Searcharxiv⌕ Search

arXiv subjects

Jens Stücker

Publications and source records attributed to Jens Stücker.

At least 19 recordsLinked to original sources
arXiv2026

Dynamical evolution of dark matter subhaloes in the Milky Way: role of the Galactic disc

Dark matter (DM) subhaloes orbiting inside the Milky Way (MW) are promising targets for DM searches, and reliable predictions for the detectability and spatial distribution of their signals are crucial for probing the nature of DM. Recent work showed that tidal forces from the baryonic components of the MW boost the efficiency of subhalo mass-loss, although the underlying physical processes remain insufficiently understood. This study focuses on clarifying the role of the Galactic disc. By using $N$-body simulations, we examine how the dynamical evolution of subhaloes varies with the inclination angle between their orbits and the Galactic disc. Subhaloes whose orbits are inclined by only a few degrees with respect to the Galactic disc pass through it quickly, which enhances tidal shock heating and leads to a pronounced increase in mass-loss efficiency. In contrast, when a subhalo orbit is exactly coplanar with the Galactic disc, adiabatic shielding suppresses the energy input from tidal shocks, resulting in a lower mass-loss efficiency. Tidal stripping lowers the DM density within subhaloes, thereby attenuating the luminosity of their DM signals. Consequently, we expect that subhaloes located at distances of $\sim 0.3$--$2$\,kpc from the Galactic disc plane emit only weak signals, whereas those remaining embedded in the disc are more promising candidates for indirect DM detection, provided that contamination from baryonic emission sources can be carefully modelled and subtracted. Although their mass-loss histories differ significantly, the structural evolution of subhaloes is still well described by the tidal tracks reported in the literature.

astro-ph.CO↗
arXiv2026

JZ-FMM: GPU-native differentiable N-body simulations with the Fast Multipole Method

N-body simulations are an essential tool for modelling the evolution of gravitating systems. Differentiable N-body simulations allow solving complicated reconstruction problems by connecting observations of evolved systems to simple informative priors on their initial conditions. While so far most reconstruction efforts focus on mildly non-linear large scale dynamics, the internal dynamics of galaxies and clusters should exhibit rich information about their formation. The exploration of reconstruction possibilities in this highly non-linear regime naturally requires highly performant differentiable N-body simulations with small scale forces. Here, we present JZ-FMM, a GPU-native differentiable implementation of the fast multipole method for N-body simulations. We show that the GPU oriented design of tree structure, dual tree traversal and multipole translation operators improves performance over established CPU-native and hybrid N-body codes by notably more than an order of magnitude. Further, we show that gradients can be evaluated efficiently and accurately so that JZ-FMM can be used as an essential building block in future field-level reconstruction efforts. We apply the new code to the reconstruction problem of a tidally stripped satellite to show that deeply non-linear problems can indeed be solved efficiently with simulation gradients, although appropriate care must be taken to navigate the complicated optimization landscape.

astro-ph.CO↗
arXiv2026

Validation of the Hybrid Bias Expansion model for the galaxy bispectrum

The Hybrid Bias Expansion model (also known as Hybrid Effective Field Theory, HEFT) provides a promising way to extend the range of validity of perturbative large-scale structure modelling by replacing perturbative gravitational evolution with the nonlinear displacement field measured from $N$-body simulations. While this approach has already been shown to improve the modelling of the power spectrum, its validity at the bispectrum level has not yet been established. In this work we perform a first systematic real-space validation of the Hybrid bispectrum model using DESI-like LRG and ELG mock catalogues constructed at fixed cosmology on volumes similar to those of DESI's LRG samples. We find that the model remains self-consistent up to $k_{\rm max}^B \simeq 0.25\,h\,{\rm Mpc}^{-1}$, while clear signs of breakdown appear for a similar EFT tree-level bispectrum approach at $k_{\rm max}^B \gtrsim 0.13\,h\,{\rm Mpc}^{-1}$. We also show that adding matter cross-statistics significantly improves the precision of the recovered bias parameters, while a partial third-order extension including only the $δ^3$ operator does not extend the validity range. Finally, we find a strong hierarchy among the bispectrum basis terms when grouped by total bias-operator order, with the lowest-order sectors dominating the total amplitude, which has important implications in emulation strategies.

