arXiv subjects

Peter Behroozi

Publications and source records attributed to Peter Behroozi.

At least 19 recordsLinked to original sources

arXiv2026

Why a Multi-Cycle Treasury Program is Essential to Maximize JWST's Scientific Contributions to Galaxy Formation

The Genesis Initiative is an extragalactic community-wide effort to coordinate white papers in response to the STScI Multi-Cycle Treasury (MCT) concept call. Two of JWST's primary science goals are to discover and characterize the first stars and galaxies, and to understand how they evolved into the galaxies we see today - questions whose fundamental importance has been affirmed by the past three decadal surveys. JWST's first five years have surpassed expectations, revealing a bounty of unexpected discoveries about the early universe. Yet these discoveries have raised as many questions as they have answered, and key mission goals remain unmet. Transforming this excitement into a quantitative physical understanding of early galaxy formation requires a uniform, statistically controlled legacy dataset that the current portfolio of smaller, heterogeneous programs cannot provide. The history of the Hubble Space Telescope offers a clear precedent. Its most transformative extragalactic results came from large, coordinated surveys far exceeding what the standard time-allocation process typically approves, and these uniform datasets remain scientifically productive decades later. We argue that JWST is now poised for the same leap. Answering the field's central open questions demands coordinated, community-led surveys of ~1000-2000 hr, with uniform and well-characterized selection functions that are structurally impossible within single-cycle programs. Such an investment is time-critical while JWST's instruments remain at peak performance, and would maximize synergy with ALMA, Euclid, Roman, Rubin, and future facilities. We therefore urge STScI to open a MCT channel, enabling the bold, ambitious surveys needed to understand how galaxies evolve from first light to the present day.

astro-ph.IM↗

arXiv2026

The stellar-to-halo mass relation of central galaxies across three orders of halo mass

The stellar content of galaxies is tightly connected to the mass and growth of their host dark matter halos. Observational constraints on this relation remain limited, particularly for low-mass groups, leaving uncertainties in how galaxies assemble their stars across halo mass scales. Accurately measuring the brightest central galaxy (BCG) stellar-to-halo mass relation (SHMR) over a wide mass range is therefore crucial for understanding galaxy formation and the role of feedback processes. Here we present the SHMR spanning $M_{\rm halo} \sim 10^{12}$-$10^{15}\,M_\odot$, using halo masses derived from eROSITA eRASS1 X-ray data and BCG stellar masses based on SDSS photometry. By stacking X-ray spectra of optically selected groups, we recover robust average halo gas temperatures for each bin, which are then converted to halo masses via the $M$-$T_X$ relation. We find that the SHMR peaks near $M_{\rm halo} \sim 10^{12}\,M_\odot$, with a declining stellar fraction at higher masses. This trend reflects a combination of processes that reduce the efficiency of stellar mass growth in massive halos, such as AGN feedback, reduced cooling efficiency, and the increasing dominance of ex-situ assembly, while halos continue to grow through mergers and accretion. Our measurements are consistent over the full mass range with previous observational studies, including weak lensing, X-ray analyses of individual clusters, and kinematical and dynamical methods. Comparisons with hydrodynamical simulations show good agreement at low masses but reveal significant discrepancies in the normalization at cluster scales, highlighting the sensitivity of BCG stellar growth to feedback prescriptions and halo assembly history. These results provide the first X-ray-based observational SHMR covering three orders of magnitude in halo mass, establish a robust benchmark for testing galaxy formation models.

astro-ph.GA↗

arXiv2026

PANORAMIC: The Dawn of Massive Quiescent Galaxies I. Number Density and Cosmic Variance from 1000 arcmin$^2$ NIRCam Imaging

