arXiv subjects

Richard Stiskalek

Publications and source records attributed to Richard Stiskalek.

35 records · Page 2Linked to original sources

arXiv2026

Constraining cosmological simulations with peculiar velocities: a forward-modeling approach

Numerical simulations are a key tool to decipher the dynamics of gravitation. Yet, they fail to spatially reproduce the Universe we observe, limiting comparison between observations and simulations to a statistical level. This is highly problematic for rare, faint or well studied nearby objects that are observed in a single environment. The computational cost of recovering this environment in random simulations is prohibitive. We present Hamlet-PM, a method that enables the constraining of initial conditions for cosmological simulations so as to produce evolved numerical universes that can be directly compared to observations of the Local Universe: constrained simulations. Our method implements the field-level forward modeling of the early-time density field from sparse and noisy measurements of late-time peculiar velocities. The dynamics are integrated with a particle-mesh gravity solver, thus probing the mildly non-linear regime. The code is applied to the Cosmicflows-4 compilation of peculiar velocities up to z < 0.05 (160 Mpc/h). The constrained ICs a re-simulated with a high precision N-body code. A series of one hundred dark-matter only cosmological constrained simulations with a resolution of 512^3 particles in a 500^3 [Mpc/h]3 box is presented. Special attention is given to twelve prominent nearby galaxy clusters, whose simulated counterparts are matched on criteria of mass and separation. We provide a mass estimate constrained by the dynamical environment for each cluster. Field-level forward modeling of the initial conditions produces highly constrained cosmological simulations. Currently, this method already overtakes in quality the pipeline in use in the peculiar-velocity community, although systematic biases still need to be addressed. Furthermore, improving the model is easy thanks to the inherent flexibility of the Bayesian approach.

astro-ph.CO↗

arXiv2026

The galaxy-environment connection revealed by constrained simulations

The evolution of galaxies is known to be connected to their position within the large-scale structure and their local environmental density. We investigate the relative importance of these using the underlying dark matter density field extracted from the Constrained Simulations in BORG (CSiBORG) suite of constrained cosmological simulations. We define cosmic web environment through both dark matter densities averaged on a scale up to 16 Mpc/h, and through cosmic web location identified by applying DisPerSE to the CSiBORG haloes. We correlate these environmental measures with the properties of observed galaxies in large surveys using optical data (from the NASA-Sloan Atlas) and 21-cm radio data (from ALFALFA). We find statistically significant correlations between environment and colour, neutral hydrogen gas (HI) mass fraction, star formation rate and Sérsic index. Together, these correlations suggest that bluer, star forming, HI rich, and disk-type galaxies tend to reside in lower density areas, further from filaments, while redder, more elliptical galaxies with lower star formation rates tend to be found in higher density areas, closer to filaments. We find analogous trends with the quenching of galaxies, but notably find that the quenching of low mass galaxies has a greater dependence on environment than the quenching of high mass galaxies. We find that the relationship between galaxy properties and the environmental density is stronger than that with distance to filament, suggesting that environmental density has a greater impact on the properties of galaxies than their location within the larger-scale cosmic web.

astro-ph.GA↗

arXiv2026

Testing subhalo abundance matching with galaxy kinematics

The rotation velocities of disc galaxies trace dark matter halo structure, providing direct constraints on the galaxy--halo connection. We construct a Bayesian forward model to connect the dark matter halo population predicted by $Λ$CDM with an observed sample of disc galaxies (SPARC) through their maximum rotation velocities. Our approach combines a subhalo abundance matching scheme (accounting for assembly bias) with a parameterised halo response to galaxy formation. When assuming no correlation between selection in the SPARC survey and halo properties, reproducing the observed velocities requires strong halo expansion, low abundance matching scatter ($<0.15$ dex at $1σ$) and a halo proxy that strongly suppresses the stellar masses in satellite haloes. This is in clear tension with independent clustering constraints. Allowing for SPARC-like galaxies to preferentially populate low $\Vmax$ haloes at fixed virial mass greatly improves the goodness-of-fit and resolves these tensions: the preferred halo response shifts to mild contraction, the abundance matching scatter increases to $\sint = 0.19^{+0.13}_{-0.11}$ dex and the proxy becomes consistent with clustering. However, the inferred selection threshold is extreme, implying that SPARC galaxies occupy the lowest ${\sim}16$ per cent of the $\Vmaxhalo$ distribution at fixed $\Mvir$. Moreover, even with selection, the inferred scatter remains in statistical disagreement with the low-mass clustering constraints, which are most representative of the SPARC galaxies in our sample. Our analysis highlights the advantage of augmenting clustering-based constraints on the galaxy--halo connection with kinematics and suggests a possible tension using current data.

