← back

The Stellar-Mass Dependence of Feedback-Driven Outflows in DESI DR1 Star-Forming Galaxies

πŸ“„ arXiv:2611.02580 Β· πŸ“₯ PDF Β· 2026-11-02 Β· astro-ph.CO

Authors: Yimeng Tang [arXiv Β· scholar] , Yifei Luo [arXiv Β· scholar] , Conghao Zhou [arXiv Β· scholar]

πŸ“„ Abstract

Stellar feedback plays a key role in regulating star formation and shaping the evolution of star-forming galaxies. Low-mass galaxies provide one of the most powerful laboratories for testing stellar feedback and its impact on galaxy evolution. Numerical simulations predict that repeated feedback-driven gas inflows and outflows generate a "breathing mode," producing a positive correlation between mass-normalized gas velocity dispersion and recent star formation in dwarf galaxies. Whether this kinematic signature persists at higher stellar masses has remained untested. We investigate this question using $\sim1200$ star-forming galaxies with Dark Energy Spectroscopic Instrument (DESI) spectroscopy and COSMOS2020 photometry spanning $7.9<\log M_\star/M_\odot<11.5$ within $z<0.25$. Ionized-gas kinematics are measured with pPXF, and recent star formation histories are reconstructed through joint spectrophotometric fitting with Prospector. We confirm the predicted positive correlation throughout the low-mass regime with $\log M_\star/M_\odot \lesssim 9.8$, quantitatively consistent with previous observations and FIRE-2 simulations. Above $\log M_\star/M_\odot \sim 9.8$, the correlation weakens rapidly and becomes consistent with zero, providing the first empirical constraint on the stellar mass scale above which the kinematic signature of feedback-driven outflow cycles becomes negligible. Our results establish a characteristic stellar mass above which stellar feedback no longer dominates galaxy-scale gas kinematics, providing a new observational benchmark for models of feedback-regulated galaxy evolution. Future joint analyses of gas kinematics and stellar structure will test whether this transition is accompanied by the broader galaxy response predicted by the breathing mode scenario.

πŸ“„ arXiv abstract page πŸ“₯ PDF