in simple terms for our audience: many of those galaxies we thought were moderately sized based on their visible light actually contain a much larger population of faint, low-luminosity stars that contribute significantly to the overall mass.

Jocelyn: It’s a huge confirmation for the JWST community, isn't it? The study validates the telescope’s unique ability to peer into these faint spectral lines that older surveys simply couldn't resolve consistently across a large sample.

Subrahmanyanyan: What I find most compelling in the results is how consistently this hidden mass appears. It moves beyond suggesting an anomaly for one specific object and establishes a general, robust trend for entire populations of early cosmic structures at redshift z-infinity.

Vera: And that consistency is key because it allows us to move from questioning whether this was a fluke observation to asserting that this represents a universal physical process in the the early universe.

Jocelyn: It feels like the paper provides us with a definitive census, giving us a solid baseline for what stellar populations should look like when they form under conditions of rapid cosmic assembly.

Subrahmanyanyan: This baseline is incredibly powerful because it gives future researchers something concrete to compare against theoretical predictions alone.

Vera: Now that we understand the *what* and the *how much* based on these findings, I’m curious about what this means for how we should approach modeling these systems moving forward.

Paper discussion segment 2 — Vera and Jocelyn discuss the paper's summary of the paper 'Hidden mass in early galaxies revealed by bottom-heavy initial mass functions' and its implications. Explain in simple terms; do not repeat what earlier segments covered.: Jocelyn: So, we have this amazing data set from the JWST–IMFERNO program, which is a major step forward, but let’s focus on the specific quantification of this hidden mass in the paper “Hidden mass in early galaxies revealed by bottom-heavy initial mass functions.”

Vera: The core summary is astounding because it quantifies exactly how much stellar material was missing from our previous estimates. For the oldest two galaxies in this sample, for example, they found that these bottom-heavy IMFs increase their total stellar masses by a factor of three to four.

Jocelyn: That scale of increase is staggering when we consider the faint stars involved—it’s like suddenly realizing that what we thought was a medium-sized building actually contains three or four times the weight in foundation and structure.

Subrahmanyanyan: This finding is critical because it directly amplifies the existing tension with older galaxy formation models, as these galaxies are much more massive than those models predicted at that early epoch.

Vera: It’s a huge shift in mass estimation, so if we summarize this for our listeners, it means that the galaxies we observed were far larger and denser than our previous calculations allowed for them to be.

Jocelyn: And because this is happening across multiple samples at z-zero point seven, it gives us a very robust indication that the entire population of early massive galaxies was undercounted by existing literature.

Subrahmanyanyan: What I find most interesting about this is how the data point to specific low-mass stars being particularly abundant, which strongly suggests that the physical processes governing star formation in those environments were unique and highly efficient at low masses.

Vera: It’s a powerful confirmation of a particular type of stellar birth, meaning we are seeing evidence that the initial conditions favored creating these faint, numerous stars over massive ones.

Jocelyn: It feels like the paper provides us with an unprecedented census, giving us a solid baseline for what stellar populations should look like when they form under conditions of rapid cosmic assembly.

Vera: This baseline is incredibly powerful because it gives future researchers something concrete to compare against theoretical predictions alone.

Jocelyn: Now that we understand the *what* and the *how much* based on this massive hidden mass, I’m curious about what this means for how we should approach modeling these systems moving forward.

Paper discussion segment 3 — Vera and Jocelyn discuss the improvements the paper suggests of the paper 'Hidden mass in early galaxies revealed by bottom-heavy initial mass functions' and its implications. Explain in simple terms; do not repeat what earlier segments covered.: Vera: We’ve established that a substantial hidden mass exists due to bottom-heavy star formation, as detailed in “Hidden mass in early galaxies revealed by bottom-heavy initial mass functions.” Now, we need to talk about the specific improvements and adjustments the authors suggest for future theory and research.

Jocelyn: The primary improvement suggested is a fundamental overhaul of the assumptions we currently make about stellar inputs. The authors are essentially telling theoretical modelers that their foundational physics needs major revisions if they want to match this observational reality.

Subrahmanyanyan: Specifically, they mandate that any future simulation of structure formation must incorporate a much heavier low-mass tail in its initial mass function parameters. It’s not enough just to tweak the exponent; the the entire mathematical framework for stellar birth needs adjustment to accommodate these higher numbers of small stars.