We measure the number density and field-to-field variance of massive quiescent galaxies at $z\sim3$ - 8 using the JWST/NIRCam pure-parallel imaging survey PANORAMIC together with archival observations, covering an area of 0.28 deg$^2$ ($\sim1000$ arcmin$^2$) in at least six filters. We identify quiescent galaxy candidates at $z\gtrsim3$ with $M_\ast \gtrsim 10^{10}\,M_\odot$, comprising 101 galaxies in a gold sample of high-confidence candidates and 137 in a more inclusive silver sample. We measure their evolving comoving number density, finding $(1.5$ vs. $3.1)\times10^{-5}\,\mathrm{Mpc}^{-3}$ at $z=3$ - 4 for the gold and silver samples, respectively, and a decline by more than a factor of 20 by $z\sim6$. Comparisons with empirical models and cosmological simulations show that widely used frameworks underpredict the abundance of massive quiescent galaxies at $z\gtrsim4$ by $\gtrsim1$ dex, indicating that current implementations of early star formation, feedback, and quenching do not produce enough early quenched systems. With 34 independent sightlines, we present the first direct empirical measurement of field-to-field variance for quiescent galaxies at $z>3$, finding a high cosmic variance of $σ_{\rm CV}\approx0.7\pm0.3$. This exceeds predictions from abundance-matched mock catalogs, suggesting that early quiescent galaxies are more strongly clustered, and more likely to be found near one another or in more biased regions, than expected in current galaxy-formation models. Any successful model for the emergence of early massive quiescent galaxies must reproduce both their abundance evolution and their imprint on the large-scale distribution.

astro-ph.GA↗

arXiv2026

The Ray Tracing Sampler: Bayesian Sampling of Neural Networks for Everyone

We derive a family of Markov Chain Monte Carlo (MCMC) sampling methods based on following ray paths in a medium where the refractive index $n(x)$ is a function of the desired likelihood $\mathcal{L}(x)$, extending past work on isokinetic sampling. The simplest ray tracing method propagates rays at constant speed through parameter space, leading to orders of magnitude higher resilience to heating for stochastic gradients as compared to both Hamiltonian Monte Carlo (HMC) and Stochastic Gradient HMC, as well as the ability to cross any likelihood barrier, including holes in parameter space. Using the simplest ray tracing method, we sample the posterior distributions of neural network outputs for a variety of different architectures, including a preliminary exploration of the 1.5 billion-parameter GPT-2 (Generative Pre-trained Transformer 2) architecture, all on a single consumer-level GPU. We also show that prior samplers including traditional HMC, the original microcanonical HMC, Metropolis, Gibbs, and even Monte Carlo integration are special cases within a generalized ray tracing framework, which can sample according to an arbitrary weighting function. Public code and documentation for C, JAX, and PyTorch are available at https://bitbucket.org/pbehroozi/ray-tracing-sampler/src

astro-ph.IM↗

arXiv2026

Characterizing the host galaxies and delay times of Ca-rich gap transients vs 91bg-like SNe and normal Type Ia SNe

Calcium-rich gap transients are a faint, fast-evolving class of supernovae that show strong nebular Ca emission lines. Their progenitor systems are uncertain, but they are often associated with old and quiescent host galaxies. In this work, we compare the properties of the hosts of hydrogen-poor Ca-rich gap transients to the hosts of 3 other classes of supernova (SNe): normal Type Ia, 91bg-like, and Type II. We use data from the Zwicky Transient Facility (ZTF) Census of the Local Universe (CLU) experiment to build up our 4 SNe samples and identify the host galaxies. A combination of precomputed host properties from the CLU catalog and those derived from SED fitting are used to characterize each host's stellar mass, star formation rate, and specific star formation rate (sSFR). We find that the hosts of Ca-rich gap transients and 91bg-like SNe occupy a similar parameter space of mass and sSFR, and are more massive and quiescent compared to the hosts of Type Ia and Type II SNe. Additionally, we construct delay time distributions (DTDs) for our 4 samples, finding that Ca-rich gap transients and 91bg-like SNe have the longest peak delay times $\sim 10^4$ Myr, compared to the peak delay times of Type Ia SNe ($\sim 10^3$ Myr) and Type II SNe ($\sim 10$ Myr). The similarity of host environment and DTDs for Ca-rich gap transients and 91bg-like SNe motivates further analysis of the relationship of these two transient classes.

astro-ph.HE↗

arXiv2026

MPI-Rockstar: a Hybrid MPI and OpenMP Parallel Implementation of the Rockstar Halo finder