astro-ph.GA↗

arXiv2025

The Velocity Field Olympics: Assessing velocity field reconstructions with direct distance tracers

The peculiar velocity field of the local Universe provides direct insights into its matter distribution and the underlying theory of gravity, and is essential in cosmological analyses for modelling deviations from the Hubble flow. Numerous methods have been developed to reconstruct the density and velocity fields at $z \lesssim 0.05$, typically constrained by redshift-space galaxy positions or by direct distance tracers such as the Tully-Fisher relation, the fundamental plane, or Type Ia supernovae. We introduce a validation framework to evaluate the accuracy of these reconstructions against catalogues of direct distance tracers. Our framework assesses the goodness-of-fit of each reconstruction using Bayesian evidence, residual redshift discrepancies, velocity scaling, and the need for external bulk flows. Applying this framework to a suite of reconstructions -- including those derived from the Bayesian Origin Reconstruction from Galaxies (BORG) algorithm and from linear theory -- we find that the non-linear BORG reconstruction consistently outperforms others. We highlight the utility of such a comparative approach for supernova or gravitational wave cosmological studies, where selecting an optimal peculiar velocity model is essential. Additionally, we present calibrated bulk flow curves predicted by the reconstructions and perform a density--velocity cross-correlation using a linear theory reconstruction to constrain the growth factor, yielding $S_8 = 0.793 \pm 0.035$. The result is in good agreement with both weak lensing and Planck, but is in strong disagreement with some peculiar velocity studies.

astro-ph.CO↗

arXiv2025

No evidence for local $H_0$ anisotropy from Tully--Fisher or supernova distances

Claims of local ($z \lesssim 0.05$) anisotropy in the Hubble constant have been made based on direct distance tracers such as Tully-Fisher galaxies and Type Ia supernovae. We revisit these using the CosmicFlows-4 Tully-Fisher W1 subsample, 2MTF and SFI++ Tully-Fisher catalogues, and the Pantheon+ supernova compilation (all restricted to $z < 0.05$), including a dipole in either the Tully-Fisher zero-point or the standardised supernova absolute magnitude. Our forward-modelling framework jointly calibrates the distance relation, marginalises over distances, and accounts for peculiar velocities using a linear-theory reconstruction. We compare the anisotropic and isotropic model using the Bayesian evidence. In the CosmicFlows-4 sample, we infer a zero-point dipole of amplitude $0.087 \pm 0.019$ mag, or $4.1\pm0.9$ per cent when expressed as a dipole in the Hubble parameter. This is consistent with previous estimates but at higher significance: model comparison yields odds of $877:1$ in favour of including the zero-point dipole. In Pantheon+ we infer zero-point dipole amplitude of $0.049 \pm 0.013$ mag, or $2.3\pm 0.6$ per cent when expressed as a dipole in the Hubble parameter. However, by allowing for a radially varying velocity dipole, we show that the anisotropic zero-point model captures local flow features (or possibly systematics) in the data rather than an actual linearly growing effective bulk flow caused by anisotropy in the zero-point or expansion rate. Crucially, inferring a more general bulk flow curve we find results fully consistent with expectations from the standard cosmological model.

astro-ph.CO↗

arXiv2025

CosmoBench: A Multiscale, Multiview, Multitask Cosmology Benchmark for Geometric Deep Learning