Vera: They also suggest looking into different star formation histories, and while their results were consistent with a single-age model, acknowledging that a more complex two-burst history is possible adds necessary nuance.

Jocelyn: That’s right; the authors are encouraging us to move beyond simplistic scenarios. They are suggesting that our models need to be able to account for different stages of star formation, not just one continuous event.

Subrahmanyanyan: And they also point out that the correlation between the IMF and properties like stellar velocity dispersion and iron abundance is consistent across both the early universe and local observations, implying that these trends are universal rather than unique to a single system.

Vera: It’s truly compelling that this consistency suggests a robust, universal physical process. The paper argues that these trends were already in place at redshift z-zero point seven, meaning our models need to reflect the same dynamics then and now.

Jocelyn: Furthermore, they suggest we have to seriously consider the limitations of our current methods—like how stellar population models are currently able to handle certain redder spectral features.

Subrahmanyanyan: The paper also recommends that future analyses must account for non-baryonic matter; since these galaxies have small effective radii, the virial and stellar masses should be compared directly to ensure consistency, which is a crucial check on our assumptions about dark matter.

Vera: It’s clear that by suggesting these improvements, they are asking us to move away from theoretical convenience and toward empirical reality.

Jocelyn: And I think the most important part of this segment is realizing that these models are not just being tweaked; they are being challenged to fundamentally change their physics.

Subrahmanyanyan: This is a massive call for a more complex and realistic look at stellar input across cosmic time, forcing us to see the true physical constraints.

Vera: I feel like we’ve covered so much ground today on the implications of this discovery, leading directly into our wrap-up.

Conclusion: Vera: We’ve seen how the observations in “Hidden mass in early galaxies revealed by bottom-heavy initial mass functions” have fundamentally changed our understanding of stellar populations in the ancient universe, and it's a huge story.

Jocelyn: We're looking at a definitive picture of hidden, low-mass stars that were simply too faint to see with older surveys, and it’s clear that this is just one aspect of the much larger story.

Subrahmanyanyan: It’s exciting because this evidence confirms that continuous star formation, rather than episodic bursts, was likely the dominant physical process in those early epochs.

Vera: It's really reassuring to know that the data has provided such a robust baseline, giving us a solid anchor for future observations and comparisons.

Jocelyn: And it feels like we’ve just witnessed the beginning of a whole new chapter in how we see galactic assembly across vast timescales.

Subrahmanyanyan: This is just about proving that our current theoretical models are fundamentally incomplete and need a more realistic look at stellar input, rather than just adding parameters to existing frameworks.

Vera: I think the most important thing to take away is that the scale of what these massive structures can achieve in their initial formation phases is much grander than we ever assumed.

Jocelyn: It’s great that we have such a clear picture of what those early systems were actually composed of, even if that population is faint and numerous.

Subrahmanyanyan: This work provides critical, hard evidence for the entire community to revise its assumptions about stellar physics and structure formation.

Vera: We've covered so much ground today on the findings in “Hidden mass in early galaxies revealed by bottom-heavy initial mass functions,” which has forced us to rethink our models significantly.

Jocelyn: It’s time to wrap this up, though, because I know we have another fascinating paper coming up that’s also pushing the boundaries of what we thought was possible.

Subrahmanyanyan: This is definitely a point where we need to keep challenging the current theories, Vera.

Vera: Absolutely, Jocelyn; let's move on to that next challenge in cosmic structure.

Chloe M. Cheng, Martje Slob, Mariska Kriek, Aliza G. Beverage, Pieter G. van Dokkum, Rachel Bezanson, Gabriel Brammer, Charlie Conroy, Anna de Graaff

Leiden Observatory, Leiden University · Observatories of the Carnegie Institution for Science · Astronomy Department, Yale University · Department of Physics & Astronomy and Pennsylvania Institute of Technology and PITT PACC, University of Pittsburgh · Cosmic Dawn Center (DAWN) · Niels Bohr Institute, University of Copenhagen · Center for Astrophysics | Harvard & Smithsonian · Max-Planck-Institut für Astronomie · Department of Astrophysics, University of Zurich · Department of Astronomy, Tsinghua University · Hong Kong Institute for Astronomy & Astrophysics

astro-ph.GA

Submitted: 2026-08-24

Updated: 2026-08-25

Comments: 39 pages, 10 figures, published in Nature Astronomy. This preprint has not undergone peer review (when applicable) or any post-submission improvements or corrections. The Version of Record of this article is published in Nature Astronomy, and is available online at https://doi.org/10.1038/s41550-026-02932-4