MPI-Rockstar is a massively parallel halo finder based on the Rockstar phase-space temporal halo finder code, which is one of the most extensively used halo finding codes. Compared to the original code, parallelized by a primitive socket communication library, we parallelized it in a hybrid way using MPI and OpenMP, which is suitable for analysis on the hybrid shared and distributed memory environments of modern supercomputers. This implementation can easily handle the analysis of more than a trillion particles on more than 100,000 parallel processes, enabling the production of a huge dataset for the next generation of cosmological surveys. As new functions to the original Rockstar code, MPI-Rockstar supports HDF5 as an output format and can output additional halo properties such as the inertia tensor.

astro-ph.IM↗

arXiv2026

Galaxies at z > 10: ΛCDM predicts increased Star Formation Efficiency

We show that the rest-frame UV statistics and global properties of galaxies at 7 <= z <= 14 are naturally reproduced within the standard ΛCDM framework when galaxy formation is modeled with UniverseMachine applied to the high-resolution Uchuu N-body simulation. Our model matches the UV luminosity functions over five magnitudes and reproduces the evolution of the UV (and inferred star formation rate) density once internal dust attenuation is included. Comparisons with spectroscopically confirmed JWST/HST galaxies show good agreement with the stellar mass-SFR and stellar mass-UV luminosity relations. In contrast, earlier claims of insufficient stellar masses at z=8 are inconsistent with our model and are likely driven by systematic uncertainties, including AGN contamination, dust attenuation, and the lack of JWST/MIRI constraints. A key prediction is that the star-formation efficiency increases with redshift at fixed halo mass, reaching 2-3 percent of baryons converted into stars by z=10-12. These results demonstrate that current JWST observations of early galaxy populations can be explained within the ΛCDM framework.

astro-ph.GA↗

arXiv2025

DREAMS.II. Galaxy Demographics from Direct Te-Based Metallicities at z~2-10: Tracing the Evolution of the Mass-Metallicity and Fundamental Relations

We present the statistics of line ratios and direct Te-based metallicities from JWST medium-resolution spectra of 292 galaxies at z=2-10, combining DREAMS observations with those of JADES and CEERS. To remove systematics caused by stellar mass (M*) and star formation rate (SFR), we construct stacked spectra binned by redshift within fixed M* and SFR ranges, as well as across the full ranges. We find that the [OII]3727/Hb ratio drops by a factor of five from z~3 to 8 at fixed M* and SFR, in contrast to the nearly constant [OIII]5007/Hb ratio. We derive metallicities via the direct Te method using the [OIII]4363 line, and identify that high-z galaxies lie on the low-metallicity end of the anti-correlation between ionization parameter and metallicity at z~0. Photoionization modeling demonstrates that the redshift evolution, where metallicity decreases and ionization parameter increases, self-consistently explains the observed line ratios. We then examine the mass-metallicity (MZ) and fundamental (MZ-SFR) relations. Including additional galaxies at z~10-12, we find that the MZ relation monotonically decreases from z~3 to 10 at fixed M*, while the MZ-SFR relation shows a significant decline at z>8. Based on the ChemicalUniverseMachine model, this evolutionary trend can be explained by enhanced gas inflow (outflow) by a factor of ~5 (~1.7) at z>8.

astro-ph.GA↗

arXiv2025

One Galaxy Sample to Rule Them All: Halo Occupation Distribution Modeling of DES Year 3 Source Galaxies

For the joint analysis of second-order weak lensing and galaxy clustering statistics, so-called $3{\times}2$ analyses, the selection and characterization of optimal galaxy samples is a major area of research. One promising choice is to use the same galaxy sample as lenses and sources, which reduces the systematics parameter space that describes uncertainties related to galaxy samples. Such a ``lens-equal-source'' analysis significantly improves self-calibration of photo-z systematics leading to improved cosmological constraints. With the aim to enable a lens-equal-source analysis on small scales we investigate the halo-galaxy connection of DES-Y3 source galaxies. We develop a technique to construct mock source galaxy populations by matching COSMOS/UltraVISTA photometry onto {\sc{UniverseMachine}} galaxies. These mocks predict a source halo occupation distribution (HOD) that exhibits significant redshift evolution, non-trivial central incompleteness and galaxy assembly bias. We produce multiple realizations of mock source galaxies drawn from the {\sc{UniverseMachine}} posterior with added uncertainties in measured DES photometry and galaxy shapes. We fit a modified HOD formalism to these realizations to produce priors on the galaxy-halo connection for cosmological analyses. We additionally train an emulator that predicts this HOD to $\sim2\%$ accuracy from redshift $z = 0.1 - 1.3$ that models the dependence of this HOD on 1) observational uncertainties in galaxy size and photometry, and 2) uncertainties in the {\sc{UniverseMachine}} predictions.