Cosmological simulations provide a wealth of data in the form of point clouds and directed trees. A crucial goal is to extract insights from this data that shed light on the nature and composition of the Universe. In this paper we introduce CosmoBench, a benchmark dataset curated from state-of-the-art cosmological simulations whose runs required more than 41 million core-hours and generated over two petabytes of data. CosmoBench is the largest dataset of its kind: it contains 34 thousand point clouds from simulations of dark matter halos and galaxies at three different length scales, as well as 25 thousand directed trees that record the formation history of halos on two different time scales. The data in CosmoBench can be used for multiple tasks -- to predict cosmological parameters from point clouds and merger trees, to predict the velocities of individual halos and galaxies from their collective positions, and to reconstruct merger trees on finer time scales from those on coarser time scales. We provide several baselines on these tasks, some based on established approaches from cosmological modeling and others rooted in machine learning. For the latter, we study different approaches -- from simple linear models that are minimally constrained by symmetries to much larger and more computationally-demanding models in deep learning, such as graph neural networks. We find that least-squares fits with a handful of invariant features sometimes outperform deep architectures with many more parameters and far longer training time. Still there remains tremendous potential to improve these baselines by combining machine learning and cosmology to fully exploit the data. CosmoBench sets the stage for bridging cosmology and geometric deep learning at scale. We invite the community to push the frontier of scientific discovery by engaging with this dataset, available at https://cosmobench.streamlit.app

cs.LG↗

arXiv2025

Testing the local supervoid solution to the Hubble tension with direct distance tracers

Several observational studies suggest that the local few hundred Mpc around the Local Group are significantly underdense based on source number counts in redshift space across much of the electromagnetic spectrum, particularly in near-infrared galaxy counts. This ``Keenan--Barger--Cowie (KBC) void'', ``Local Hole'', or ``local supervoid'', would have significant ramifications for the Hubble tension by generating outflows that masquerade as an enhanced local expansion rate. We evaluate models for the KBC void capable of resolving the Hubble tension with a background Planck cosmology. We fit these models to direct distances from the Tully--Fisher catalogue of the CosmicFlows-4 compilation using a field-level forward model. Depending on the adopted void density profile, we find the derived velocity fields to prefer a void size of less than 70 Mpc, which is less than 10 per cent of the fiducial size found by Haslbauer et al. based on the KBC luminosity density data. The predicted local Hubble constant is $72.1^{+0.9}_{-0.8}$, $70.4^{+0.4}_{-0.4}$, or $70.2^{+0.5}_{-0.4}$ km/s/Mpc for an initial underdensity profile that is exponential, Gaussian, or Maxwell-Boltzmann, respectively. The latter two ameliorate the Hubble tension to within $3σ$ of the 4-anchor distance ladder approach of Breuval et al. giving $73.2 \pm 0.9$ km/s/Mpc. The exponential profile does achieve consistency with this measurement at just over $1σ$, but it is disfavoured by the Bayesian evidence. The preferred models produce bulk flow curves that disagree with recent estimates from CosmicFlows-4, despite the void models being flexible enough to match such estimates.

astro-ph.CO↗

arXiv2025

The Manticore Project I: a digital twin of our cosmic neighbourhood from Bayesian field-level analysis

We present the first results from the Manticore project, dubbed Manticore-Local, a suite of Bayesian constrained simulations of the nearby Universe, generated by fitting a physical structure formation model to the 2M++ galaxy catalogue using the BORG algorithm. This field-level inference yields physically consistent realizations of cosmic structure, leveraging a nonlinear gravitational solver, a refined galaxy bias model, and physics-informed priors. The Manticore-Local posterior realizations evolve within a parent cosmological volume statistically consistent with LCDM, demonstrated through extensive posterior predictive tests of power spectra, bispectra, initial condition Gaussianity, and the halo mass function. The inferred local supervolume shows no significant deviation from cosmological expectations; notably, we find no evidence for a large local underdensity. Our model identifies high-significance counterparts for fourteen prominent galaxy clusters each within one degree of its observed sky position. Across the posterior ensemble, these counterparts are consistently detected with 2-4 sigma significance, and their reconstructed masses and redshifts agree closely with observational estimates, confirming the inference's spatial and dynamical fidelity. The peculiar velocity field recovered by Manticore-Local achieves the highest Bayesian evidence across five datasets, surpassing state-of-the-art models. Unlike methods yielding only point estimates or using simplified dynamics, Manticore-Local provides a full Bayesian posterior over cosmic structure and evolution, enabling rigorous uncertainty quantification. These results establish Manticore-Local as the most advanced constrained realization suite of the Local Universe to date, offering a robust statistical foundation for future studies of galaxy formation, velocity flows, and environmental dependencies in our cosmic neighbourhood.