astro-ph.CO↗
arXiv2026

JZ-Tree: GPU friendly neighbour search and friends-of-friends with dual tree walks in JAX plus CUDA

Algorithms based on spatial tree traversal are widely regarded as among the most efficient and flexible approaches for many problems in CPU-based high-performance computing (HPC). However, directly transferring these algorithms to GPU architectures often yields substantially smaller performance gains than expected in light of the high computational throughput of modern GPUs. The branching nature of tree algorithms leads to thread divergence and irregular memory access patterns -- both of which may severely limit GPU performance. To address these challenges, we propose a Morton (z-order) 'plane-based tree hierarchy' that is specifically designed for GPU architectures. The resulting flattened data layout enables efficient dual-tree traversal with collaborative execution across thread groups, leading to highly coalesced memory access patterns. Based on this framework we present implementations of two important spatial algorithms -- exact $k$-nearest neighbour search and friends-of-friends (FoF) clustering. For both cases, we observe more than an order-of-magnitude performance improvement over the closest competing GPU libraries for large problem sizes ($N \gtrsim 10^7$), together with strong scaling to distributed multi-GPU systems. We provide an open-source implementation, 'JZ-Tree' (JAX z-order tree), which serves as a foundation for efficient GPU implementations of a broad class of tree-based algorithms.

cs.DC↗
arXiv2026

STRAWBERRY: Finding haloes in the gravitational potential

Here, we present a novel algorithm that discriminates between bound and unbound particles by consideration of the gravitational potential from an accelerated reference frame -- also referred to as `the boosted potential'. Particles are considered bound if their energy does not exceed the escape energy of a potential well -- given by the closest saddle-point that connects to a deeper potential minimum. This approach has core benefits over previous approaches, since it does not require any ad-hoc thresholds (such as over-density criteria), it includes the gravitational effect of all particles in the binding criterion (improving over widely used self-potential binding checks) and it only operates with instantaneous information (making it simpler than approaches based on dynamical histories). We show that particles typically become bound between their first peri- and apo-centeric passage and that bound and unbound populations show very distinct characteristics through their distribution in phase space, their density profiles, their virial ratios, and their redshift evolution. Our findings suggest that it is possible to understand haloes as two-component systems, with one component being bound, virialized, of finite extent and evolving slowly in quasi-equilibrium and the other component being unbound, unvirialized and evolving rapidly.

astro-ph.CO↗
arXiv2025

Non-halo structures and their effects on gravitationally lensed galaxies

While the $Λ$CDM model succeeds on large scales, its validity on smaller scales remains uncertain. Recent works suggest that non-halo dark matter structures, such as filaments and walls, could significantly influence gravitational lensing and that the importance of these effects depends on the dark matter model: in warm dark matter scenarios, fewer low-mass objects form and thus their mass is redistributed into the cosmic-web. We investigate these effects on galaxy-galaxy lensing using fragmentation-free Warm Dark Matter (WDM) simulations with particle masses of m$_χ$ = 1 keV and m$_χ$ = 3 keV. Although these cosmological scenarios are already observationally excluded, the fraction of mass falling outside of haloes grows with the thermal velocity of the dark matter particles, which allows for the search for first-order effects. We create mock datasets, based on gravitationally-lensed systems from the BELLS-Gallery, incorporating non-halo contributions from these simulations to study their impact in comparison to mocks where the lens has a smooth mass distribution. Using Bayesian modelling, we find that perturbations from WDM non-halo structures produce an effect on the inferred parameters of the main lens and shift the reconstructed source position. However, these variations are subtle and are effectively absorbed by standard elliptical power-law lens models, making them challenging to distinguish from intrinsic lensing features. Most importantly, non-halo perturbation does not appear as a strong external shear term, which is commonly used in gravitational lensing analyses to represent large-scale perturbations. Our results demonstrate that while non-halo structures can affect the lensing analysis, the overall impact remains indistinguishable from variations of the main lens in colder WDM and CDM scenarios, where non-halo contributions are smaller.