astro-ph.CO↗

arXiv2025

Machine Learning the Dark Matter Halo Mass of the Milky Way

Although numerous dynamical techniques have been developed to estimate the total dark matter halo mass of the Milky Way, it remains poorly constrained, with typical systematic uncertainties of 0.3 dex. In this study, we apply a neural network-based approach that achieves high mass precision without several limitations that have affected past approaches; for example, we do not assume dynamical equilibrium, nor do we assume that neighboring galaxies are bound satellites. Additionally, this method works for a broad mass range, including for halos that differ significantly from the Milky Way. Our model relies solely on observable dynamical quantities, such as satellite orbits, distances to larger nearby halos, and the maximum circular velocity of the most massive satellite. In this paper, we measure the halo mass of the Milky Way to be log_10 M_vir / M_Sun = 12.20^{+0.163}_{-0.138}. Future studies in this series will extend this methodology to estimate the dark matter halo mass of M31, and develop new neural networks to infer additional halo properties including concentration, assembly history, and spin axis.

astro-ph.GA↗

arXiv2025

Halo Properties from Observable Measures of Environment: II. Central versus Satellite Classification

A physical understanding of galaxy formation and evolution benefits from an understanding of the connections between galaxies, their host dark matter halos, and their environments. In particular, interactions with more-massive neighbors can leave lasting imprints on both galaxies and their hosts. Distinguishing between populations of galaxies with differing environments and interaction histories is therefore essential for isolating the role of environment in shaping galaxy properties. We present a novel neural-network based method, which takes advantage of observable measures of a galaxy and its environment to recover whether it (1) is a central or a satellite, (2) has experienced an interaction with a more massive neighbor, and (3) is currently orbiting or infalling onto such a neighbor. We find that projected distances to, redshift separations of, and relative stellar masses with respect to a galaxy's 25 nearest neighbors are sufficient to distinguish central from satellite halos in $> 90\%$ of cases, with projection effects accounting for most classification errors. Our method also achieves high accuracy in recovering interaction history and orbital status, though the network struggles to distinguish between splashback and infalling systems in some cases due to the lack of velocity information. With careful treatment of the uncertainties introduced by projection and other observational limitations, this method offers a new avenue for studying the role of environment in galaxy formation and evolution.

astro-ph.GA↗

arXiv2025

Insights into the dependence of galaxy properties on the environment with explainable machine learning models

Galaxies reside within dark matter halos, but their properties are influenced not only by their halo properties but also by the surrounding environment. We construct an interpretable neural network framework to characterize the surrounding environment of galaxies and investigate the extent to which their properties are affected by neighboring galaxies in IllustrisTNG300 data ($z=0$). Our models predict galaxy properties (stellar mass and star formation rate) given dark matter subhalo properties of both host subhalo and of surrounding galaxies, which serve as an explainable, flexible galaxy-halo connection model. We find that prediction accuracy peaks when incorporating only the nearest neighboring galaxy for stellar mass prediction, while star formation rate prediction benefits from information from up to the third-nearest neighbor. We determine that environmental influence follows a clear hierarchical pattern, with the nearest neighbor providing the dominant contribution that diminishes rapidly with additional neighbors. We confirm that central and satellite galaxies, as well as different galaxy categories based on mass and star-forming activity, exhibit distinct environmental dependencies. Environmental dependence for low-mass galaxies ($\log(M_*/M_\odot) < 10$) shows 35-50% environmental contribution compared to just 8-30% for massive centrals, while satellite galaxies experience consistently stronger environmental effects than centrals across all populations. Furthermore, we find that the most significant attribute from neighboring subhalos for predicting target galaxy properties is its distance to the nearest neighboring galaxy. These quantitative results offer guidance for constructing more sophisticated empirical and semi-analytic models of galaxy formation that explicitly include environmental dependence as a function of galaxy type and mass.

astro-ph.GA↗

arXiv2025

Lonely Little Red Dots: Challenges to the AGN-nature of little red dots through their clustering and spectral energy distributions