astro-ph.CO↗

arXiv2025

Inferring the Ionizing Photon Contributions of High-Redshift Galaxies to Reionization with JWST NIRCam Photometry

JWST is providing constraints on the history of reionization owing to its ability to detect faint galaxies at $z\gg6$. Modeling this history requires understanding both the ionizing photon production rate ($ξ_{\rm ion}$) and the fraction of those photons that escape into the intergalactic medium ($f_{\rm esc}$). Observational estimates of these quantities generally rely on spectroscopy for which large samples with well-defined selection functions are limited. To overcome this challenge, we present and release an implicit likelihood inference pipeline, PHOTONIOn, trained on mock photometry to predict the escaped ionizing luminosity of individual galaxies ($\dot{N}_{\rm ion}$) based on photometric magnitudes and redshifts. We show that PHOTONIOn is able to reliably infer $\dot{N}_{\rm ion}$ from photometry. This is in contrast to traditional SED-fitting approaches which rely on $f_{\rm esc}$ prescriptions that often over-predict $\dot{N}_{\rm ion}$ for LyC-dim galaxies, even when given access to spectroscopic data. We have deployed PHOTONIOn on a sample of 4,559 high-redshift galaxies from the JADES Deep survey, finding gentle redshift evolutions of $\log_{10}(\dot{N}_{\rm ion}) = (0.08\pm0.01)z + (51.60\pm0.06)$ and $\log_{10}(f_{\rm esc}ξ_{\rm ion}) = (0.07\pm0.01)z + (24.12\pm0.07)$. Late-time values for the ionizing photon production rate density are consistent with theoretical models and observations. We measure the evolution of the IGM ionized fraction to find that observed populations of star-forming galaxies are capable of driving reionization in GOODS-S to completion by $z\sim 5.3$ without the need for AGN or other exotic sources, consistent with other studies of the same field. The $20\%$ of UV-brightest galaxies ($M_{\rm UV}<-18.5$) reionize $35\%$ of the survey volume, demonstrating that UV faint LyC emitters are crucial for reionization.

astro-ph.GA↗

arXiv2024

Symmetry in Hyper Suprime-Cam galaxy spin directions

We perform a Bayesian analysis of anisotropy in binary galaxy spin directions in the Hyper-Suprime Cam Data Release 3 catalogue, in response to a recent claim that it exhibits a dipole (Shamir 2024). We find no significant evidence for anisotropy, or for a direction-independent spin probability that differs from 0.5. These results are unchanged allowing for a quadrupole or simply searching for a fixed anisotropy between any two hemispheres, and the Bayes factor indicates decisive evidence for the isotropic model. Our principled method contrasts with the statistic employed by Shamir 2024, which lacks a strong theoretical foundation.

astro-ph.GA↗

arXiv2024

Evaluating the variance of individual halo properties in constrained cosmological simulations