astro-ph.CO↗
arXiv2025

Probabilistic Lagrangian bias estimators and the cumulant bias expansion

The spatial distribution of galaxies is a highly complex phenomenon currently impossible to predict deterministically. However, by using a statistical $\textit{bias}$ relation, it becomes possible to robustly model the average abundance of galaxies as a function of the underlying matter density field. Understanding the properties and parametric description of the bias relation is key to extract cosmological information from future galaxy surveys. Here, we contribute to this topic primarily in two ways: (1) We develop a new set of probabilistic estimators for bias parameters using the moments of the Lagrangian galaxy environment distribution. These estimators include spatial corrections at different orders to measure bias parameters independently of the damping scale. We report robust measurements of a variety of bias parameters for haloes, including the tidal bias and its dependence with spin at a fixed mass. (2) We propose an alternative formulation of the bias expansion in terms of "cumulant bias parameters" that describe the response of the logarithmic galaxy density to large-scale perturbations. We find that cumulant biases of haloes are consistent with zero at orders $n > 2$. This suggests that: (i) previously reported bias relations at order $n > 2$ are an artefact of the entangled basis of the canonical bias expansion; (ii) the convergence of the bias expansion may be improved by phrasing it in terms of cumulants; (iii) the bias function is very well approximated by a Gaussian -- an avenue which we explore in a companion paper.

astro-ph.CO↗
arXiv2024

Gaussian Lagrangian Galaxy Bias

Understanding $\textit{galaxy bias}$ -- that is the statistical relation between matter and galaxies -- is of key importance for extracting cosmological information from galaxy surveys. While the bias function $f$ -- that is the probability of forming galaxy in a region with a given density field -- is usually approximated through a parametric expansion, we show here, that it can also be measured directly from simulations in a non-parameteric way. Our measurements show that the Lagrangian bias function is very close to a Gaussian for halo selections of any mass. Therefore, we newly introduce a Gaussian bias model with several intriguing properties: (1) It predicts only strictly positive probabilities $f > 0$ (unlike expansion models), (2) It has a simple analytic renormalized form and (3) It behaves gracefully in many scenarios where the classical expansion converges poorly. We show that the Gaussian bias model describes the galaxy environment distribution $p(δ| \mathrm{g})$, the scale dependent bias function $f$ and the renormalized bias function $F$ of haloes and galaxies generally equally well or significantly better than a second order expansion with the same number of parameters. We suggest that a Gaussian bias approach may enhance the range of validity of bias schemes where the canonical expansion converges poorly and further, that it may make new applications possible, since it guarantees the positivity of predicted galaxy densities.

astro-ph.CO↗
arXiv2024

Probabilistic Estimators of Lagrangian Shape Biases: Universal Relations and Physical Insights

The intrinsic alignment of galaxies is a key factor in modeling weak-lensing observations and can serve as a valuable signal for both cosmological and astrophysical studies. Modelling this signal requires understanding how galaxy shapes form, and their relations to the large-scale gravitational field -- typically encoded in the value of large-scale shape-bias parameters. In this article we contribute to this topic in three ways: (i) developing new estimators of Lagrangian shape-biases (ii) applying them to measure the shape-biases of dark-matter halos (iii) interpreting these measurements to gain insight on the process of halo-shape formation. We show that our estimators produce results consistent with previous literature, and that they possess advantages with respect to previous methods, namely that the measurement of each bias parameter is completely independent from the others, and that bias parameters can be defined for each individual object. We measure universal relations between shape-bias parameters and peak-significance, $ν$. This relation for the first-order shape-bias parameter is linear at high $ν$, and converges to zero at low $ν$, which we interpret as strong evidence against the proposed scenario according to which galaxy shapes arise due to post-formation interaction with the large-scale tidal-field. We anticipate our estimators to be very useful in analyzing hydrodynamical simulations to extract physical understandings of galaxy shape formation, as well as establishing priors on the values of intrinsic-alignment biases.

