Feng Wang

Context . Galactic cluster pairs are often interpreted as binary clusters, but physical proximity alone does not establish gravitational binding. Aims . We tested whether high-confidence Galactic cluster-pair candidates can be interpreted as classical bound binary clusters and whether their recent orbits show any weaker evidence of past compactness. Methods . We analysed 83 high-confidence cluster-pair candidates using present-day two-body energy diagnostics, 200 Myr Galactic orbit backtracking, and control populations drawn from the Hunt-Reffert (HR24) open-cluster catalogue. We computed η = E _k / E _g , E _orb = E _k − E _g , and an expansion factor EF = D _now / D _min . We compared the observed expansion factor distribution with both the 21444 unique separation-limited HR24 pair pool and a 10 ⁵ -draw replacement sample used only as a smoothed auxiliary baseline. Results . All 83 pairs fail the two-body binding condition: the median E _k / E _g is 6568, the 16th–84th percentile range is 436–26173, and the minimum is 72.7. The bound fraction remains zero after increasing all cluster masses by factors of three or ten. A two-sided Spearman test finds no age trend ( ρ = −0.107, p = 0.337 for the null of zero monotonic association). Orbit backtracking shows that 61.45% of the pairs were closer during the past 200 Myr, and 37.35% reached D _min < 50 pc. Relative to the unique HR24 pool, common-parent tests do not reject the random-pair null at the 5% level (Kolmogorov-Smirnov p = 0.205; Anderson-Darling p = 0.069). One-sided empirical tests of random size-83 draws give p = 0.093–0.104. The replacement baseline gives smaller formal p-values but reuses a finite pair pool. Conclusions . The robust result is negative: the current high-confidence Galactic cluster-pair sample should not be interpreted as a catalogue of classical bound binary clusters. A fossil-complex interpretation remains possible for a subset of unbound pairs, but the population-level evidence of excess past compactness is weak rather than decisive.

Context . Open clusters are fundamental benchmarks for calibrating stellar evolution models. However, the precise determination of cluster parameters is often hindered by the broadening of the main sequence due to unresolved binaries and stellar rotation, particularly near the main-sequence turnoff (MSTO). Ruprecht 147, as the oldest (∼2.5 Gyr) and nearest (. ∼ 300 pc) open cluster, provides a unique opportunity to resolve these degeneracies. Aims . We aim to simultaneously constrain the fundamental parameters (age, metallicity, distance, and extinction) and the underlying population properties (binary fraction and rotational distribution) of Ruprecht 147 using a rigorous statistical framework that explicitly models the “Kraft break” in stellar rotation. Methods . We constructed a 4D differentiable emulator based on PARSEC v2.0 stellar evolutionary tracks, utilizing JAX for GPU-accelerated inference. This emulator continuously interpolates mass, age, metallicity, and rotation rate (.). We performed a hierarchical Bayesian analysis on high-precision Gaia data release 3 photometry using the No-U-Turn Sampler (NUTS) to sample the high-dimensional posterior distribution. Results . Our analysis yields a precise distance modulus of ( m − M ) ₀ = 7.416 _−0.002 ^+0.003 mag and an age of . = 2.340 ± 0.002 Gyr. We derive a super-solar metallicity of [M/H] = 0.087 _−0.003 ^+0.001 dex. The inferred rotation distribution shape parameters (. _rot ≈ 7.9, . _rot ≈ 5.1) indicate a mean rotation rate of ω ≈ 0.46 for turnoff stars. We find a photometric binary fraction of . _bin ≈ 33.9%. Conclusions . The application of differentiable programming enables the robust breaking of the age-metallicity-extinction degeneracy. The detection of rapid rotation at the MSTO confirms the inefficiency of magnetic braking for stars with thin convective envelopes (the Kraft break), even at an advanced age of 2 Gyr. Its high metallicity suggests that Ruprecht 147 likely originated in the inner Galactic disk and migrated to its current location.

