Cosmic Stillness: High Quiescent Galaxy Fractions Across Upper Mass Scales in the Early Universe to z = 7 with JWST

TL;DR

Using JWST data from CEERS, NEP, and JADES, this work identifies a robust sample of high-redshift quenched galaxies (3 < z < 7) via sSFR-based BAGPIPES SED fitting, revealing strong evolution in the massive passive fraction and corresponding declines in comoving number densities with redshift. Morphological analysis with GALFIT shows passives are predominantly compact with high Sérsic indices, and multiwavelength data uncover AGN signatures in a subset of the sample. The study highlights UVJ's limitations at z > 3 and demonstrates the efficacy of an sSFR-driven approach for detecting recently quenched systems, including at z > 6.5 where a population of Little Red Dots emerges, underscoring the need for deeper spectroscopy to confirm their nature. Overall, the results constrain early quenching pathways and provide a benchmark for future wide-area JWST surveys and theoretical models.

Abstract

We present a detailed investigation into the abundance and morphology of high redshift quenched galaxies at $3 < z < 7$ using James Webb Space Telescope data in the NEP, CEERS and JADES fields. Within these fields, we identify 90 candidate passive galaxies using specific star formation rates modelled with the BAGPIPES SED fitting code, which is more effective at identifying recently quenched systems than the classical UVJ method. With this sample of galaxies, we find number densities broadly consistent with other works and a rapidly evolving passive fraction of high mass galaxies ($\log_{10}{(M_{\star}/M_{\odot})} > $ 9.5) between $3 < z < 5$. We find that the fraction of galaxies with low star formation rates and mass 9.5 $ < \log_{10}{(M_{\star}/M_{\odot})} < $ 10.5 decreases from $\sim$25% at $3 < z < 4$ to $\sim$2% at $5 < z < 7$. Our passive sample of galaxies is shown to exhibit more compact light profiles compared to star-forming counterparts and some exhibit traces of AGN activity through detections in either the X-ray or radio. At the highest redshifts ($z > 6.5$) passive selections start to include examples of 'little red dots' which complicates any conclusions until their nature is better understood.

Figures (11)

• Figure 1: Example of fitted spectral energy distribution to the galaxy in NEP field with the unique ID 55884, whose redshift is fit by Le PHARE. The black points represent measurements from the JWST after calibration, and the orange line represents the fitted SED. Also, the lighter orange line represents the whole fit, while the darker orange marker represents where the measured band is being fitted. The Balmer break is clearly visible, and the fitted SED after the Balmer break is within our measured points, hence this is an example of an object passing the visual inspection.
• Figure 2: The initial, lenient, Le Phare passive sample which was then run through BAGPIPES to obtain the final passive sample plotted on UVJ diagrams. Galaxies have been split into normal (star-forming), passive and robust passive categories, as well as redshift and mass bins, as indicated in the legend. Where passive and robust passive signify that 50 % or 97.5% respectively, of the galaxy's sSFR PDF falls below the threshold set by Equation \ref{['eq:sSFR']}.
• Figure 3: Redshift versus mass diagram for the passive sample defined by our sSFR cut in Equation \ref{['eq:sSFR']}. The orange dots indicate the quiescent galaxies that have not been selected via UVJ selection. The blue squared indicate the quiescent galaxies that have been selected via UVJ selection.
• Figure 4: The evolution of co-moving number density of massive quenched galaxies across redshift and mass cuts for our combined fields. The x-axis values are averages of the redshift bin intervals they span. The uncertainties in the figure incorporate cosmic variance as well as the Poisson noise. We also include values from previous work, Valentino_2023carnall2023, for comparison.
• Figure 5: The evolution of comoving number density of massive robustly quenched galaxies across redshift and stellar mass cuts for our combined fields. The x-axis values are averages of the redshift bin intervals they span. The uncertainties in the figure incorporate cosmic variance as well as the Poisson noise. We also include values from previous work, Valentino_2023carnall2023, for comparison.