Photometric classification of quasars from DES and photo-$z$ estimation with Machine Learning

2026-06-01T11:20:51Z

This paper presents a comprehensive study of quasar photometric classification and redshift estimation using machine learning techniques. We cross-matched photometric data from the Dark Energy Survey Data Release 2 (DES DR2) with spectroscopic classifications from the Sloan Digital Sky Survey Data Release 16 (SDSS DR16), yielding an initial sample of 168,738 point-like objects. Using a K-Nearest Neighbors (KNN) algorithm with PSF magnitudes in the $g$, $r$, $i$, and $z$ bands, we achieved high-precision quasar/galaxy classification against stellar contaminants, reaching a recall of 0.77 at 0.99 precision. Photometric redshifts were subsequently estimated using a hybrid machine learning approach combining a Boosted Decision Tree from ANNz and a Decision Tree Regressor from scikit-learn. The resulting catalog spans redshifts from $z \approx 0.5$ to $z > 3$, with a distinct population recovered at $z \approx 4$. A stacked outlier classifier was developed to mitigate catastrophic redshift errors. The full photometric redshift sample contains 872,372 objects and remains reliable for cosmological applications at $z \approx 4$. The cleaned catalog contains 675,683 objects and is suitable for large-scale structure studies in the range $0 < z < 3$. This robustly characterized quasar catalog provides a valuable resource for future cosmological investigations.

Observational constraints on 3-forms dark energy

2026-06-01T10:52:35Z

3-forms are natural candidates for describing the late-time accelerated expansion of the Universe, as they can inherently reproduce a positive cosmological constant when lacking an evolving potential. When such a potential is present, a 3-form field may exhibit either quintessence-like or phantom-like behaviour. In this paper, we consider a late-time effective dark energy model described by a 3-form with a Gaussian potential, stable during the dark-energy-dominated era. We constrain this model observationally by performing a Markov Chain Monte Carlo (MCMC) analysis employing a comprehensive cosmological dataset, including Planck PR4 cosmic microwave background (CMB) data, DESI DR1 baryon acoustic oscillation (BAO) measurements, Pantheon+ Type Ia supernovae data, low-$z$ Cepheid calibrators, and DES Y1 large-scale structure observations. We demonstrate that the 3-form model successfully increases the predicted Hubble parameter of CMB and BAO data from $67.89\pm0.36{\rm km/s/Mpc}$ of $Λ$CDM model to $68.29^{+0.56}_{-0.61}{\rm km/s/Mpc}$ by approaching the potential peak at the right time, thus mildly reducing the tension with the late-time observation. Overall, the 3-form field serves as a promising candidate of phantom-like dark energy from both theoretical and observational points of view.

COSMOS: A numerical relativity code specialized for PBH formation

2026-06-01T10:40:31Z

Primordial black holes (PBHs) are black holes generated in the early universe without having gone through stellar evolution. In the standard formation process, PBHs are formed from super-horizon primordial fluctuations with non-linearly large initial amplitude. In order to simulate the non-linear gravitational dynamics of PBH formation, one has to rely on numerical relativity solvers to approximate the solution of the Einstein equations. COSMOS is a C++ package for solving the Einstein equations in 3+1 dimensions, providing simple tools for the simulation of PBH formation. In order to resolve the collapsing region, non-Cartesian scale-up coordinates and a fixed mesh-refinement procedure are implemented. In COSMOS, a massless scalar field and a perfect fluid with a linear equation of state are implemented as matter fields. To achieve a practically acceptable computational speed, OpenMP is used for the parallelization. COSMOS has no other dependencies, which makes for an easier installation.

The role of the galaxy stellar mass function in determining the cosmological distribution of astrophysical transients with applications to fast radio bursts and merging binary black holes

2026-06-01T10:30:50Z

The cosmological distribution and formation rate of compact astrophysical objects such as fast radio bursts (FRBs) are typically assumed to be proportional to a linear combination of cosmological star formation rate and stellar mass. In the literature, a template for star formation rate, which is just a function of redshift, is typically used. In this work, we point out the importance of galaxy stellar mass function which captures the host galaxy information of observed FRBs as well as the redshift evolution of galaxy stellar mass. Using this information and taking the stellar mass distribution of a sample of localized FRBs at face value, we find that FRB formation efficiency per stellar mass may have to be more efficient (by a factor of $\approx 3$) than previously calculated, in order to reproduce the observed volumetric rate of FRBs at $z=0$. We show that cosmological population studies of FRBs have to include host galaxy information along with its redshift evolution in order to obtain unbiased results. This consideration is also applicable to other transients, e.g. gamma-ray bursts and merging binary black hole events. We show that our approach may open up the possibility to distinguish between different scenarios of merging binary black holes formation with a detection of few thousand gravitational wave events.