astro-ph.CO↗
arXiv2024

Characterizing Structure Formation through Instance Segmentation

Dark matter haloes form from small perturbations to the almost homogeneous density field of the early universe. Although it is known how large these initial perturbations must be to form haloes, it is rather poorly understood how to predict which particles will end up belonging to which halo. However, it is this process that determines the Lagrangian shape of protohaloes and is therefore essential to understand their mass, spin and formation history. Here, we present a machine-learning framework to learn how the protohalo regions of different haloes emerge from the initial density field. This involves one neural network to distinguish semantically which particles become part of any halo and a second neural network that groups these particles by halo membership into different instances. This instance segmentation is done through the Weinberger method, in which the network maps particles into a pseudo-space representation where different instances can be distinguished easily through a simple clustering algorithm. Our model reliably predicts the masses and Lagrangian shapes of haloes object-by-object, as well as summary statistics like the halo-mass function. We find that our model extracts information close to optimal by comparing it to the degree of agreement between two N-body simulations with slight differences in their initial conditions. We publish our model open-source and suggest that it can be used to inform analytical methods of structure formation by studying the effect of systematic manipulations of the initial conditions.

astro-ph.CO↗
arXiv2023

The BACCO simulation project: biased tracers in real space

We present an emulator for the two-point clustering of biased tracers in real space. We construct this emulator using neural networks calibrated with more than $400$ cosmological models in a 8-dimensional cosmological parameter space that includes massive neutrinos an dynamical dark energy. The properties of biased tracers are described via a Lagrangian perturbative bias expansion which is advected to Eulerian space using the displacement field of numerical simulations. The cosmology-dependence is captured thanks to a cosmology-rescaling algorithm. We show that our emulator is capable of describing the power spectrum of galaxy formation simulations for a sample mimicking that of a typical Emission-Line survey at $z \sim 1$ with an accuracy of $1-2\%$ up to nonlinear scales $k \sim 0.7 h \mathrm{Mpc}^{-1}$.

astro-ph.CO↗
arXiv2023

Phase-space simulations of prompt cusps: simulating the formation of the first haloes without artificial fragmentation

The first generation of haloes forms from the collapse of the smallest peaks in the initial density field. $N$-body simulations of this process suggest a prompt formation of a steep power-law cusp, but these calculations are plagued by numerical artifacts which casts some doubt on this result. Here, we develop new simulation methods based on the dark matter phase-space sheet approach and present results which are entirely free of artificial clumps. We find that a cusp with density $ρ\propto r^{-1.5}$ is indeed formed promptly, subsequently accreting a more extended halo and participating in the hierarchical growth of later halo generations. However, our simulations also suggest that the presence of artificial clumps just before peak collapse can significantly shallow the inner profiles of the cusps. We use $N$-body simulations with controlled amounts of small-scale power to place a conservative upper limit on the scales affected by artificial clumps. Finally, we used these results to simulate the collapse of the first generation of peaks of various types and in different cosmologies, finding prompt cusps to form in all cases. We conclude that prompt cusps are a generic feature of the collapse of peaks on the free-streaming scale of the initial density field, and their structure can safely be studied using $N$-body simulations provided care is taken to excise the region potentially affected by artificial clumps.

astro-ph.GA↗
arXiv2023

The effect of stellar encounters on the dark matter annihilation signal from prompt cusps

Prompt cusps are the densest quasi-equilibrium dark matter objects; one forms at the instant of collapse within every isolated peak of the initial cosmological density field. They have power-law density profiles, $ρ\propto r^{-1.5}$ with central phase-space density set by the primordial velocity dispersion of the dark matter. At late times they account for $\sim 1\%$ of the dark matter mass but for $>90\%$ of its annihilation luminosity in all but the densest regions, where they are tidally disrupted. Here we demonstrate that individual stellar encounters, rather than the mean galactic tide, are the dominant disruptors of prompt cusps within galaxies. Their cumulative effect is fully (though stochastically) characterised by an impulsive shock strength $B_* = 2πG\intρ_*({\bf x}(t))\, \mathrm{d}t$ where $ρ_*$, the total mass density in stars, is integrated over a cusp's entire post-formation trajectory. Stellar encounters and mean tides have only a small effect on the halo annihilation luminosity seen by distant observers, but this is not true for the Galactic halo because of the Sun's position. For a 100 GeV WIMP, Earth-mass prompt cusps are predicted, and stellar encounters suppress their mean annihilation luminosity by a factor of two already at 20 kpc, so that their annihilation emission is predicted to appear almost uniform over the sky. The Galactic Center $γ$-ray Excess is thus unaffected by cusps. If it is indeed dark matter annihilation radiation, then prompt cusps in the outer Galactic halo and beyond must account for 20-80% of the observed isotropic $γ$-ray background in the 1 to 10 GeV range.