This paper presents CSST-PSFNet , a deep learning method for high-fidelity point-spread function (PSF) reconstruction developed for the Chinese Space Station Survey Telescope (CSST). The model integrates a residual neural network, a lightweight transformer architecture, and a variational latent representation to address key challenges in CSST imaging, including severe PSF undersampling, interband variability, and smooth spatial variation across the focal plane. Trained and validated on high-resolution star–PSF pairs generated by the CSST Main Survey Simulator, CSST-PSFNet achieves improved pixel-level accuracy and more precise recovery of shape parameters relevant to weak lensing compared to the widely used PSFEx . On both the standard test dataset and a blurred dataset representing the upper bound of expected on-orbit PSF degradation, the model achieves a size residual precision below 0.005 and an ellipticity residual precision below 0.002. A weak-label adaptation experiment further shows that the model can recover PSFEx -level performance when the true PSF is unknown, demonstrating robustness in controlled degradation scenarios and weak-label adaptation experiments. These results indicate that CSST-PSFNet provides a flexible and extensible framework for future on-orbit PSF calibration in large-scale CSST surveys, with potential applications in weak-lensing cosmology and precision astrophysical measurements.

The Square Kilometre Array (SKA) telescope is poised to yield massive observational data every year. The continuum imaging pipeline is a key and significant SKA data processing pipeline. An exact assessment of the continuum imaging outcomes not only serves to ascertain the overall operational integrity of the telescope but also furnishes invaluable references for subsequent data processing. In response to the urgent requirement for objective and accurate imaging quality assessment (QA) of the SKA continuum imaging pipeline, we empirically investigated the performance and availability of four source finders based on the full-scale SKA1-LOW simulation data generated by the OSKAR and the Radio Astronomy Simulation, Calibration, and Imaging Library (RASCIL2), carefully analyzing the performance of the source finder under different scenarios such as different thermal noise levels, imaging sizes, and field of view. We finally optimized the QA pipeline for SKA1-LOW according to empirical results, resulting in a discernible enhancement in both the performance and the accuracy of the pipeline.

Next-generation radio interferometers, with their enhanced spatial, temporal, and high-frequency resolution, will pose considerable challenges to data processing and storage. Frequency averaging reduces the volume of data, but would cause smearing of the bandwidth. The examination of bandwidth smearing is essential in the processing of radio interferometer data to achieve scientific objectives. Existing analysis methods can quantify parameters such as peak intensity loss in restored images under ideal assumptions. However, determining the shape of the source smearing in dirty images requires using complex observational simulation methods, which limits comprehensive studies of bandwidth smearing from frequency averaging and its optimal application. In this paper, we introduce a semianalytical method (SAM) and validate its accuracy through the simulation method. The SAM efficiently and precisely evaluates the impact of the smearing effect from frequency averaging for various telescope pointing positions, offering a plausible “distorted beam” representation of the non-phase-centered beam shape of point sources in the dirty image. Based on this beam shape, the specific shape of smearing after frequency averaging can be accurately and comprehensively described. This not only enables the determination of the extent of smearing, prediction of accuracy loss, and formulation of effective observation and data processing strategies prior to data processing, but also facilitates the development of more advanced frequency averaging methods.

Be stars are rapidly rotating B-type stars that exhibit Balmer emission lines in their optical spectra. These stars play an important role in studies of stellar evolution and disk structures. In this work, we carried out a systematic search for Be stars based on LAMOST spectroscopic data. Using low-resolution spectra from LAMOST DR11, we constructed a data set and developed a classification model that combines long short-term memory networks and convolutional neural networks , achieving a testing accuracy of 97.86%. The trained model was then applied to spectra with signal-to-noise ratios greater than 10, yielding 55,667 B-type candidates. With the aid of the MKCLASS automated classification tool and manual verification, we finally confirmed 40,223 B-type spectra. By cross-matching with published Hα emission-line star catalogs, we obtained a sample of 8298 Be stars, including 3787 previously reported Be stars and 4511 newly discovered. Furthermore, by incorporating color information, we classified the Be star sample into Herbig Be stars and Classical Be stars. In total, we identified 3363 Classical Be stars and 35 Herbig Be stars. The B-type and Be star catalogs derived in this study, together with the code used for model training, have been publicly released to facilitate community research.

