Morphology of Radio Sources in Representation Space

2026-09-09T15:51:05Z

Understanding radio source morphologies and their classification remains challenging. We previously developed a deep clustering method based on self-supervised learning to classify a subsample of radio sources from the LOFAR Two-meter Sky Survey DR2. This yielded a labelled subset used to fine-tune an ensemble of classifiers. We aim to identify rare morphological classes in a subsample of LoTSS-DR3, which contains > 13 million sources, beyond the 12 classes previously recognised in the DR2 sample. We further aim to characterise the resulting class distribution. We applied the classifier ensemble to derive class probabilities and representations for our samples. Among DR3 sources with low class probabilities, we searched for new clusters in representation space. Moreover, we use these clusters as centroids to classify the low-probability subset. We found that 88% of the sources fall into clearly separated clusters in representation space, coinciding with the DR2 clusters. Within the remaining sources, we identified representatives of six additional morphological classes, including rare morphologies like winged and FR-ambiguous sources. The number of sources in the resulting 18 classes exhibit a strongly skewed distribution, with four dominant classes constituting 69% of the DR3 sample, while rare morphologies account for only a small fraction. The identification of additional morphological classes shows the possibility of open-set recognition with representation learning in an unexplored dataset. Unlike anomaly detection, which flags individual sources as uncommon, this approach identifies representatives of novel morphological classes, providing a framework for discovering novel source populations. The highly skewed class distribution poses a fundamental challenge for constructing balanced training datasets and highlights the need for open-set approaches in future observations.

Collisionless whistler heat-flux instability in ultra-high-$β$ plasmas

2026-09-09T15:50:28Z

Kinetic instabilities, notably the whistler heat-flux instability (WHFI), are known to suppress thermal transport significantly in the moderate- to high-$β$ plasmas relevant to many astrophysical systems. This paper explores WHFI-regulated heat transport in a new regime: ultra-high-$β$ plasmas with $β_{e} \gtrsim L_{\mathrm{T}}/ρ_e$. Extrapolating previous theories of the WHFI to ultra-high-$β$ plasmas, we propose that the magnetic energy in unstable whistler fluctuations becomes comparable to that of the background magnetic field at saturation. We corroborate this hypothesis using 1D3V and 2D3V kinetic simulations using the particle-in-cell code OSIRIS. We find that, in ultra-high-$β$ plasmas, the heat flux is localised and no longer regulated primarily by resonant pitch-angle scattering of electrons; instead, thermal energy is transported predominantly by advection at the whistler phase velocity. Heat-flux suppression is observed in 1D3V and 2D3V simulations; however, we show that the saturation of the WHFI and the regulation of heat flux are sensitive to dimensionality in the ultra-high-$β$ regime. The amplitude and phase velocity of the heat-flux-regulating whistler waves scale differently with $β_e$, yielding parallel heat fluxes, normalised to the free-streaming value, of $q_{e\parallel} / q_\mathrm{fs} \approx 4.7 β_{e}^{-1}$ and $q_{e\parallel} / q_\mathrm{fs} \approx 0.3 β_{e}^{-1/2}$ in 2D3V and 1D3V simulations, respectively. We perform 2D3V simulations with background magnetic fields inclined to the temperature gradient, showing perpendicular heat transport remains negligible. We develop a heuristic theory from kinetic equations that explains these phenomena. Our work extends our understanding of how the WHFI modifies thermal transport to regimes applicable to high-energy-density physics and the reionised intergalactic medium.

Extending the Little Red Dot population at intermediate redshift with VIPERS

2026-09-09T15:24:39Z

Little red dots (LRDs) are a recently identified population of compact, red sources characterised by a distinctive V-shaped UV-to-optical continuum and broad emission lines. Their physical structure and evolution remain uncertain, while their properties at intermediate redshifts are largely unexplored. We identify and characterise LRD-like sources at 0.50.5, bridging local and high-redshift studies. Their abundance, compact morphology, environmental properties, and multi-wavelength emission provide new constraints on the evolution and physical origin of this population.

