First results of AMBRA : Abundant Seeds and Early Mergers as a Pathway to the First Massive Black Holes
Yihao Zhou (周亦豪), Aklant Kumar Bhowmick, Tiziana Di Matteo, Patrick LaChance, Rupert Croft, Laura Blecha, Simeon Bird, Paul Torrey, Lars Hernquist · 2026
Read First results of AMBRA : Abundant Seeds and Early Mergers as a Pathway to the First Massive Black Holes by Yihao Zhou (周亦豪), Aklant Kumar Bhowmick, Tiziana Di Matteo, Patrick LaChance, Rupert Croft, Laura Blecha, Simeon Bird, Paul Torrey, Lars Hernquist (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-01123.txt.
Introduction
Zhou and colleagues present AMBRA, combining ASTRID's volume with BRAHMA's gas-based black hole seeding models to explain JWST's massive black holes. AMBRA uses a heavy-seed prescription with seeds of ten thousand to a hundred thousand solar masses. AMBRA's redshift-8 black hole density runs over ten times ASTRID's. Black holes like GN-z11 and CEERS-1019 merge early, with mergers supplying about 50 percent of mass by redshift 11. AMBRA predicts about 4 LISA-detectable mergers yearly at redshift 8 and above, three orders of magnitude above ASTRID.
Catalog record
| CREATOR | Yihao Zhou (周亦豪), Aklant Kumar Bhowmick, Tiziana Di Matteo, Patrick LaChance, Rupert Croft, Laura Blecha, Simeon Bird, Paul Torrey, Lars Hernquist |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 71423 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.01123
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
First results of AMBRA : Abundant Seeds and Early Mergers as a Pathway to the First Massive Black Holes Yihao Zhou (周亦豪), Aklant Kumar Bhowmick, Tiziana Di Matteo, Patrick LaChance, Rupert Croft, Laura Blecha, Simeon Bird, Paul Torrey, Lars Hernquist Abstract We present the first results from AMBRA simulation ( A STRID with M BH seeding from BRA HMA) evolved to z = 8 z=8 . AMBRA combines the large cosmological volume and statistical power of ASTRID with the physically motivated gas-based black hole seeding models from BRAHMA . Motivated by JWST’s discoveries of massive black holes (BHs) at z ≳ 9 z\gtrsim 9 , AMBRA adopts a lenient heavy-seed prescription from the BRAHMA suite, allowing for the formation of 4 × 10 4 − 5 4\times 10^{4-5} M ⊙ M_{\odot} seeds in halos with star-forming, metal-poor gas. The seeding model is motivated by scenarios in which heavy seeds form through stellar collisions in star clusters or from the rapid growth of Population III remnants. The improved seeding model enables AMBRA to form BH seeds much earlier and more efficiently compared to ASTRID . This significantly enhances early BH growth, producing a z = 8 z=8 BH number density more than an order of magnitude higher than that in ASTRID over the mass range 10 5 − 7 10^{5-7} M ⊙ M_{\odot} . BHs reaching masses consistent with GN-z11 and CEERS-1019 typically originate in highly compact density peaks and undergo multiple early mergers. In these systems, ∼ 50 % \sim 50\% of BH masses by z = 11 z=11 is from BH mergers, after which gas accretion becomes the dominant growth channel. Without this early merger-driven assembly, ASTRID cannot reproduce the high-mass BH detected by JWST. Our results indicate that abundant early seed formation combined with frequent mergers can explain several JWST massive BH candidates without requiring sustained super-Eddington accretion. As a testable prediction, AMBRA yields ≈ 4 \approx 4 LISA detectable BH merger events per year at z ≥ 8 z\geq 8 , which is three orders of magnitude higher than that in ASTRID . I Introduction In recent years, the James Webb Space Telescope (JWST) has rapidly transformed our view of massive black hole (BH) assembly in the early universe by unveiling a growing population of massive BHs at high redshift. While quasars at z\gtrsim 6 had already established that black holes with M_{\rm BH}\gtrsim 10^{8-9} M_{\odot} exist within the first Gyr ( 87 ; 91 ; 28 ) , JWST is now identifying a much larger number density of low-luminosity AGN and BH candidates at similar redshifts ( 34 ; 51 ; 78 ) . Beyond increasing the sample size, JWST provides new constraints on the host environments of these high- z massive BHs. Based on JWST/NIRCam observations, the EIGER Collaboration ( 41 ) found a strong diversity among the observed quasar fields at z\geq 6 ( 27 ) . Using a compilation of JWST-identified low-luminosity AGNs at 5<z<6 , 2 performed a clustering analysis and