Constrained cosmological simulations play an important role in modelling the local Universe, enabling investigation of the dark matter content of local structures and their formation. We introduce an internal method for quantifying the extent to which the variance of individual halo properties is suppressed by the constraints imposed on the initial conditions. We apply it to the Constrained Simulations in BORG (CSiBORG) suite of $101$ high-resolution realisations across the posterior probability distribution of initial conditions from the Bayesian Origin Reconstruction from Galaxies (BORG) algorithm. The method is based on the overlap of the initial Lagrangian patch of a halo in one simulation with those in another, measuring the degree to which the haloes' particles are initially coincident. This addresses the extent to which the imposed large-scale structure constraints reduce the variance of individual halo properties. We find consistent reconstructions of $M\gtrsim10^{14}~M_\odot / h$ haloes, indicating that the constraints from the BORG algorithm are sufficient to pin down the masses, positions, and peculiar velocities of clusters to high precision, though we do not assess how well they reproduce observations of the local Universe. The effect of the constraints tapers off towards lower mass, and the halo spins and concentrations are largely unconstrained at all masses. We document the advantages of evaluating halo consistency in the initial conditions and describe how the method may be used to quantify our knowledge of the halo field given galaxy survey data analysed through the lens of probabilistic inference machines such as BORG.

astro-ph.CO↗

arXiv2024

Probing general relativistic spin-orbit coupling with gravitational waves from hierarchical triple systems

Wave packets propagating in inhomogeneous media experience a coupling between internal and external degrees of freedom and, as a consequence, follow spin-dependent trajectories. These phenomena, well known in optics and condensed matter physics, are referred to as spin Hall effects. Similarly, the gravitational spin Hall effect is expected to affect the propagation of gravitational waves on curved spacetimes. In this general-relativistic setup, the curvature of spacetime acts as impurities in a semiconductor or inhomogeneities in an optical medium, leading to a frequency- and polarization-dependent propagation of wave packets. In this letter, we study this effect for strong-field lensed gravitational waves generated in hierarchical triple black hole systems in which a stellar-mass binary merges near a more massive black hole. We calculate how the gravitational spin Hall effect modifies the gravitational waveforms and show its potential for experimental observation. If detected, these effects will have profound implications for astrophysics and tests of general relativity.

gr-qc↗

arXiv2024

Frequency- and polarization-dependent lensing of gravitational waves in strong gravitational fields

The propagation of gravitational waves can be described in terms of null geodesics by using the geometrical optics approximation. However, at large but finite frequencies the propagation is affected by the spin-orbit coupling corrections to geometrical optics, known as the gravitational spin Hall effect. Consequently, gravitational waves follow slightly different frequency- and polarization-dependent trajectories, leading to dispersive and birefringent phenomena. We study the potential for detecting the gravitational spin Hall effect in hierarchical triple black hole systems, consisting of an emitting binary orbiting a more massive body acting as a gravitational lens. We calculate the difference in time of arrival with respect to the geodesic propagation and find that it follows a simple power-law dependence on frequency with a fixed exponent. We calculate the gravitational spin Hall-corrected waveform and its mismatch with respect to the original waveform. The waveform carries a measurable imprint of the strong gravitational field if the source, lens, and observer are sufficiently aligned, or for generic observers if the source is close enough to the lens. We present constraints on dispersive time delays from GWTC-3, translated from limits on Lorentz invariance violation. Finally, we address the sensitivity of current and future ground detectors to dispersive lensing. Our results demonstrate that the gravitational spin Hall effect can be detected, providing a novel probe of general relativity and the environments of compact binary systems.

gr-qc↗

arXiv2023

On the fundamentality of the radial acceleration relation for late-type galaxy dynamics

Galaxies have been observed to exhibit a level of simplicity unexpected in the complex galaxy formation scenario posited by standard cosmology. This is particularly apparent in their dynamics, where scaling relations display much regularity and little intrinsic scatter. However, the parameters responsible for this simplicity have not been identified. Using the Spitzer Photometry & Accurate Rotation Curves galaxy catalogue, we argue that the radial acceleration relation (RAR) between galaxies' baryonic and total dynamical accelerations is the fundamental $1$-dimensional correlation governing the radial (in-disk) dynamics of late-type galaxies. In particular, we show that the RAR cannot be tightened by the inclusion of any other available galaxy property, that it is the strongest projection of galaxies' radial dynamical parameter space, and that all other statistical radial dynamical correlations stem from the RAR plus the non-dynamical correlations present in our sample. We further provide evidence that the RAR's fundamentality is unique in that the second most significant dynamical relation does not possess any of these features. Our analysis reveals the root cause of the correlations present in galaxies' radial dynamics: they are nothing but facets of the RAR. These results have important ramifications for galaxy formation theory because they imply that to explain statistically late-type galaxy dynamics within the disk it is necessary and sufficient to explain the RAR and lack of any significant, partially independent correlation. While simple in some modified dynamics models, this poses a challenge to standard cosmology.