From stardust to interstellar grain growth in the first galaxies: a cosmological transition in dust evolution near z ~ 8.9

2026-09-09T15:07:30Z

When and how did dust begin to shape galaxies? Motivated by the identification of an apparent redshift break in galaxy dust masses, suggesting substantially lower dust masses at $z \gtrsim 9$, we investigate dust enrichment during the first billion years of cosmic history. We aim to determine whether the observed evolution marks a transition in the dominant dust-production mechanism and to identify the physical conditions under which such a transition is expected to occur. Using JWST, ALMA, and NOEMA observations, we measure ultraviolet dust attenuation and dust masses. We apply a censored change-point analysis and compare the observations with dust evolution modelling. The analysis identifies a preferred transition near redshift z ~ 8.9, corresponding to about 570 Myr after the Big Bang. The evidence for a break is strongest in dust mass and dust-to-stellar mass ratio, mostly estimated from JWST NIRSpec spectrophotometric fitting but also partly from sub-mm data. The ultraviolet attenuation measurements are consistent with a transition at the same epoch but do not independently require one. The models are consistent with the onset of efficient interstellar grain growth above a characteristic metallicity. We interpret the transition near z ~ 8.9 as the emergence of grain-growth-dominated dust evolution from an earlier regime dominated by supernova-produced grains. Population III enrichment can modify the earliest chemical-enrichment history but leaves the timing of the dust transition nearly unchanged and is not required for its emergence.

Associating binary black holes with galactic centres using lensed gravitational waves

2026-09-09T15:02:59Z

Binary black holes (BBHs) have been theorised to form through various different channels, but current LIGO-Virgo-KAGRA (LVK) sky locations make the direct study of formation channels a challenge. Association of strongly lensed BBHs to their hosts has been shown to be possible with current detector networks at upgraded sensitivity and survey telescopes like Euclid, with localisations that can reach sub-galactic scales in the source plane. With Hubble Space Telescope-like high-resolution imaging, we report that strong-lensing localisation of BBHs allows us to confidently associate the BBH with specific structures within their host galaxies, specifically, galactic centres (GCs). We demonstrate that a BBH localised to the GC of the galaxy can reject light-tracing BBH populations ($p < 5\%$) in a single observation, and in most cases demonstrate high confidence rejection ($p < 1\%$), providing a conservative basis to exclude light-tracing spatial distributions associated with field-like or non-central formation channels. Such association therefore directly supports formation channels confined to GCs, such as active galactic nuclei and nuclear stellar clusters.

Cosmic-ray electron propagation in the peculiar barred spiral galaxy NGC 2442

2026-09-09T14:54:22Z

Face-on spiral galaxies offer a favorable geometry for studying magnetic-field structures and cosmic-ray (CR) propagation because projection effects and structural overlap are reduced. We investigate cosmic-ray electron (CRE) transport in the nearby face-on spiral galaxy NGC 2442 and assess how its environment and magnetic-field structure influence propagation. We combine radio continuum (RC) observations from ASKAP at 943 MHz, MeerKAT at 1.28 and 1.7 GHz, and ATCA at 5 GHz with optical H$α$ and infrared data, and compare them with 2D CRE transport simulations. NGC 2442 has a steep integrated RC spectrum, with $α=-0.96\pm0.04$ for the total emission and $α_{\rm nt}=-1.21\pm0.04$ for the synchrotron emission over 408 MHz-5 GHz. A break near 1 GHz indicates substantial radiative aging. Under equipartition, we derive a mean magnetic-field strength of $10.8\,μ{\rm G}$. RC-SFR smoothing gives effective CRE propagation lengths of $\sim0.65$-$0.89$ kpc at 943-1700 MHz and $\sim0.44$ kpc at 5 GHz, corresponding to diffusion coefficients of order $10^{28}\,{\rm cm^2\,s^{-1}}$. We identify a steep-spectrum synchrotron ``island'' in the southeast, with $α\sim-1.09$ and no clear H$α$, infrared, FUV, or NUV counterpart, indicating that CREs are unlikely to be injected in situ. Our 2D CRPropa simulations show that anisotropic diffusion along ordered magnetic fields enables CREs to reach the island more efficiently than isotropic diffusion. NGC 2442 therefore shows that environmental disturbances and ordered magnetic fields can strongly regulate CRE propagation in disturbed spiral galaxies.

The solar Galactocentric distance and local kinematic parameters from Gaia DR3 using Bottlinger's equations

