Search for: All records

Abstract Galaxies grow alongside central supermassive black holes (SMBHs) through fueling and feedback. However, the origins of this coevolution remain unclear and vary across modeling frameworks. Using semianalytic models (SAMs), we trace SMBH mass assembly acrossM_BH ∼ 10⁶⁻¹⁰M_⊙. We find significant discrepancies between observations and physics-based models of the local black hole mass function (BHMF), likely from differences in the stellar mass function and scaling relations used to infer the BHMF. Most physics-based models agree atz ∼ 1–4 and broadly match the JWST broad-line active galactic nucleus (AGN) BHMFs atz = 4–5. These models also reproduce the observed bolometric AGN luminosity evolution, except the SAMDark Sage, which predicts an excess. Interestingly, this pronounced “knee” in the bolometric AGN luminosity function predicted byDark SagearoundL_bol ∼ 10⁴⁶erg s⁻¹is consistent with the inferred abundance and luminosity of “little red dots” atz = 5–6, under the assumption that they are powered entirely by AGN activity. In contrast to other models,Dark Sagedeploys multiple growth channels for SMBHs that include mergers, hot-mode accretion, merger-driven cold-accretion, and secular-instability-driven accretion. We analyze the black hole mass buildup and accretion histories inDark Sage, which, unlike other models, also allows for super-Eddington accretion, and we find that, on average, SMBHs primarily grow through secular disk instabilities and merger-driven cold gas accretion modes. We also find that black hole mergers contribute the majority of the growth of ∼60% of the total mass budget only for the most massive SMBHs byz= 0. 

Abstract As the next generation of large galaxy surveys come online, it is becoming increasingly important to develop and understand the machine-learning tools that analyze big astronomical data. Neural networks are powerful and capable of probing deep patterns in data, but they must be trained carefully on large and representative data sets. We present a new “hump” of the Cosmology and Astrophysics with MachinE Learning Simulations (CAMELS) project: CAMELS-SAM, encompassing one thousand dark-matter-only simulations of (100 h −1 cMpc) 3 with different cosmological parameters (Ω m and σ 8 ) and run through the Santa Cruz semi-analytic model for galaxy formation over a broad range of astrophysical parameters. As a proof of concept for the power of this vast suite of simulated galaxies in a large volume and broad parameter space, we probe the power of simple clustering summary statistics to marginalize over astrophysics and constrain cosmology using neural networks. We use the two-point correlation, count-in-cells, and void probability functions, and we probe nonlinear and linear scales across 0.68 < R <27 h −1 cMpc. We find our neural networks can both marginalize over the uncertainties in astrophysics to constrain cosmology to 3%–8% error across various types of galaxy selections, while simultaneously learning about the SC-SAM astrophysical parameters. This work encompasses vital first steps toward creating algorithms able to marginalize over the uncertainties in our galaxy formation models and measure the underlying cosmology of our Universe. CAMELS-SAM has been publicly released alongside the rest of CAMELS, and it offers great potential to many applications of machine learning in astrophysics: https://camels-sam.readthedocs.io .