Observations with the James Webb Space Telescope (JWST) reveal a previously unseen population of compact red objects, known as ``little red dots`` (LRDs). We study a new photometrically selected sample of 124 LRDs in the redshift range $z$ $\sim$ 3 - 10 selected from NIRCam coverage of the CEERS, NEP-TDF, JADES and JEMS surveys. For JADES, the NEP-TDF and CEERS, we compare SED models with and without AGN components and analyse the impact of an AGN component on the goodness of fit using the Bayesian information criterion (BIC). We find that whilst the $χ^{2}$ of the majority of models containing AGN components is improved compared to models without AGN components, we show that the BIC suggests models without AGN are a more appropriate fit to LRD SEDs, especially when MIRI data is available. We also measure LRD clustering in the CEERS field, JADES field, and NEP-TDF, where we compare the spatial distribution of LRDs and galaxies with Kolmogorov-Smirnov tests of equality of distribution. We find that the neighbourhood of LRDs tends to be less dense compared to galaxies at all selections and masses and at similar redshifts. We further measure upper limit estimates for the halo masses of LRDs using abundance matching. Whilst the population of LRDs could be a mixture of several different inherent populations, as a whole it does appear that these systems are mostly hosting compact galaxies or star clusters in formation.

astro-ph.GA↗

arXiv2025

A Promise for the JWST era: Massive black holes directly collapsed from wave dark matter haloes, and Star formation in and around their accretion flows

There are several puzzling phenomena from recent JWST observations, which seem to push the standard ΛCDM cosmology over the edge. All those puzzles can be understood in a coherent way if we assume that first massive black holes (MBHs, the "heavy eggs") formed by directly collapsing from wave CDM haloes, which can be even earlier than the formation of first galaxies (the "chickens"). We elucidate two false obstacles that have been believed to prevent wave CDM haloes from collapsing into black holes (namely "maximum stable mass" and "bosenova") and point out that general-relativistic instabilities (e.g., the non-axisymmetric instability numerically found in spinning scalar-field boson stars) could serve as the concrete mechanisms for direct-collapse black holes (DCBHs) born from wave dark matter. Once the MBHs formed, star formation bursts in and around the accretion flows, characteristic of a special mode in compact regions of high gravitational accelerations.

astro-ph.GA↗

arXiv2025

Using Cumulative Number Densities to Compare Galaxies across Cosmic Time

Comparing galaxies across redshifts at fixed cumulative number density is a popular way to estimate the evolution of specific galaxy populations. This method ignores scatter in mass accretion histories and galaxy-galaxy mergers, which can lead to errors when comparing galaxies over large redshift ranges (Delta z > 1). We use abundance matching in the LCDM paradigm to estimate the median change in number density with redshift and provide a simple fit (+0.16 dex per unit Delta z) for progenitors of z = 0 galaxies. We find that galaxy descendants do not evolve in the same way as galaxy progenitors, largely due to scatter in mass accretion histories. We also provide estimates for the 1-sigma range of number densities corresponding to galaxy progenitors and descendants. Finally, we discuss some limits on number density comparisons, which arise due to difficulties measuring physical quantities (e.g., stellar mass) consistently across redshifts. A public tool to calculate number density evolution for galaxies, as well as approximate halo masses, is available online.

astro-ph.CO↗

arXiv2025

Trinity VII. Predictions for the Observable Correlation Functions of Accreting Black Holes

The quasar correlation function assesses the occurrence of quasar pairs as a function of separation, which is strongly influenced by quasar host halo masses. The empirical Trinity model recently inferred the redshift-dependent relationship between supermassive black hole (SMBH) mass, galaxy mass, and halo mass, using constraints other than correlation functions (e.g., quasar luminosity functions, active galactic nuclei occupation fractions, and SMBH mass-bulge mass relations). Hence, comparing the predicted quasar correlation functions from Trinity to real observations is an important test of Trinity's inferred SMBH -- halo relation. In this work, we use a compilation of observed two-point projected and redshift-space correlation functions from $0 \leq z \leq 3.5$. We find that Trinity accurately predicts quasar correlation functions within observed error bars, although observations do not have much constraining power at lower redshifts due to smaller observable volumes and lower quasar number densities. This finding is consistent with Trinity having the correct placement of quasars within their host galaxies and dark matter halos, without requiring quasar clustering constraints during model fitting. Using Trinity, we also predict the clustering as a function of quasar bolometric luminosity, finding that existing survey uncertainties are too large to show measurable differences ($\lesssim 0.3$ dex change in bias for $10^{42}$ erg s$^{-1}$ compared to $10^{46}$ erg s$^{-1}$ SMBHs across redshifts). This fact arises because most SMBH growth (and hence quasar luminosity) occurs in halos in a similar mass range ($10^{12}-10^{13} M_\odot$).