2026-09-09T14:43:30Z

The solar Galactocentric distance, $R_0$, together with the local kinematic parameters of the Milky Way, provides fundamental constraints on the Galactic structure and rotation. Their accurate determination has become possible with the hight astrometric precision of Gaia DR3 catalogue. We determine the Galactocentric distance, the Solar peculiar motion components $(u_0,v_0,w_0)$, and the local angular velocity of Galactic rotation, $ω_0$, using a self-consistent approach based on Bottlinger's equations within a nonlinear least-squares framework. Bottlinger's equations were applied to approximately 13 million luminous giant stars $(M_G<4)$ selected from Gaia DR3. The analysis was carried out for two subsamples: a thin disk dominated sample (|z|<0.5 kpc) and a mixed thin and thick disk sample (|z|<1.0 kpc). The kinematic parameters were estimated using a nonlinear generalized least-squares method. We obtained $R_0=8.165\pm0.024$kpc and a local angular velocity of Galactic rotation $ω_0=28.477\pm0.002$ km/s/kpc for the thin disk dominated sample. The Solar peculiar motion components are $(u_0,v_0,w_0)=(9.261\pm0.006,16.593\pm0.008,7.686\pm0.008)$km/s. Including mixed thin and thick disk stars resulted in a systematic shift in $R_0$, yielding $R_0=8.333\pm0.022$kpc, while the Solar peculiar motion and local angular velocity remain consistent within uncertainties. The Galactic kinematic parameters derived are consistent with recent independent determinations. The Solar peculiar motion parameters and the local angular velocity remain stable for different vertical selections, the recovered Galactocentric distance exhibits a systematic dependence on the adopted stellar sample, indicating that the vertical structure of the tracer population should be taken into account when determining $R_0$.

The internal kinematics and chemistry of 20 Milky Way strings

2026-09-09T13:24:32Z

Context: the recent discovery of filamentary stellar structures in the Milky Way disk raises the question of their formation in the context of the Galaxy evolution. Aims: in this work, we aim to kinematically and chemically characterize several filamentary stellar structures by looking for clues about their origin and formation mechanism. Methods: recent works using machine-learning techniques and $Gaia$ data have identified many previously unknown Galactic strings. We cross-matched such data with the kinematics derived from the most recent Gaia DR3 catalog as well as chemical data provided by the Galactic Archaeology with HERMES (GALAH) survey. Results: we found that most strings contain one or more open clusters and are kinematically hotter than open clusters (with internal velocity dispersions in the 1--7 km/s range) but cooler than field stars. 18 objects appear chemically homogeneous ($σ_{[Fe/H]}$ <= 0.1 dex), while two strings have a broad [Fe/H] distribution similar to the field stars but exhibit a different mean metallicity, which may identify them as migrating stellar structures. For seven objects the GALAH observations focus on the embedded open clusters; therefore, the chemical abundances are not representative of the whole structure. An object shows clues of chemical gradients along its extension in Galactic longitude. Conclusions: these objects may be evaporating or disrupted open clusters in which the chemical homogeneity is preserved in an unbound and expanding cloud of stars or co-moving stars belonging to a former and now disrupting star formation hub. In a few cases they appear to be moving groups with no chemical homogeneity, likely formed by the dynamical action of the Galactic bar or spiral arms. More data are needed to confirm such hypotheses.

Enabling New Discoveries with Machine Learning

2026-09-09T12:44:04Z

The next generation of telescopes such as the Square Kilometre Array and the Vera C. Rubin Observatory will produce enormous quantities of data, too large for traditional analysis techniques. Machine learning has proven invaluable in handling massive data volumes and automating tasks traditionally done by human scientists. The question is, can machine learning also automate scientific discovery? Recent advances in the field have made this feasible but only when combined with expert human input through active learning. Astronomaly is a publicly available framework for active anomaly detection in astronomy. It learns to make recommendations of interesting sources to a user, allowing scientists to quickly uncover rare and new classes of objects in large datasets. In this paper, we review some of the discoveries made using Astronomaly, discuss the latest deep learning-based approaches and explore the critical role of the cyber-human interface in unveiling cosmic mysteries hidden in massive astronomical datasets.

The quenching time and timescale distribution of z~2 quiescent galaxies from precise colour distribution analysis

2026-09-09T12:38:48Z

Understanding when and how galaxies quench their star formation is crucial for understanding the dominant physical processes at play. The spectral energy distribution (SED) of galaxies encodes significant information on their past histories: the relative importance of different physical processes influences the observed distribution of SED shapes in the galaxy population. We use a simulation based inference (SBI) approach to directly constrain the distribution of formation times, quenching times and quenching timescales within the massive galaxy population at z >~ 2 from their broad band photometric colour distribution at 1.710.3. We measure a quenched galaxy fraction of 0.24+/-0.02, with the number density of quenched galaxies rising rapidly 2.5Gyr after the Big Bang (z<~2.6). Galaxies must quench rapidly to achieve the precise bimodal colour distribution: defining the quenching timescale as the time from peak star formation rate (SFR_peak) -> 0.5xSFR_peak, the quenching timescale distribution has a mode at 97_{-25}^{+31}Myr, a median of 182+/-16Myr and a tail to ~700Myr. To achieve full quiescence takes a median time of ~400Myr. Comparing to direct number density measurements of quenched galaxies at z>2 the combination of recent and rapid quenching inferred from the fossil record suggests a substantial rejuvenation and/or merger rate for quenched galaxies observed directly at z>3.5.

