PI: Debora Sijacki

The James Webb Space Telescope (JWST) is revolutionizing our understanding of the properties of galaxies all the way back to the very early Universe, with a leap in sensitivity, especially in measuring spectra, of more than three orders of magnitude. JWST observations have led to numerous surprising results and have enabled new avenues of research in galaxy formation. One salient example is the inference of bursty star formation in high-redshift galaxies, whereby galaxies likely go through cycles of intense star formation followed by periods of virtually no star formation, i.e., quiescence.

These first quiescent galaxies forming in the early Universe offer a unique probe into the physical mechanisms behind the cessation of star formation, as the cosmic time of their assembly is particularly short. Specifically, spectroscopic JWST studies suggest a fast build-up of stellar mass within the first one or two billion years of the Universe, followed by rapid suppression of star formation, also known as `quenching’, within a few tens of millions of years, which is challenging our current understanding of galaxy formation.

While the majority of early quiescent galaxies identified to date are massive systems, JWST is opening a new window on the first generations of quiescent low-mass galaxies populating the high-redshift Universe. A recently observed quiescent galaxy, JADES-GS-z7-01-QU, has a remarkably low stellar mass at a redshift z = 7.3.

To understand the likely physical mechanisms at play that led to the formation of objects like JADES-GS-z7-01-QU, we took advantage of novel galaxy formation simulations from the Azahar suite (see Fig. 1). These simulations, performed with the adaptive mesh-refinement RAMSES code on DiRAC facilities as a part of dp012 project, include galaxy formation models ranging from magneto-hydrodynamics with a magneto-thermo-turbulent prescription for star formation to more sophisticated setups incorporating full on-the-fly radiative transfer and cosmic-ray physics.

We analyzed a sample of 10 simulated galaxies at very high resolution in the early Universe and found that the inclusion of cosmic rays and radiative transfer leads to much burstier star-formation histories, as well as a greater variety of star-formation burst intensities. We generated a forward-modeled mock spectrum during a `mini-quenching’ event at redshift z = 7.5 and compared it with observational data for the high-redshift quiescent galaxy JADES-GS-z7-01-QU, showing very good agreement.

This study underscores the critical role of incorporating comprehensive physical processes, such as cosmic rays and radiative transfer, into hydrodynamical simulations to capture more accurately the burstiness of star formation in high-redshift galaxies. By quantifying the burstiness of galaxies in Azahar using observable metrics, such as detailed mock spectra, we are able to bridge the gap between simulations and observations, aiding the physical interpretation of JWST findings. Refining our understanding of galaxy formation models through such detailed comparisons with JWST data will ultimately shed light on the puzzling early stages of galaxy formation.

This article is based on results published as Dome T. et al., MNRAS 537, 629 (2025).

Figure 1. Central panel: Large-scale view of the Azahar simulation capturing galaxy formation in the very early Universe. Galaxies (bright white compact regions) form at the intersection of cosmic filaments. Surrounding panels: Zoom-in views into a varied collection of galaxies from the Azahar simulation, ranging from massive spiral systems to complex galaxy mergers. Colours show neutral hydrogen (blue), ionized hydrogen (gray), JWST stellar emission (white), and emanating radiation in pink-orange hues.