Quenching that shapes the first billion years

by Akash Vani (MPA Garching)

Monday 15 Dec 2025, 14:30 → 15:30 Asia/Kolkata

A-304 and Zoom

Recent JWST observations of the first billion years revealed rapid quenching, structurally evolved galaxies and massive galaxies, exposing tensions with current numerical simulations. I present an upgraded semi-analytic framework called LGalaxies built on top of Gadget-4 in which we systematically tested some of the proposed solutions and introduce a self-consistent, physically motivated prescription for high-efficiency star formation and 'regulated' stellar feedback based on gas surface densities at high redshift within the ΛCDM cosmology. We also update galaxy merging and structural evolution by incorporating gas density-dependent processes and black-hole fueling channels, such as cold-gas inflows triggered by disk instabilities that are frequent at intermediate z. These mechanisms drive morphological transformation and compact galaxy sizes. The upgrades produce a boosted UV luminosity function over 10<z<13, more compact and massive quenched galaxies across 0<z<4, and, owing to already implemented advanced environmental effects, quenched satellite populations already at z>4. Across 0<z<13, from dwarfs to the most massive galaxies, the model reproduces key observables such as stellar mass functions, galaxy size - mass relations, number densities of quenched systems, and the UV luminosity function, while maintaining agreement for the galaxy population at z=0. Methodologically, we introduce a new calibration framework that is ~10× faster than prior implementations in the framework and a redesigned codebase that substantially improves speed and efficiency. We also describe ongoing efforts, such as a high-resolution dark-matter simulation tailored to early-universe galaxy formation, deep learning accelerated parameter exploration, and new drivers of morphological evolution relevant to high-z galaxies. Taken together, these advances, implemented within the fast, modular LGalaxies framework, offer a viable path to resolving several JWST-driven discrepancies and provide a computationally efficient platform for interpreting early galaxy formation.