The JWST Emission Line Survey (JELS): Multi-wavelength properties of Paschen line-emitters at Cosmic Noon

2026-09-09T12:22:08Z

Context: Narrowband JWST/NIRCam selection isolates star-forming galaxies in well-defined redshift slices, while ALMA and archival far-IR/submm data reveal obscured growth. We combine these to measure Paschen-line SFRs, infer nebular extinction from EAZY Av, and probe FIR-to-submm emission via weighted-median stacking, to constrain galaxy growth at cosmic noon. Aims: We characterize stellar, nebular, and FIR properties of Paschen-selected galaxies in COSMOS, in two ensembles at z~1.50 (Pa-alpha) and z~2.65 (Pa-beta), quantifying short- vs long-timescale SF and dust attenuation, and constraining obscured SFR and gas masses. Methods: We select robust narrowband-excess samples from the JELS F466N/F470N mosaic (Pa-alpha: 77, Pa-beta: 31 sources), remove AGN via X-ray/radio matching, and assemble multiwavelength photometry (JWST, HST, Spitzer, Herschel, SCUBA-2, AzTEC, ALMA). Photo-z/stellar properties come from EAZY/STARDUST; line SFRs use recent Paschen calibrations, with nebular extinction from EAZY Av. Results: EAZY fits show stellar masses log(Mstar/Msun)~7.4-10.3 and modest SFRs (0.1-27 Msun/yr). Line SFRs are 0.1-22.0 and 0.7-47.9 Msun/yr for Pa-alpha/beta. Most sources show short-term bursts (line SFRs exceed SED SFRs), weakly decreasing with mass. Nebular attenuation correlates with mass and SFR. Stacking gives no >3sigma detections beyond MIRI. The weighted-median SED gives SFR_UV,med = 0.48+-0.11 and 3.92+-0.47 Msun/yr for Pa-alpha/beta. ALMA Band 4 gives the tightest constraints: median 3-sigma limits SFR_IR<53 (Pa-alpha) and <44 Msun/yr (Pa-beta); gas mass <10^11.2 (Pa-alpha) and <10^10.6 Msun (Pa-beta). Conclusions: Paschen-selected samples show widespread short-timescale star formation, but FIR/sub-mm data give only upper limits on obscured SFR/ISM mass. Deeper spectroscopy and ALMA continuum data are needed to constrain obscured SF and gas reservoirs.

Search for refractive substructure in active galactic nuclei using ground-based VLBI observations

2026-09-09T12:15:50Z

We present the results of the first systematic search for refractive scattering substructure in active galactic nuclei (AGN) using ground-based very long baseline interferometry (VLBI) data. We selected 1074 AGNs showing a steep frequency dependence of their angular sizes and compared the observed long-baseline visibility amplitudes with the predictions of the refractive substructure model for scatter-broadened images. Nine sources showed consistency with the expected level of refractive noise, including the quasar TXS 2005+403 (J2007+4029), which confirms previously reported results for this source. Remarkably, the majority of these AGNs are located in the Cygnus region -- one of the key ``hotspots'' of strong Galactic scattering. For five selected AGNs, we identified at least two epochs with sufficiently high data quality showing long-baseline detections, suggesting the presence of scattering substructure. We showed that refractive scattering substructure in AGN is a rare effect, requiring high sensitivity, sufficient source brightness, and lines of sight with particularly strong turbulence in the interstellar medium. Nevertheless, AGNs can serve as powerful probes of the turbulent interstellar medium. We discuss strategies for improving the sensitivity of VLBI observations and identify the most promising sources for reliable detection and investigation of refractive substructure in future observations.