DOI: 10.1038/s41550-026-02932-4

Code: https://github.com/cconroy20/alf

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 70/100

The gist: JWST observations have revealed that massive galaxies formed and evolved far faster than predicted by galaxy formation models, with many having already assembled a large mass in stars about 12

Key concepts

Initial Mass Function (IMF): The IMF describes the distribution of stellar masses formed in a system. This paper focuses on 'bottom-heavy' IMFs, meaning there is an unusually high number of low-mass stars compared to what was previously assumed in galaxy formation models.
Hidden Mass: This refers to the significant population of faint, low-luminosity stars within early galaxies that were not detected by older surveys. These unseen stars contribute substantially to the total mass of these ancient objects.
Redshift (z-infinity): Redshift is a measure used in astronomy to determine how far away an object is and how much the universe has expanded since its light was emitted. The study establishes trends for this hidden mass across early cosmic structures at z-infinity.
Stellar Population Models: These are theoretical tools used to predict what stellar populations should look like based on certain assumptions about star formation. The paper suggests these models need major revisions to match the observed reality of these early galaxies.

Terminology

Abundance — 17× in this paper · explained in 1 episode
Velocity Dispersion — 11× in this paper · explained in 2 episodes
Core — 6× in this paper · explained in 2 episodes
Milky Way — 6× in this paper
Consistency — 5× in this paper · explained in 3 episodes
Calibration — 4× in this paper · explained in 7 episodes
Metallicity — 3× in this paper · explained in 3 episodes
Population Synthesis — 3× in this paper · explained in 2 episodes
James Webb Space Telescope — 2× in this paper
alpha — 2× in this paper · explained in 1 episode
Alignment — 1× in this paper · explained in 1 episode
Aperture — 1× in this paper · explained in 1 episode
Dawn JWST Archive — 1× in this paper
European Southern Observatory — 1× in this paper
Initial Mass Function (IMF) — 1× in this paper · explained in 2 episodes
Physical Constraints — 1× in this paper · explained in 7 episodes

Summary

JWST observations have revealed that massive galaxies formed and evolved far faster than predicted by galaxy formation models, with many having already assembled a large mass in stars about 12 billion years ago [1–7]. However, masses of distant galaxies are highly uncertain, as they assume a distribution of stellar birth masses (the initial mass function [IMF]) similar to that in the Milky Way (MW).

To address this uncertainty, we provide robust measurements of the IMF beyond the local Universe. We examine nine massive, quiescent galaxies at z about 0.7, observed via the JWST IMFERNO program. These targets were selected from the LEGA-C survey [8] and utilized ultra-deep spectra from LEGA-C combined with deep spectra from JWST-IMFERNO to achieve a combined rest-frame wavelength coverage of about 3700 - 10780..

To constrain the IMF, we simultaneously fit the entire rest-frame wavelength range covered by LEGA-C and IMFERNO with the absorption line fitter (alf), a flexible full-spectrum stellar population synthesis (SPS) code [12, 15]. We parameterize the IMF as a double broken power-law.

The primary finding is that these distant galaxies have excess low-mass stars. In other words, they have more bottom-heavy IMFs than the MW. For the oldest two galaxies, which are direct descendants of JWST’s “impossibly early” galaxies, the bottom-heavy IMFs increase their stellar masses by a factor of 3 – 4.

The results show that several of our galaxies have alpha IMF significantly higher than 1, indicating that they have an excess of low-mass stars compared to the MW. This is demonstrated in Figure 2, which shows the IMF mismatch parameter (alpha IMF). Furthermore, when comparing stellar and virial masses (Figure 3), we find that the stellar masses increase when applying our variable IMF, with the galaxies with the largest mass offsets shifting the most. This stellar mass increase generally brings the stellar and virial masses into agreement.

Our finding of bottom-heavy IMFs at early times is consistent with the two-phase formation model for massive galaxies. The formation redshifts of our oldest and most bottom-heavy galaxies imply that JWST’s distant, massive galaxies also had bottom-heavy IMFs at early times, possibly making them a factor of 3 – 4 more massive than originally reported. These larger masses would suggest an even greater tension with galaxy formation models [1–7].