Tachyonic particle production: quantum 2PI formalism with momentum exchanging collisions

2026-06-01T10:29:45Z

Oscillating spacetime curvature can drive particle production during reheating, whose accurate modeling requires the use of non-perturbative out-of-equilibrium methods. Tachyonic instabilities have previously been studied using 2-Particle Irreducible (2PI) formalism in the Hartree approximation, which however misses important momentum exchanging interactions. We present a self-consistent approximation scheme for reducing the non-local next-to-leading order 2PI equations of motion to local quantum kinetic equations, which can be solved with standard methods. We pay special attention to interactions involving unstable modes during tachyonic instabilities.

Are Primordial Black Holes Truly Fine-Tuned?

2026-06-01T10:28:28Z

Single-field inflationary models which generate primordial black holes through the enhancement of the curvature primordial power at small scales are commonly criticized and frequently dismissed because they require a large amount of fine-tuning in the parameters setting the ultra slow-roll phase. However, the standardly adopted definition of fine-tuning has a clear drawback: the more the primordial black hole abundance is small and cosmologically harmless, the larger the parameter space is fine-tuned. A well-defined indicator of fine-tuning should assign large values to scenarios requiring significant parameter adjustment to reproduce the desired primordial black hole abundance, while yielding values of order unity in cases governed by standard sensitivity. Motivated by such arguments, we use the (modified version of) Wilson's naturalness criterion for quantifying the fine-tuning and naturalness and we show that the primordial black hole models are not technically unnatural.

Probing inflationary features with galaxy ultraviolet luminosity function observables

2026-06-01T10:10:58Z

We use the galaxy ultraviolet luminosity function measurements at $z=4-8$ to constrain modification to standard inflationary power spectrum. These observables are sensitive to the matter power spectrum which itself depends on inflationary initial conditions. We consider a specific model where a bump feature is introduced to the standard power law inflation spectrum. We find that the galaxy luminosity observables can probe such modifications at wavenumbers $0.5\lesssim k \lesssim 50$ Mpc$^{-1}$. We obtain upper limits on the amplitude of bump-like features at the mentioned wavenumbers. We obtain constraints which are similar to previous constraints on these models using measurements of optical depth of reionization. However, the galaxy luminosity functions are a more direct probe for these type of models and, therefore, can complement indirect constraints coming from measurements of IGM properties.

Self-lensing of moving gravitational-wave sources can break the microlensing crossing timescale degeneracy

2026-06-01T10:00:24Z

When a moving gravitational-wave (GW) source travels behind a massive astrophysical object, its signal is gravitationally lensed, showing a waveform distortion similar to a Paczyński curve. We present a first study on the lensing signature of a massive black hole (MBH) on a frequency-dependent GW signal from a moving compact binary merger (CBC) source, focused on ground-based GW detectors. For both light and GW sources in a Keplerian circular orbit around a MBH lens, the self-lensing geometry breaks the microlensing degeneracy in the Einstein radius crossing timescale $t_{\rm E}$. The duration of the curve ($2 t_{\rm E}$) becomes independent on the MBH mass $M_{\rm MBH}$, and provides a direct value of the orbital distance $d_{\rm LS}$ of the source around the MBH. However, $M_{\rm MBH}$ remains unknown. In GW signals, the redshifted mass $M_{{\rm MBH},z}$ can additionally be analytically inferred from the interference pattern, once the modulation period $T$, the GW frequency $f$, and $t_{\rm E}$ are known: $M_{{\rm MBH},z}\simeq 2.5\times 10^6\,M_\odot\,(t_{\rm E}/[100\,{\rm s}])\,(f\,T)^{-1}$. If this lensing signature is not considered, it may be confused with other waveform distortions, especially in the modeling of overlapping CBC signals in next generation ground-based GW detectors. The observation of one of these curves and its associated parameters may help (1) constrain the orbital distance $d_{\rm LS}$ of sources, especially around low-mass MBHs at the center of star clusters and galaxies, (2) additionally estimate the mass $M_{{\rm MBH},z}$ of these MBHs, and (3) infer the orbital inclination of the binary. Simultaneously obtaining $d_{\rm LS}$ and $M_{{\rm MBH},z}$ through self-lensing can help constrain the astrophysical environments where GW signals come from.

