Context. Investigating substellar populations in young massive clusters offers crucial insights into the formation of brown dwarfs (BDs) and the role of environmental conditions in shaping their properties. Westerlund 2 (Wd2), as one of the nearest dense and massive clusters in the Milky Way, represents an ideal laboratory for studying the effects of high stellar density and ionizing radiation from massive stars on BD formation. Aims. This paper presents deep JWST/NIRCam observations of Wd2 in a large number of filters between 1.15 and 4.1 µm. Our analysis focuses on identifying and characterizing the BD population within the cluster. Methods. We carried out point spread function (PSF) photometry on deep JWST/NIRCam data obtained in four wide and six medium-band filters, using DOLPHOT. The resulting catalog was used to identify BD candidates in Wd2 through spectral energy distribution (SED) fitting with atmospheric models. The methodology was validated using synthetic JWST photometry of spectroscopically confirmed young BDs together with simulated populations of typical contaminants, allowing us to define an empirical selection criteria. Results. The 50% detection limit of our NIRCam catalog is ∼0.015–0.02 M⊙, at the distance, age, and extinction of Wd2, providing the deepest view of a supermassive star cluster to date. The validation demonstrates that the adopted SED-fitting methodology reliably recovers young brown dwarfs while remaining robust against contamination from reddened evolved stars. Applying this procedure to Wd2, we identify 353 substellar candidates. Most candidates (301) lie above the 10 Myr isochrone in the Hertzsprung-Russell diagram consistent with cluster membership and are defined as strong candidates. Comparison with a control field indicates a contamination level of ≲5% for the strong candidate sample down to masses of ∼0.01–0.015 M⊙. We identify 73 candidates exhibiting an infrared excess indicative of circumstellar disks. These objects occupy the expected infrared-excess locus in de-reddened color diagrams and represent 27.7+3.4−3.2% of the candidates detected in F410M. Conclusions. These results demonstrate the capability of JWST/NIRCam medium-band photometry to identify robust samples of candidate brown dwarfs in distant, embedded, and crowded massive star clusters using imaging observations alone. The resulting catalog provides a well-characterized sample for future spectroscopic confirmation and subsequent studies of the substellar population of Wd2.
Context. Surface convection in cool stars sets the entropy jump between the atmosphere and the deep convective envelope and therefore affects the radius and structure of one-dimensional stellar models. In standard local treatments, this entropy jump is controlled through an effective convective efficiency. Aims. We have developed a phase-space closure that relates the mean stratification to the local distribution of entropy and velocity among convective elements. Our aim is to describe how the surface entropy transition emerges from the statistics of these elements rather than prescribing it directly. Methods. We first constructed a maximum-entropy model for the convective bulk, assuming that the realised states are bounded by a deep entropy level S. This prediction was then confronted with three-dimensional radiation-hydrodynamics simulations from the M3DIS grid. The comparison motivated an extension to a two-population closure in which an optimal bulk component coexists with a radiatively cooled plume component. Results. The maximum-entropy construction predicts a one-sided exponential entropy distribution, and its low-entropy tail is recovered in the deep convective part of the simulations. Closer to the surface, the distribution is reorganised by radiative cooling and separates into two thermodynamic components. Using a radiative entropy branch, an adiabatically descending cooled branch, and a population fraction controlled by optical depth, the two-population closure reproduces the mean entropy stratification of the four reference simulations over most of the displayed range without adjusting the entropy profile itself. Conclusions. The surface transition can therefore be interpreted as a continuous population decomposition, rather than as a sharp boundary between radiative and convective layers. This provides a natural route towards one-dimensional stellar models in which the surface entropy jump is determined from the statistical organisation of the flow instead of being set by a calibrated mixing-length efficiency.
