ASTRID and TNG300 simulations map distinct black-hole evolutionary channels from z=2 to z=0.5
Two large cosmological simulations demonstrate that black-hole location in scaling relations encodes merger, stripping, and wandering histories from cosmic noon onward. The work connects local dwarf-galaxy observations to high-redshift quasar growth through environment-dependent channels that single-object studies miss.
The study follows individual black holes rather than host galaxies across two large-volume cosmological simulations, capturing central, satellite, and off-nuclear wandering BHs. This approach reveals that high-mass centrals grow via mergers then quench hosts through kinetic AGN feedback, while tidal stripping shifts satellites to lower stellar mass at fixed black-hole mass, creating weakly accreting systems. Wandering BHs in ASTRID arise from satellite accretion and remain undermassive with minimal subsequent growth.
These channels occupy distinct loci in both the M_BH-M_star and specific accretion rate–specific star-formation rate planes, acting as fossils of dynamical and feedback history. The result links local dwarf-galaxy overmassive black holes to the rapid early growth seen in high-redshift quasars, showing how environment since cosmic noon sets the scatter observed today.
The analysis highlights under-covered connections between AGN activity in dense environments and the assembly of the local black-hole mass function. Future wide-field surveys and deeper JWST spectroscopy of low-mass quenched systems can test the predicted fractions of stripped and wandering black holes.
Key limitation is reliance on sub-grid prescriptions for dynamical friction and AGN feedback; higher-resolution zoom simulations with resolved nuclear dynamics would strengthen the predictions.
Weller et al.: Deep JWST spectroscopy of quenched satellites at 1<z<2 will detect overmassive central BHs with M_BH/M_star ratios exceeding local relations by a factor of three in at least 15 percent of systems by 2028.
Sources (3)
- [1]Primary Source(https://arxiv.org/abs/2607.07793)
- [2]Supporting Source(https://arxiv.org/abs/2305.13390)
- [3]Supporting Source(https://ui.adsabs.harvard.edu/abs/2023ApJ...952..133R)