astro-ph.GA↗

arXiv2025

The SAGA Survey. V. Modeling Satellite Systems around Milky Way-mass Galaxies with Updated UniverseMachine

Environment plays a critical role in shaping the assembly of low-mass galaxies. Here, we use the UniverseMachine (UM) galaxy-halo connection framework and the Data Release 3 of the Satellites Around Galactic Analogs (SAGA) Survey to place dwarf galaxy star formation and quenching into a cosmological context. UM is a data-driven forward model that flexibly parameterizes galaxy star formation rates (SFR) using only halo mass and assembly history. We add a new quenching model to UM, tailored for galaxies with stellar masses $\lesssim 10^9$ solar masses, and constrain the model down to a stellar mass $\gtrsim 10^7$ solar masses using new SAGA observations of 101 satellite systems around Milky Way (MW)-mass hosts and a sample of isolated field galaxies in a similar mass range from the Sloan Digital Sky Survey (SDSS). The new best-fit model, 'UM-SAGA,' reproduces the satellite stellar mass functions, average SFRs, and quenched fractions in SAGA satellites while keeping isolated dwarfs mostly star forming. The enhanced quenching in satellites relative to isolated field galaxies leads the model to maximally rely on halo assembly to explain the observed environmental quenching. Extrapolating the model down to a stellar mass $\sim 10^{6.5}$ solar masses yields a quenched fraction of $\gtrsim$ 30% for isolated field galaxies and $\gtrsim$ 80% for satellites of MW-mass hosts at this stellar mass. This specific prediction can soon be tested by spectroscopic surveys to reveal the relative importance of internal feedback, cessation of mass and gas accretion, satellite-specific gas processes, and reionization for the evolution of faint low-mass galaxies.

astro-ph.GA↗

arXiv2025

ChemicalUniverseMachine I: Uncovering the Cosmic Evolution of Metals in the Galaxy-ISM-CGM Ecosystem

We present an empirical chemical evolution model that explains the distribution of metals in the interstellar medium (ISM) and the circumgalactic medium (CGM) of galaxies based on the UniverseMachine and NeutralUniverseMachine models in the framework of $Λ$CDM structure formation. We parameterize the fractions of outflowing metals returned and mixed into the multi-phase ISM of the star-forming regions ($f_{\rm H2}$) and into the neutral gas regions ($f_{\rm HI}$); metal production, transfer, and dilution are caused by star formation, galaxy mergers, and gas inflow from the inter-galactic medium, respectively, with rates determined by the (Neutral)UniverseMachine models. Using a Markov Chain Monte Carlo algorithm, we explore the posterior distributions of metal return and mixing consistent with observed mass-metallicity relations in HII regions (at $0<z<5$), HI damped Lyman-alpha systems (at $1<z<4$), and the CGM (at $z=0$). We find that the fraction of metals present in the ISM, $f_{\rm H2}+f_{\rm HI}$, increases with halo mass from $\sim20$\% at $10^{10}M_\odot$ to $\sim80$\% at $10^{13}M_\odot$. These fractions increase mildly at higher redshifts, to $\sim30$\% at $10^{10}M_\odot$ and $\sim80$\% at $10^{13}M_\odot$ at $z=5$. Interestingly, there is no significant redshift evolution of $f_{\rm H2}+f_{\rm HI}$ at fixed circular velocity, suggesting that metal distribution between the ISM and CGM is universally determined by the halo potential well depth. CGM metal enrichment is thus slow in high-$z$ halos with deep potential wells. While $f_{\rm H2}$ monotonically increases with halo mass, $f_{\rm HI}$ peaks at $\sim10^{12}-10^{13} M_\odot$, suggesting that reinfall may be inefficient in larger-mass halos.

astro-ph.GA↗