Aims. The ratio between infrared and visible albedo, pIR/pV, provides a simple proxy for the spectral slope of asteroid surfaces between the visible and near-infrared wavelength ranges. The aim of this work is to assess whether the different spectral units of Phobos occupy the same spectral parameter space as primitive asteroid populations in order to confirm, or not, its possible asteroidal origin. Methods. For each of the three spectral units of Phobos (red, transitional, and blue), we extracted averaged, normalised reflectance spectra by comparing normalised reflectance in the visible (0.50–0.60 m) and near-infrared (3.1–3.8 m) wavelength ranges from the Mars Express OMEGA instrument to compute a proxy for the infrared-visible albedo ratio (RP). Then, using the database of visible and infrared albedos compiled in the Minor Planet Physical Properties Catalogue (MP3C), we computed asteroid pIR/pV values (RA) for comparison with those of Phobos. Results. The RP values for the red, transitional, and blue regions vary between 2.90 and 3.12, 2.32 and 2.35, 1.64 and 1.85, respectively. These values span a range occupied by primitive asteroid populations, with the transitional and red regions coinciding with the high ratio regime within the pIR/pV domain, commonly associated with red-sloped primitive asteroids, including D- and T-type objects as well as Jupiter trojans.
We are witnessing a surge in observations of the cosmic dawn and the epoch of reionisation, driving an increasing demand for fast and robust theoretical interpretation frameworks. In response, machine learning -- emulation in particular -- has emerged as a powerful approach for accelerating and enhancing inference pipelines. In this work, we present 21cmEMUv3, an emulator trained on 21cmFASTv3 simulations that model both atomically and molecularly cooling galaxies. 21cmEMUv3 is conditioned on σ_8 and ten astrophysical parameters to produce seven summary observables: (i) the cylindrical 21cm power spectrum (PS), emulated for the first time at high resolution and accuracy across a wide redshift range of z ∼ 6--30; (ii) the spherically averaged 21cm PS; (iii) the mean neutral fraction of the intergalactic medium; (iv) the mean 21cm spin temperature; (v) the global 21cm signal; (vi) the ultraviolet luminosity functions; and (vii) the Thomson scattering optical depth. Notably, the cylindrical 21cm PS is emulated via score-based diffusion, while the remaining six summaries are emulated via long short-term memory networks, all achieving łesssim1% median accuracy. We used the emulator to reinterpret current 21cm PS upper limits from HERA, for the first time using state-of-the-art hydrodynamical simulations to inform priors on star formation inside molecularly cooling galaxies. We find that our inferred soft-band X-ray luminosity per unit star formation rate is consistent with extrapolations of high-mass X-ray binaries to the low-metallicity regimes expected in the first galaxies, excluding values below 10^39.4 erg s -1 M^ -1 _⊙ $ at $95% confidence. Finally, we produced forecasts for the detection of the cosmic 21cm PS with the Square Kilometre Array for different array configurations. The 21cmEMU package is publicly available. yr
Frequency-dependent radio time delays have been measured in Sgr A* light curves, with higher-frequency emission generally leading lower-frequency emission. These frequency-dependent lags have been interpreted in terms of an organized jet-like outflow, expanding plasma components, or magnetic-flux eruptions, although their origin and recurrence remain uncertain. We aim to search for and quantify frequency-dependent radio time delays in a new high-cadence Karl G. Jansky Very Large Array (VLA) observation of Sgr A*. Using the improved broad 12--37 GHz frequency coverage on the VLA, we measured the timing of the dominant variability feature and compared the resulting lag--wavelength slope with previous measurements. Sgr A* was observed with the VLA on 28 May 2022 in the Ku, K, and Ka bands over an sim6 hour track. We extracted both band-averaged and 2 GHz subband light curves and measured the delays using a peak-time method and a locally normalized cross-correlation function. The light curves show one broad variability feature across the observed frequency range, with the maximum occurring progressively later at lower observing frequencies. Both timing methods recover a positive lag--wavelength relation, with a slope of order 40 . This value agrees with time lag results reported over more than 15 years from the first VLA observations at 22--43 GHz to later multifrequency VLA and ALMA campaigns. -1 The sign and magnitude of the estimated delay are consistent with outward propagation through an optically thick synchrotron-emitting region. The recurrence of comparable delays over almost two decades suggests that frequency-dependent radio lags are a persistent feature of Sgr A* variability. Broadband, multi-epoch VLA monitoring of Sgr A* provides a useful probe of its radio variability, but robust lag measurements require high cadence and improved $(u,v)$ coverage of both Sgr A* and the surrounding extended emission such as by combining different VLA configurations.