provided an empirical estimate of their host dark matter (DM) halo mass: \log M_{\rm halo}\approx 10^{11}\ h^{-1}M_{\odot} . These discoveries suggest that early BH formation and growth may be both more common and more diverse than implied by the pre-JWST quasar census. Moreover, JWST is also pushing the redshift frontier for AGN by detecting a handful of objects at z\sim 9-11 , such as CEERS-1019 GN-z11 , and UHZ1 ( 47 ; 49 ; 12 ; 31 ) , which challenge our understanding of BH formation and growth theories. Explaining the emergence of \sim 10^{6-7} M_{\odot} black holes by z=9 is especially difficult ( 85 ; 39 ; 37 ) . Even under optimistic assumptions of continuous Eddington-limited accretion, the available time is short and requires a near-unity duty cycle. One possibility is to invoke sustained episodes of super-Eddington accretion ( 66 ; 36 ; 48 ) . However, feedback can substantially reduce the efficiency of such rapid growth ( 64 ; 50 ; 65 ) . As a result, producing the most massive JWST-era black hole candidates through accretion alone remains extremely challenging. Large-volume cosmological simulations are essential for studying this problem because they can self-consistently capture the complex interplay between BH growth and the evolving galaxy population across a wide range of environments. The ASTRID simulation ( 11 ; 57 ; 93 ) was designed to provide statistically representative predictions for BH evolution in a large volume ( 250\,h^{-1}{\rm Mpc} per side) across cosmic history. ASTRID successfully reproduces a range of pre-JWST constraints on galaxies and BHs at high redshift ( 57 ; 56 ) , including the BH mass function, luminosity function, and M_{\rm BH}-M_{\rm gal} scaling relations. However, like many other cosmological simulations in the literature, ASTRID adopts a BH seeding prescription based purely on a threshold halo mass. This prescription cannot distinguish between different physical seeding channels. At the same time, ASTRID does not produce the earliest, rapidly growing BH population implied by JWST detections, highlighting the need for improved BH formation models. Several cosmological simulations have begun to implement more physically motivated prescriptions for BH seeding that depend on local gas conditions ( 80 ; 82 ; 26 ; 15 ; 8 ; 9 ) . In particular, the BRAHMA suite ( 7 ; 4 ; 8 ; 10 ; 9 ) has systematically explored a wide range of gas-based seeding scenarios using prescriptions tied to properties such as gas density, metallicity, Lyman Werner (LW) radiation, gas spin, and environmental richness. These studies demonstrate that the choice of seed model can strongly influence early BH growth and the resulting high-redshift BH populations observable with JWST. 6 used constrained BRAHMA cosmological simulations of rare ( 5\sigma ) overdense peaks to study the assembly of the earliest z\sim 9 –11 BHs discovered by JWST under systematic variations of heavy seed models ( \sim 10^{4} – 10^{5}\,M_{\odot} ). They found that, under standard assumptions for stellar and AGN feedback, exceptionally high abundances of heavy seeds are required to assemble BHs with the current mass estimates of GN-z11 and CEERS-1019 . The required abundances are significantly larger than what is typically expected under canonical DCBH formation scenarios, which require extremely high LW fluxes ( \gtrsim 1000\,J_{21} ; 72 ; 76 ). This is mainly because AGN feedback strongly suppresses early BH accretion in these BRAHMA simulations. With a large enough number of heavy \sim 10^{5}~M_{\odot} seeds, mergers can significantly accelerate BH growth at early times, which subsequently also boosts accretion at later times. In contrast, lower-mass seeds ( \lesssim 10^{4}\,M_{\odot} ), even if formed in larger numbers, grow more slowly due to longer merger delay times resulting from weaker dynamical friction. From a physical perspective, several promising pathways have been proposed that could produce heavy ( \sim 10^{4} – 10^{5}\,M_{\odot} ) seeds more efficiently. These include enhanced stellar and BH collisions in ultra-dense ( \gtrsim 10^{8}\,\mathrm{cm^{-3}} ) nuclear star clusters (NSCs) ( 45 ; 63 ) , supra-exponential BH accretion in NSCs ( 55 ) , rapid BH mergers of \sim 10^{4}\,M_{\odot} seeds forming in feedback-free star clusters within early proto-galaxies via gravo-gyro instability ( 22 ) , hyper-Eddington growth of a small fraction of light Pop III seeds