We note that while the results are consistent when excluding individual features (such as NaI, CaT, or the Wing-Ford band), the exact shape of the IMF and trends with bottom-heaviness are not sensitive to any one feature. However, limitations exist; for instance, we fix the slope of the IMF beyond 1 M, as our method is only sensitive to the relative fraction of low-to-high-mass stars (see [12]).

Improvements for AI systems

The provided methodology offers several critical advancements that require integration into next-generation AI systems designed for high-stakes astronomical data analysis. These improvements move beyond simple pattern recognition toward robust, multi-physics parameter inference.

Improvement: Integrate a generalized, non-parametric Double Broken Power Law (DBPL) model into the core of the Stellar Population Synthesis (SPS) engine, replacing fixed functions like Kroupa or Salpeter.

Specific Implementation: The AI system must be trained to treat alpha 1 and alpha 2 (the two low-mass slopes in the DBPL) as free parameters, allowing the MCMC sampling to explore a wide range of IMF shapes beyond m=0.5 M and m=1.0 M.
Capability: The AI system can autonomously identify and quantify bottom-heavy populations (alpha IMF > 1 relative to MW/Kroupa) across entire galaxy samples, not just individual targets, providing a statistically significant measure of hidden stellar mass that was previously obscured by fixed IMF assumptions.

Improvement: Develop a unified data processing module capable of simultaneously fitting heterogeneous datasets—specifically, merging low-resolution optical spectra (LEGA-C) with high-resolution near-infrared data (JWST-IMFERNO).

Specific Implementation: The AI must handle the flux scaling and spectral continuity across vastly different rest-frame wavelength ranges (about 3700 Å to about 10780 Å), ensuring that the fitting process is globally constrained by forcing the data from multiple surveys to be on a single, unified flux scale.
Capability: The AI system can break strong degeneracies between elemental abundance patterns and IMF variations by leveraging the combined spectral coverage, enabling it to provide precise chemical compositions alongside a robust estimate of stellar mass.

Improvement: Implement an automated Differential Velocity Dispersion Corrector that accounts for instrumental and kinematic misalignment when comparing data from different surveys (e.g., LEGA-C vs. IMFERNO).

Specific Implementation: The AI must perform the iterative correction shown in Figure 7: calculating sigma v, IMFERNO / sigma v, LEGA-C and applying this ratio to the virial masses (M vir) for galaxies where misalignment is detected.
Capability: The AI system can eliminate systematic errors in mass estimation caused by differing slit orientations or rotational support, ensuring that derived M/L ratios are physically accurate across a diverse sample of early-type galaxies, minimizing false positives in mass determination.

Improvement: Refine the Markov Chain Monte Carlo (MCMC) optimization process to handle the simultaneous fitting of multiple physical parameters: Age, 18 elemental abundances, and two IMF slopes (alpha 1, alpha 2).

Specific Implementation: The AI must be trained to manage a high-dimensional posterior distribution, using techniques like informed sampling or nested sampling rather than standard MCMC chains to efficiently explore the parameter space defined by Equation 1.
Capability: The AI system can provide reliable, statistically robust confidence intervals (plus or minus 1 sigma) for both stellar population parameters and IMF variations simultaneously, delivering a level of certainty that far exceeds current observational limitations.

Abstract

James Webb Space Telescope (JWST) observations have revealed that massive galaxies formed and evolved faster than predicted by galaxy formation models, with many having already assembled a large mass in stars approximately 12 billion years ago. However, masses of distant galaxies are uncertain, as they assume a distribution of stellar birth masses (the initial mass function (IMF)) similar to that in the Milky Way. Specifically, the contribution from low-mass stars, which make up the bulk of stellar mass, is not directly observed, but inferred based on an extrapolation of the Milky Way IMF. Here, we provide robust constraints on the low-mass IMF beyond the local Universe from full-spectrum models. Using ultra-deep spectra of nine massive quiescent galaxies at redshift z about0.7 from the JWST Initial Mass Function of Early Red NIRSpec Objects program, extended to bluer wavelengths with deep Very Large Telescope Large Early Galaxy Astrophysics Census spectra, we find that the most massive galaxies have excess low-mass stars. Remarkably, our oldest galaxy (formation redshift z form > 5) has the most bottom-heavy IMF. This galaxy may be a descendant of JWST's 'impossibly early' galaxies, implying that the latter may have had similarly bottom-heavy IMFs increasing their masses by a factor of approximately 4 plus or minus1. Our findings may thus amplify the tension with galaxy formation models.