astro-ph.GA↗

arXiv2022

The scatter in the galaxy-halo connection: a machine learning analysis

We apply machine learning, a powerful method for uncovering complex correlations in high-dimensional data, to the galaxy-halo connection of cosmological hydrodynamical simulations. The mapping between galaxy and halo variables is stochastic in the absence of perfect information, but conventional machine learning models are deterministic and hence cannot capture its intrinsic scatter. To overcome this limitation, we design an ensemble of neural networks with a Gaussian loss function that predict probability distributions, allowing us to model statistical uncertainties in the galaxy-halo connection as well as its best-fit trends. We extract a number of galaxy and halo variables from the Horizon-AGN and IllustrisTNG100-1 simulations and quantify the extent to which knowledge of some subset of one enables prediction of the other. This allows us to identify the key features of the galaxy-halo connection and investigate the origin of its scatter in various projections. We find that while halo properties beyond mass account for up to 50 per cent of the scatter in the halo-to-stellar mass relation, the prediction of stellar half-mass radius or total gas mass is not substantially improved by adding further halo properties. We also use these results to investigate semi-analytic models for galaxy size in the two simulations, finding that assumptions relating galaxy size to halo size or spin are not successful.

astro-ph.GA↗

arXiv2021

The dependence of subhalo abundance matching on galaxy photometry and selection criteria

Subhalo abundance matching (SHAM) is a popular technique for assigning galaxy mass or luminosity to haloes produced in N-body simulations. The method works by matching the cumulative number functions of the galaxy and halo properties, and is therefore sensitive both to the precise definitions of those properties and to the selection criteria used to define the samples. Further dependence follows when SHAM parameters are calibrated with galaxy clustering, which is known to depend strongly on the manner in which galaxies are selected. In this paper we introduce a new parametrisation for SHAM and derive the best-fit SHAM parameters as a function of various properties of the selection of the galaxy sample and of the photometric definition, including Sérsic vs Petrosian magnitudes, stellar masses vs r-band magnitudes and optical (SDSS) vs HI (ALFALFA) selection. In each case we calculate the models' goodness-of-fit to measurements of the projected two-point galaxy correlation function. In the optically-selected samples we find strong evidence that the scatter in the galaxy-halo connection increases towards the faint end, and that AM performs better with luminosity than stellar mass. The SHAM parameters of optically- and HI-selected galaxies are mutually exclusive, with the latter suggesting the importance of properties beyond halo mass. We provide best-fit parameters for the SHAM galaxy-halo connection as a function of each of our input choices, extending the domain of validity of the model while reducing potential systematic error in its use.

astro-ph.GA↗

arXiv2020

Are stellar--mass binary black hole mergers isotropically distributed?

The Advanced LIGO and Advanced Virgo gravitational wave detectors have detected a population of binary black hole mergers in their first two observing runs. For each of these events we have been able to associate a potential sky location region represented as a probability distribution on the sky. Thus, at this point we may begin to ask the question of whether this distribution agrees with the isotropic model of the Universe, or if there is any evidence of anisotropy. We perform Bayesian model selection between an isotropic and a simple anisotropic model, taking into account the anisotropic selection function caused by the underlying antenna patterns and sensitivity of the interferometers over the sidereal day. We find an inconclusive Bayes factor of $1.3:1$, suggesting that the data from the first two observing runs is insufficient to pick a preferred model. However, the first detections were mostly poorly localised in the sky (before the Advanced Virgo joined the network), spanning large portions of the sky and hampering detection of potential anisotropy. It will be appropriate to repeat this analysis with events from the recent third LIGO observational run and a more sophisticated cosmological model.

astro-ph.HE↗