Accurate inner stellar density slopes from projected surface densities in galaxies

2026-06-01T09:15:21Z

The inner slope of the three-dimensional stellar density in dwarf galaxies (rho'[0]) is a sensitive probe of possible departures from the collisionless cold dark matter (CDM) paradigm, since cored stellar distributions (rho'[0]=0) cannot easily reside within the cuspy potentials CDM predicts for low-mass systems. Photometry alone offers an observationally inexpensive way to constrain rho'(0), making this approach particularly attractive for the faint galaxies most relevant to dark matter (DM) studies. Inferring volume densities, however, requires deprojecting the observed stellar surface density, Sigma(R), a procedure that is notoriously ambiguous in the presence of noise. To avoid explicit deprojection, we derive an expression (Eq.~[9]]) to obtain rho'(0) directly from the radial derivatives of Sigma(R), assuming spherical symmetry and smooth finite density profiles. All projected profiles are shown to have the same central functional form, independent of the underlying volume density (Eq.~[20]). As a result, the derivatives of Sigma(R) can be extrapolated to the center using constraints from larger radii, which in turn yields rho'(0). As an illustration, we apply the method to six ultra-faint dwarf (UFD) galaxies, finding that all of them have a surface density with the same shape, from which the presence of stellar cores is inferred (rho'[0] simeq 0). The technique also has the ability to diagnose rho'[0]>0, corresponding to galaxies with a central stellar mass deficit potentially linked to black-hole scouring, MONDian dynamics, or deviations from CDM.

Blue-tilted Runnings and the JWST Early Galaxy Tension

2026-06-01T08:46:45Z

The recent James Webb Space Telescope (JWST) observations reported the unexpectedly large abundance of massive galaxies with stellar masses of $\sim 10^{10}~M_{\odot}$ at high redshifts $z \simeq 6.5 - 9$ compared with the prediction of the standard $Λ$CDM model. As a possible solution to the tension, we consider a blue-tilted spectrum of density perturbations with a positive running. We find that, for $α_s \simeq 0.02$ and $β_s \simeq 0.02$, a joint analysis with CMB observations shows that the tension can be resolved at the 1$σ$ confidence level. Such a blue-tilted spectrum is also plausible from the perspective of primordial black hole formation on much smaller scales in the early Universe.

CSST large-scale structure analysis pipeline: IV. Cosmic Voids Identified from Galaxy Group Samples as Probes of the Large-scale Structure

2026-06-01T08:45:59Z

Because groups are directly associated with halos, they allow for considerably simpler theoretical modeling than approaches based on individual galaxies. We therefore propose to use voids identified in galaxy group catalogs, referred to as group-voids, to investigate the cosmic large-scale structure (LSS). Using the reference mock galaxy redshift survey (MGRS) designed for the Chinese Space-station Survey Telescope (CSST), we build two galaxy group catalogs representing ideal and conservative scenarios, derived from galaxy samples with 100\% and roughly 30\% spectroscopic redshift completeness, respectively. We then identify voids in these two mock group catalogs, as well as in the underlying halo catalog, and measure two void statistics, the void size function (VSF) and the void density profile, within five redshift intervals spanning $z=0$ to $1.0$. We compare the statistics obtained from two kinds of voids: those defined by galaxy groups (group-voids) and those defined by dark matter halos (halo-voids). In the void-finding process, we adopt the brightest central galaxy (BCG) as the group center to improve the accuracy of the inferred void centers. Our analysis shows that void statistics derived from group-voids with spectroscopic redshift completeness of at least 40\% can faithfully reproduce the corresponding statistics from halo-voids. Even when the redshift completeness of galaxies falls to as low as 30\%, we can still reliably describe group-voids via halo-voids by incorporating a redshift error term. This indicates that group-voids are a promising tool for probing LSS and offer a valuable complement to standard void studies, which is especially advantageous for emulator-based methods.

Mass and Spin Growth of Very Massive Stars in Star Clusters Potentially Associated with Little Red Dots

2026-06-01T08:19:06Z

Using gravitational $N$-body simulations, we investigate the evolution of mass and spin for very massive stars (VMSs) in dense star clusters, which may be potentially associated with Little Red Dots (LRDs). Our results show that VMS masses can reach $10^3$--$10^4\,M_\odot$, depending on the initial conditions of the host clusters. Notably, the VMS mass increases by up to a factor of three when accounting for the bloated state at the Hayashi track induced by stellar collisions, provided that this state is maintained at accretion rates exceeding $3 \times 10^{-2}\,M_\odot\,{\rm yr}^{-1}$. In all cases, the spin of the VMS, when normalized to the dimensionless black hole (BH) spin parameter, exceeds $10$. While our model may overestimate VMS masses and spins due to the omission of post-main-sequence evolution and the loss of mass and angular momentum during collisions, we nonetheless demonstrate that VMSs formed in dense star clusters can be highly spinning. Such a rapidly spinning VMS is expected to collapse into an intermediate-mass BH surrounded by a massive accretion disk. This BH-disk system could trigger powerful explosions and emit burst gravitational waves, similar to those observed in GW190521 and GW231123, for which the remnant BH masses are estimated to be $\gtrsim 100\,M_\odot$.