Open clusters play a critical role in the study of both stellar and Galactic evolution. We present a comprehensive analysis of the open cluster NGC 6134 using Gaia DR3 astrometric and photometric data, supplemented with multiwavelength observations from the Two Micron All Sky Survey, Wide-field Infrared Survey Explorer, NEOWISE, and Transiting Exoplanet Survey Satellite (TESS). A total of 714 high-probability members were identified using the Hierarchical Density-based Spatial Clustering of Applications with Noise algorithm. The cluster parameters were derived by Bayesian isochrone fitting with BASE-9, resulting in log (age/yr) = 9.1683 ± 0.019, [Fe/H] = 0.0911 ± 0.04, and A _V = 1.2095 ± 0.015 mag. The cluster is dynamically relaxed (t _rh = 16.59 ± 0.25, ≪age) with significant mass segregation (Λ_MSR = 1.71 ± 0.26). Galactic orbital integration using galpy indicates that the cluster resides in the thin disk on a slightly eccentric orbit. We identified three candidates of blue straggler stars (BSSs). The first candidate was modeled as an EA-type semidetached eclipsing binary using TESS photometry and Markov Chain Monte Carlo fitting, and Candidate 2 was fitted with a spectral energy distribution (SED) using the Virtual Observatory SED Analyzer tool, providing estimates of its temperature, surface gravity, and metallicity based on the SED-fitting results. However, the third candidate lacks sufficient observational data, which prevents further analysis. These results highlight binary mass transfer as a dominant channel of BSS candidates in NGC 6134, providing observational evidence for the formation mechanism of blue stragglers.

Hunting for open clusters (OCs) in Gaia data is a hot topic for astronomical big data analysis. Significant progress has been made in searching for OCs in Gaia data using machine learning methods. Based on our previous research, we applied a hybrid unsupervised clustering algorithm (friends-of-friends and pyUPMASK) and a binary classification algorithm (Random Forests) to perform fine-grained blind searching beyond $5\:$kpc on Gaia DR3. After isochrone-fitting, cross-matching, and visual inspection, we obtained 2932 plausible candidate clusters, 436 of which have been published as candidates in other catalogs. We performed a Kings’s model profile fitting and a comparative study of theoretical tidal radii and the observed mass–radius correlation to determine the physical reality of these OCs. After physical reality checks, we validated the remaining 872 candidate clusters using statistical analysis and dynamical binding distinction. Statistical analysis shows that the distributions of these candidates’ proper motion, OC age, and metallicity are consistent with other related studies. We also analyzed the intrinsic dispersion of morphological features, sky maps, ages, and metallicities of these candidate clusters. Following rigorous verifications and analytical validations, we have ultimately identified 739 OC candidates with high confidence.

Deconvolution in radio interferometry faces challenges due to incomplete sampling of the visibilities in the spatial frequency domain caused by a limited number of antenna baselines, resulting in an ill-posed inverse problem. Reconstructing dirty images into clean ones is crucial for subsequent scientific analysis. To address these challenges, we propose an U-Net based method that extracts high-level information from the dirty image and reconstructs a clean image by effectively reducing artifacts and sidelobes. The U-Net architecture, consisting of an encoder-decoder structure and skip connections, facilitates the flow of information and preserves spatial details. Using simulated data of radio galaxies, we train our model and evaluate its performance on the testing set. Compared with the CLEAN method and the Visibility and Image Conditioned Denoising Diffusion Probabilistic Model (VIC-DDPM), our proposed model can effectively reconstruct both extended sources and faint point sources with higher values in the Structural Similarity Index Measure (SSIM) and the Peak Signal-to-Noise Ratio (PSNR). Furthermore, we investigate the impact of noise on the model performance, demonstrating its robustness under varying noise levels.

Radio telescopes produce visibility data about celestial objects, but these data are sparse and noisy. As a result, images created on raw visibility data are of low quality. Recent studies have used deep learning models to reconstruct visibility data to get cleaner images. However, these methods rely on a substantial amount of labeled training data, which requires significant labeling effort from radio astronomers. Addressing this challenge, we propose VisRec, a model-agnostic semi-supervised learning approach to visibility data reconstruction in radio astronomy. Specifically, VisRec consists of both a supervised learning module and an unsupervised learning module. In the supervised learning module, we introduce a set of data augmentation functions to produce diverse visibility examples. In comparison, the unsupervised learning module in VisRec augments unlabeled data and uses reconstructions from non-augmented visibility as pseudo-labels for training. This hybrid approach allows VisRec to effectively leverage both labeled and unlabeled data. This way, VisRec performs well even when labeled data is scarce. Our evaluation results show that VisRec is applicable to various models, and outperforms all baseline methods in terms of reconstruction quality, robustness, and generalizability.