The ASTRID simulation: the evolution of Supermassive Black Holes

Abstract

We present the evolution of black holes (BHs) and their relationship with their host galaxies in Astrid , a large-volume cosmological hydrodynamical simulation with box size 250 ℎ −1 Mpc containing 2 × 5500 3 particles evolved to 𝑧 = 3. Astrid statistically models BH gas accretion and AGN feedback to their environments, applies a power-law distribution for BH seed mass 𝑀 sd , uses a dynamical friction model for BH dynamics and executes a physical treatment of BH mergers. The BH population is broadly consistent with empirical constraints on the BH mass function, the bright end of the luminosity functions, and the time evolution of BH mass and accretion rate density. The BH mass and accretion exhibit a tight correlation with host stellar mass and star formation rate. We trace BHs seeded before 𝑧 > 10 down to 𝑧 = 3, finding that BHs carry virtually no imprint of the initial 𝑀 sd except those with the smallest 𝑀 sd , where less than 50% of them have doubled in mass. Gas accretion is the dominant channel for BH growth compared to BH mergers. With dynamical friction, Astrid predicts a significant delay for BH mergers after the first encounter of a BH pair, with a typical elapse time of about 200 Myrs. There are in total 4.5 × 10 5 BH mergers in Astrid at 𝑧 > 3, ∼ 10 3 of which have X-ray detectable EM counterparts: a bright kpc scale dual AGN with 𝐿 X > 10 43 erg s −1 . BHs with 𝑀 BH ∼ 10 7−8 𝑀 experience the most frequent mergers. Galaxies that host BH mergers are unbiased tracers of the overall 𝑀 BH − 𝑀 * relation. Massive (> 10 11 𝑀 ) galaxies have a high occupation number ( 10) of BHs, and hence host the majority of BH mergers.

The ASTRID Simulation: Galaxy Formation and Reionization

“…We showed that hydrogen reionization has a strong effect on star formation rates and gas fractions. Our companion paper (Ni et al 2021) examines the statistics of black holes.…”

“…We include a treatment of super-massive black holes (SMBHs). The SMBH model is discussed in more detail in Ni et al (2021) and summarised here. Our accretion and feedback models are similar to those in the BlueTides simulation Feng et al (2016a), and are based on earlier work by Springel et al (2005); Di .…”

“…In addition to describing the techniques used in performing the simulation, we present some initial results for the population of high redshift galaxies, compared to available observations in Section 3. Results for the black hole population may be found in our companion paper (Ni et al 2021).…”

Machine Learning and Cosmology

“…(1) simulations must be produced at large HPC facilities, using O(100K-1M) GPU-hours (e.g. AbacusSummit [77]; Uchuu [78]); or O(10M-100M) CPU-hours [79][80][81][82] (2) outputs are stored on large, high-throughput file systems, totalling O(1 PB); (3) the analysis (generation of data products) is executed in cluster environments (often using CPUs) requiring a small percent of the cost incurred in running the simulations; (4) ML models are trained on these data products (using GPUs or other accelerators). The next decade will only see these requirements increase, as upcoming surveys explore both broader sky area and greater depth, demanding simulations with both larger volume and finer resolution.…”

Massive Black Hole Mergers with Orbital Information: Predictions from the ASTRID Simulation

“…Furthermore, we use a dynamical friction model (tested and validated in Chen et al 2021) to evolve the binary black holes and include the sinking and merger of MBHs in the simulation in a more physical way. Here we briefly summarize the black hole seeding and dynamics treatment in Astrid, and refer to and Ni et al (2021) for detailed presentations of physical models for star formation and black holes.…”