eROSITA cosmology with galaxy groups: Hot gas budget out to the virial radius

2026-09-09T11:25:51Z

Non-gravitational processes that expel hot gas beyond the virial regions of groups and clusters of galaxies, known collectively as baryonic feedback, play a key role in reshaping the matter distribution of the Universe on Mpc scales. We use eROSITA observations of a complete sample of 25 galaxy groups selected from the first public release of the eROSITA-DE data (eRASS1) and identified with the Two Micron Redshift Survey optical group catalogue (2MRS). We extract and fit surface brightness (SBx) profiles and present hot gas mass and hot gas fraction profiles out to $R_{200}$. We perform a Bayesian analysis of $M_{\mathrm{gas}}-M_{\mathrm{tot}}$, $L_{\mathrm{X}}-M_{\mathrm{tot}}$, and $L_{\mathrm{X}}-M_{\mathrm{gas}}$ relations, taking into account the aperture covariance effects. At $R_{500}$, we report uniformly flat SBx profiles with a mean $β$ parameter of $0.38 \pm0.04$, steepening to $β= 0.76\pm0.19$ beyond $R_{500}$. We measure a sub-cosmic hot gas fraction at the median mass of our sample $M_{500} = 2.54\times10^{13}M_{\odot}$ of $ f_{\mathrm{gas,500}} = 4.32\pm0.42\%$. Similarly, at $R_{200}$ and the median mass $M_{\mathrm{ 200}} = 3.69\times10^{13}M_{\odot}$, we obtain $f_{\mathrm{gas,200}}=5.78\pm0.69\%$. Our $f_{\mathrm{gas}}-M_{\mathrm{tot}}$ and $L_{\mathrm{X}}-M_{\mathrm{tot}}$ relations show significant deviations from the predictions of the strong feedback variants of the FLAMINGO simulation ($2.5σ$ to $8.0σ$ tension), while fiducial FLAMINGO and BAHAMAS provide the closest match to our measurements. Using our measured baryon fractions and the SP(k) model, we infer a $10\%-15\%$ reduction in the matter power spectrum at $k = 5\ h\ \mathrm{Mpc}^{-1}$ relative to a dark matter-only universe, in agreement with fiducial FLAMINGO and BAHAMAS, while revealing a growing tension on smaller scales with the strong feedback variants.

Jeans criterion in hydrostatic stratified media

2026-09-09T10:50:17Z

The classical Jeans stability criterion neglects spatial gradients in background physical quantities such as density and pressure. Here, we revisit the Jeans analysis for a non-rotating fluid in hydrostatic equilibrium, explicitly accounting for these gradients and deriving a modified dispersion relation governing perturbation propagation. Using physical arguments, we show why the Jeans swindle is not required to derive the Jeans criterion for perturbations that are small compared to the size of the host system. We use linear perturbation analysis alongside an averaging procedure to study the behavior of the local Eulerian perturbations in a hydrostatic medium, showing that background gradients enter the dispersion relation in a highly non-trivial manner, precluding the derivation of a general gravitational stability criterion applicable to an arbitrary hydrostatic system. We demonstrate that, in the local short-wavelength limit, the standard Jeans criterion remains valid despite the presence of nonzero background gradients, and that all local perturbations in a hydrostatic system are stable against local gravitational collapse. We conclude that it is not possible to derive a general local gravitational stability criterion valid for an arbitrary hydrostatic background; however, in the local short-wavelength limit, the standard Jeans criterion remains valid.

Dust destruction in bubbles driven by multiple supernovae explosions

2026-09-09T10:16:15Z

Dust lifetime derived from an isolated supernova (SN) evolution in the interstellar medium is known to be an order of magnitude shorter than the time needed to replenish dust mass by its production in various Galactic sources. We show, using 3-D numerical hydrodynamical simulations, that destruction of dust in the case of multiple SNe in a star cluster is markedly different from that in an isolated SN. We find that the mass of dust destroyed in the bubble does not grow for a considerable time, while SNe continue to explode. This regime is attained at saturation timescale, which is proportional to SNe rate in cluster. We show that the mass of dust destroyed in bubble per SN decreases for higher SN rate. Thus, the destruction efficiency -- defined as the ratio of the the total mass of dust destroyed by clustered SNe to that destroyed by the same number of isolated SNe -- in bubbles evolved in a homogeneous medium drops for massive clusters, e.g. around clusters with $M_\ast \gtrsim 4\times 10^4 M_\odot$ it is less $0.4$\%. For lower mass clusters, the efficiency is proportional to the average time delay between SNe. We found that each cluster with $M_\ast \lesssim 4\times 10^4 M_\odot$ destroys the same mass of dust as a single isolated SN. In a clumpy medium in bubbles formed around clusters with $M_\ast \sim 4\times 10^4 M_\odot$ and up to 4 times around $M_\ast \sim 8\times 10^5 M_\odot$. We argue that the interstellar dust swept up by multiple SNe almost completely survives in the shells of bubbles around such massive clusters. Therefore, the destruction of the interstellar dust is controlled by SNe in low-mass clusters. We point out that the interstellar dust lifetime for a given SN rate is at least a factor $\sim 10$ longer as compared to the estimates derived from an isolated SN. (abridged)