astro-ph.CO↗
arXiv2023

Tidal Stripping in the Adiabatic Limit

We present a model for the remnants of haloes that have gone through an adiabatic tidal stripping process. We show that this model exactly reproduces the remnant of an NFW halo that is exposed to a slowly increasing isotropic tidal field and approximately for an anisotropic tidal field. The model can be used to predict the asymptotic mass loss limit for orbiting subhaloes, solely as a function of the initial structure of the subhalo and the value of the tidal field at pericentre. Predictions can easily be made for differently concentrated host-haloes with and without baryonic components, which differ most notably in their relation between pericentre radius and tidal field. The model correctly predicts several empirically measured relations such as the `tidal track' and the `orbital frequency relation' that was reported by Errani & Navarro (2021) for the case of an isothermal sphere. Further, we propose applications of the `structure-tide' degeneracy which implies that increasing the concentration of a subhalo has exactly the same impact on tidal stripping as reducing the amplitude of the tidal field. Beyond this, we find that simple relations hold for the bound mass, truncation radius, WIMP annihilation luminosity and tidal ratio of tidally stripped NFW haloes in relation to quantities measured at the radius of maximum circular velocity. Finally, we note that NFW haloes cannot be completely disrupted when exposed adiabatically to tidal fields of arbitrary magnitudes. We provide an open-source implementation of our model and suggest that it can be used to improve predictions of dark matter annihilation.

astro-ph.CO↗
arXiv2022

Shedding light on low-mass subhalo survival and annihilation luminosity with numerical simulations

In this work, we carry out a suite of specially-designed numerical simulations to shed further light on dark matter (DM) subhalo survival at mass scales relevant for gamma-ray DM searches, a topic subject to intense debate nowadays. Specifically, we have developed and employed an improved version of DASH, a GPU $N$-body code, to study the evolution of low-mass subhaloes inside a Milky Way-like halo with unprecedented accuracy, reaching solar-mass and sub-parsec resolution in our simulations. We simulate subhaloes with varying mass, concentration, and orbital properties, and consider the effect of the gravitational potential of the Milky Way galaxy itself. More specifically, we analyze the evolution of both the bound mass fraction and annihilation luminosity of subhaloes, finding that most subhaloes survive until present time, even though in some cases they lose more than 99% of their mass at accretion. Baryons in the host induce a much more severe mass loss, especially when the subhalo orbit is more parallel to the galactic disk. Many of these subhaloes cross the solar galactocentric radius, thus making it easier to detect their annihilation fluxes from Earth. We find subhaloes orbiting a DM-only halo with a pericentre in the solar vicinity to lose 70-90% of their initial annihilation luminosity at redshift zero, which increases up to 99% when baryons are also included in the host. We find a strong relation between subhalo's mass loss and the effective tidal field at pericentre. Indeed, much of the dependence on concentration, orbital parameters, host potential and baryonic components can be explained through this single parameter. In addition to shedding light on the survival of low-mass galactic subhaloes, our results can provide detailed predictions that will aid current and future quests for the nature of DM.