Cosmology with the angular cross-correlation of gravitational-wave and galaxy catalogs: forecasts for next-generation interferometers and the Euclid survey

2026-06-01T07:58:35Z

We study the angular power spectrum of gravitational-wave and galaxy catalogs in tomographic redshift and distance bins as a probe of late-time cosmology, focusing specifically on next-generation ground-based interferometers in combination with the Euclid photometric survey. We assess the potential of this technique to constrain the Hubble constant and the matter energy density. Our analysis incorporates realistic gravitational-wave source populations, error modelling calibrated on recent detector designs, and accounts for nuisance parameters. We show that the tomographic angular cross-correlation could determine the Hubble constant to percent or sub-percent precision depending on the binning choice, configuration and operation time of gravitational-wave observatories. This conclusion holds even when marginalising over the unknown tracer biases, primordial power-spectrum parameters and baryon density. In particular, we show that the combination of the galaxy auto-correlation spectra and the cross-correlation of gravitational waves and galaxy surveys can lead to an improvement of up to a factor ${\sim}10$ in constraining power over either of the two probes taken individually. However, this prospect crucially relies on the presence of multiple gravitational-wave interferometers able to yield precise sky localisation. We also discuss the use of a spectroscopic redshift catalog, as well as the detectability of the clustering bias of gravitational-wave sources.

Accelerated inference of microlensed gravitational waves with machine learning

2026-06-01T07:28:08Z

Gravitational waves (GWs) within the LIGO-Virgo-KAGRA sensitivity band can be microlensed by stellar and intermediate-mass black holes, producing a frequency-dependent modulation of the signal amplitude. Microlensing analyses, however, are costly due to the increased dimensionality of the parameter space and waveform computation time. As a proof of concept, we show that the deep-learning-based framework Deep Inference for Gravitational-Wave Observations (DINGO), which employs a simulation-based inference approach to estimate posterior distributions, can perform efficient parameter inference for GW microlensing by an isolated point-mass lens. Using simulated microlensed GW signals, we train a lensed-DINGO network and compare its performance with traditional Bayesian parameter estimation carried out with Bilby. Our framework can be used to rapidly identify microlensed events in large GW catalogs. When the lensed-DINGO network is combined with importance sampling, we find that although sample efficiencies are somewhat reduced compared to the unlensed-DINGO network, owing to the richer structure of microlensed signals, it still achieves $\mathcal{O}(10\times)$ speed-up relative to Bilby. We further show that this framework is useful to efficiently estimate the background Bayes-factor distribution, which is crucial for assessing the significance of candidate lensed events. However, for foreground (lensed) events, the sampling efficiency can sometimes drop when analysed with the unlensed-DINGO network, providing a diagnostic indicator of out-of-distribution data. Our approach can be straightforwardly generalised to more complex and realistic lens models, enabling detailed studies of microlensed GWs.

Reconnection-Driven Injection and Stochastic Reacceleration during Cosmological Magnetogenesis

2026-06-01T06:58:26Z

We investigate whether magnetic reconnection can provide suprathermal proton seed particles during cosmological magnetogenesis prior to nonlinear structure formation. Previous work showed that pressure-anisotropy-driven stochastic acceleration alone is strongly limited by cosmological expansion and Coulomb cooling. Here, we extend this framework by adding a phenomenological reconnection-driven source term to the Fokker--Planck equation for the isotropic ion distribution, with the injection power tied to the magnetic-energy growth rate during magnetogenesis. We find that reconnection can act as a fast injection channel and can produce a visible suprathermal tail. However, the resulting nonthermal energy fraction remains very small, of order $10^{-7}$ in the fiducial model, implying a negligible nonthermal pressure contribution to the pre-structure intergalactic medium. This limitation arises because the extremely high-beta plasma contains only a small magnetic-energy reservoir, even when reconnection itself is locally fast. Using a test-particle shock reacceleration estimate, we further show that the reconnection-produced tail can enhance the suprathermal proton population available for later structure-formation shocks by about an order of magnitude. Nevertheless, the associated hadronic gamma-ray emission from low-density cluster outskirts is expected to remain far below current detectability. We therefore conclude that reconnection during cosmological magnetogenesis is unlikely to dominate the cosmic-ray energy budget directly, but may provide a low-level seed population for subsequent shock acceleration.