Tracing the Evolution of the Balmer Break from Cosmic Dawn to Cosmic Noon with JWST

Adarsh Kuruvanthodi, Daniel Schaerer, Rui Marques-Chaves, Andrea Weibel, Damien Korber, Corinne Charbonnel

First listed 2026-07-09 | Last updated 2026-07-08

Abstract

The Balmer break (BB) is a key spectral feature for constraining stellar population ages, star formation histories, and redshifts of high-redshift sources. The redshift evolution and distribution of BB strength, together with the properties of BB galaxies, constrain stellar population characteristics and the nature of star formation across cosmic epochs. However, a systematic and unbiased characterization of BB strengths across the full galaxy population remains limited with the James Webb Space Telescope (JWST). We aim to characterize the redshift evolution of BB strength over $z=3.5$-10 and its distribution across different epochs using photometry. We also examine correlations between BB strength and key physical parameters within these redshift intervals. We further assess the implications of BB galaxies for the nature of star formation and stellar populations at $z>3.5$. We used the JWST NIRCam photometric observations taken as part of various programs, including CEERS, JADES, FRESCO, and PRIMER. We estimated the BB strength of the objects with two adjacent broadband filters in various redshift windows between redshifts 3 and 10, which exclude strong line contamination. We employ the SED-fitting code CIGALE for both SED fitting and the generation of mock galaxy simulations. We find that the median Balmer break strength (expressed as a flux ratio) increases from cosmic dawn to cosmic noon, from 1.1 to 1.5, primarily driven by the age of the stellar population. These estimations are in agreement with the latest spectroscopic estimations in the literature. We identify objects with extremely large BB strengths (BB$>3.0$) at $z=3.5$-4 and $z=7$-10, indicating strong extinction combined with an old stellar population and the presence of Little Red Dots (LRDs) in the former, and predominantly LRDs in the latter.[abridged]

Short digest

Using JWST/NIRCam photometry from CEERS, JADES, FRESCO, and PRIMER, this paper measures Balmer-break strengths with adjacent broadband filters chosen to avoid strong line contamination, building a population-level view across z=3.5-10 rather than focusing only on individual spectroscopic standouts. The headline result is that the median Balmer break strengthens from about 1.1 at cosmic dawn to about 1.5 by cosmic noon, in line with recent spectroscopic estimates and primarily driven by older stellar populations at lower redshift. Interpreting the measurements with CIGALE-based models, the characteristic stellar ages decline toward high redshift, from roughly 350 Myr for constant star formation and 50 Myr for bursty histories at lower-z bins to about 20 and 10 Myr, respectively, at z>7. The most extreme BB>3 systems appear at z=3.5-4 and z=7-10, linking the strongest breaks to dusty old populations plus Little Red Dots in the former interval and to predominantly LRD-like sources in the latter.

Key figures to inspect

• Figure 5. This is the core population-level evidence: it shows the full Balmer-break distributions in each redshift bin, their medians, and where they fall relative to the minimum and maximum values expected from constant-star-formation and instantaneous-burst models. It is the best figure for seeing both the typical BB strength at each epoch and the high-BB tail that exceeds classical stellar-population expectations.
• Figure 6. This is the headline evolution plot, summarizing how the median Balmer break changes with redshift for both the photometric and spectroscopic samples. It directly supports the paper’s main claim that BB strength rises from cosmic dawn toward cosmic noon and shows consistency with external spectroscopic and stacked-spectrum estimates.
• Figure 7. This figure carries the physical interpretation by showing how BB strength correlates with stellar mass, age, UV slope beta, sSFR, and EW(H-alpha) across multiple redshift bins. It is where the paper demonstrates that stronger breaks are generally associated with older, more massive, redder, and less specifically star-forming systems.
• Figure 8. This comparison figure places the sample alongside literature Balmer-break sources and theoretical model envelopes, making it especially useful for understanding where the extreme objects sit relative to normal stellar populations. It is also one of the clearest ways to connect the paper’s strongest-break outliers to the broader discussion of unusual high-redshift systems, including Little Red Dots.
• Figure 14. This figure translates the observed break strengths into age distributions for the large photometric sample under both instantaneous-burst and constant-star-formation assumptions. It matters because it makes the inferred age evolution concrete and also reveals that some of the strongest high-redshift BB sources drive implausibly old ages in the constant-star-formation interpretation.