astro-ph.GA↗
arXiv2022

Boundaries of chaos and determinism in the cosmos

According to the standard model of cosmology, the arrangement of matter in the cosmos on scales much larger than galaxies is entirely specified by the initial conditions laid down during inflation. But zooming in by dozens of orders of magnitude to microscopic (and human?) scales, quantum randomness reigns, independent of the initial conditions. Where is the boundary of determinism, and how does that interplay with chaos? Here, we make a first attempt at answering this question in an astronomical context, including currently understood processes. The boundary is a function, at least, of length scale, position, and matter type (dark matter being more simply predictable). In intergalactic voids, the primordial pattern of density fluctuations is largely preserved. But we argue that within galaxies, the conditions are at minimum chaotic, and may even be influenced by non-primordial information, or randomness independent of the initial conditions. Randomness could be supplied by events such as supernovae and jets from active galactic nuclei (AGN) and other accretion disks, which, with the help of chaotic dynamics, could broadcast any possible microscopic randomness to larger scales, eventually throughout a galaxy. This may be generated or amplified by a recently investigated process called spontaneous stochasticity, or effective randomness in turbulent systems arising from arbitrarily small perturbations.

physics.hist-ph↗
arXiv2021

Simulating the complexity of the dark matter sheet II: halo and subhalo mass functions for non-cold dark matter models

We present "sheet+release" simulations that reliably follow the evolution of dark matter structure at and below the dark matter free-streaming scale, where instabilities in traditional N-body simulations create a large population of spurious artificial haloes. Our simulations sample a large range of power-spectrum cutoff functions, parameterized through the half-mode scale $k_{\rm{hm}}$ and a slope parameter $β$. This parameter space can represent many non-cold dark matter models, including thermal relic warm dark matter, sterile-neutrinos, fuzzy dark matter, and a significant fraction of ETHOS models. Combining these simulations with additional N-body simulations, we find the following results. (1) Even after eliminating spurious haloes, the halo mass function in the strongly suppressed regime ($n_{\rm{X}}/n_{\rm{CDM}} < 5\%$) remains uncertain because it depends strongly on the definition of a halo. At these mass scales traditional halo finders primarily identify overdensities that are unbound, highly elongated, dominated by tidal fields, or far from virialized. (2) The regime where the suppression is smaller than a factor of 20 is quite robust to these uncertainties, however, and can be inferred reliably from suitable N-body simulations. (3) Parameterizing the suppression in the halo- and subhalo mass functions through the scales where the suppression reaches $20\%$, $50\%$ and $80\%$, we provide simple formulae which enable predictions for many non-cold dark matter models. (4) The halo mass-concentration relations in our sheet+release simulations agree well with previous results based on N-body simulations. (5) In general, we confirm the validity of previous N-body studies of warm dark matter models, largely eliminating concerns about the effects of artificial haloes.

astro-ph.CO↗
arXiv2021

The Boosted Potential

The global gravitational potential, $ϕ$, is not commonly employed in the analysis of cosmological simulations, since its level sets do not show any clear correspondence to the underlying density field and its persistent structures. Here, we show that the potential becomes a locally meaningful quantity when considered from a boosted frame of reference, defined by subtracting a uniform gradient term $ϕ_{\rm{boost}}(\boldsymbol{x}) = ϕ(\boldsymbol{x}) + \boldsymbol{x} \cdot \boldsymbol{a}_0$ with acceleration $\boldsymbol{a}_0$. We study this "boosted potential" in a variety of scenarios and propose several applications: (1) The boosted potential can be used to define a binding criterion that naturally incorporates the effect of tidal fields. This solves several problems of commonly-used self-potential binding checks: i) it defines a tidal boundary for each halo, ii) it is much less likely to misidentify caustics as haloes (specially in the context of warm dark matter cosmologies), and iii) performs better at identifying virialized regions of haloes -- yielding to the expected value of 2 for the virial ratio. (2) This binding check can be generalized to filaments and other cosmic structures. (3) The boosted potential facilitates the understanding of the disruption of satellite subhaloes. We propose a picture where most mass loss is explained through a lowering of the escape energy through the tidal field. (4) We discuss the possibility of understanding the topology of the potential field in a way that is independent of constant offsets in the first derivative $\boldsymbol{a}_0$. We foresee that this novel perspective on the potential can help to develop more accurate models and improve our understanding of structure formation.

astro-ph.CO↗