The Astrophysical Journal (Jan 2025)

Metallicity Gradients in Modern Cosmological Simulations. I. Tension between Smooth Stellar Feedback Models and Observations

  • Alex M. Garcia,
  • Paul Torrey,
  • Aniket Bhagwat,
  • Ruby J. Wright,
  • Qian-Hui Chen,
  • Kathryn Grasha,
  • Sophia Ridolfo,
  • Z. S. Hemler,
  • Arnab Sarkar,
  • Priyanka Chakraborty,
  • Erica J. Nelson,
  • Ryan L. Sanders,
  • Tiago Costa,
  • Mark Vogelsberger,
  • Lisa J. Kewley,
  • Sara L. Ellison,
  • Lars Hernquist

DOI
https://doi.org/10.3847/1538-4357/adea51
Journal volume & issue
Vol. 989, no. 2
p. 147

Abstract

Read online

The metallicity of galaxies, and its variation with galactocentric radius, provides key insights into the formation histories of galaxies and the physical processes driving their evolution. In this work, we analyze the radial metallicity gradients of star-forming galaxies in the EAGLE, Illustris, IllustrisTNG, and SIMBA cosmological simulations across broad mass (10 ^8.0 M _⊙ ≤ M _⋆ ≲ 10 ^12.0 M _⊙ ) and redshift (0 ≤ z ≤ 8) ranges. We find that all simulations predict strong negative (i.e., radially decreasing) metallicity gradients at early cosmic times, likely due to their similar treatments of relatively smooth stellar feedback not providing sufficient mixing to quickly flatten gradients. The strongest redshift evolution occurs in galaxies with stellar masses of 10 ^10.0 –10 ^11.0 M _⊙ , while galaxies with stellar mass 10 ^11 M _⊙ exhibit weaker redshift evolution. Our result of negative gradients at high redshift contrast with the many positive and flat gradients in the 1 6, the negative gradients observed with JWST and the Atacama Large Millimeter/submillimeter Array are flatter than those in simulations, albeit with closer agreement than at lower redshift. Overall, we suggest that these smooth stellar feedback galaxy simulations may not sufficiently mix their metal content radially, and that either stronger stellar feedback or additional subgrid turbulent metal diffusion models may be required to better reproduce observed metallicity gradients.

Keywords