The NSF-DOE Vera C. Rubin Observatory, Roman Space Telescope, Euclid, and other next-generation surveys will deliver imaging, spectroscopic, and time-domain data at scales that increasingly shift the bottleneck in astronomical machine learning (ML) projects from model design to infrastructure. We present Hyrax, an open-source, modular, GPU-enabled Python framework that supports the full ML lifecycle in astronomy: from data acquisition and training to inference and experiment comparison, with capabilities including multimodal dataset support, integrated vector databases for similarity search, and interactive two- and three-dimensional latent-space exploration for unsupervised discovery. We demonstrate Hyrax's versatility through five representative applications on real survey data: (i) unsupervised representation learning on $\sim 4\times10^5$ Rubin Legacy Survey of Space and Time (LSST) Data Preview 1 (DP1) galaxies, surfacing new merger and low-surface-brightness candidates missing from reference Euclid and Dark Energy Survey catalogs, while also isolating imaging artifacts -- all without labeled training data; (ii) hybrid density-based clustering for identifying cluster-scale gravitational lens candidates in DP1 data; (iii) multimodal early-time transient classification in the Zwicky Transient Facility leveraging light curves, spectra, images, and metadata; (iv) supervised false-positive filtering in shift-and-stack searches for distant solar system objects in the Dark Energy Camera Ecliptic Exploration Project survey; and (v) supervised detection of semi-resolved dwarf galaxies in Hyper Suprime-Cam and LSST-like imaging using synthetic source injection. Together, these results demonstrate that Hyrax provides astronomy-specific ML infrastructure that enables systematic discovery and rapid methodological iteration across next-generation astronomical surveys.
"The Dusty Universe: The Fifth Pandust Conference" took place in Tucson (AZ, USA) from November 10 until 14, 2025. The goal of this meeting was to get the dust community together to review where we are, hear exciting new results, and make plans for the future. The meeting encompassed all aspects of dust investigations including observations, theory, modeling, and laboratory studies. The conference consisted of invited review talks, contributed talks and posters. Science topics included interstellar dust (Milky Way & nearby galaxies), circumstellar dust (including Solar System & exoplanets), dust in Galaxies (including high-z), lifecycle of dust, and future needs (laboratory, theory, & observations), with a particular focus on results from JWST and ALMA, and on nanodust (including PAHs). On November 12th, we organized breakout discussion sessions covering a wide range of interesting dust-related topics. The purpose of this document is to capture the main topics/questions that were discussed, the key conclusions of these discussions, the challenges and possible solutions that were brought up, and the open questions that still remain to be answered. We hope that this document records our findings and challenges for the future generation.
Emission from the pure rotational transitions of H$_2$ traces warm molecular gas, providing insight into its temperature distribution and local heating conditions. We have extended previous power-law H$_2$ temperature models to account for differential extinction by dust as well as non-equilibrium ortho-to-para-H$_2$ ratios (OPR). The turbulent environment of the M82 starburst offers a unique opportunity to study H$_2$ out of equilibrium conditions, using ~15 pc spatially resolved measurements from MIRI/MRS on JWST. With extensive detections of H$_2$ S(1)-S(7), we use our model to assess spatial variations in local heating conditions of molecular gas across a ~500 pc region of the M82 central starburst. The average slope of the recovered H$_2$ power law temperature distribution is consistent with prior studies, and the slope strongly anti-correlates with relative [Fe II]/H$_2$ S(1)-S(2) strength, pointing to the importance of shock-heating. Our models indicate that the OPR is, on average, about half of its equilibrium value. This suppression is attributed to cloud mixing timescales which are short compared to timescales for spin conversion, with molecular gas remembering its ''cooler past''. By accounting for OPR disequilibrium, we can identify instances of recent and rapid heating to better understand the flow of energy through the interstellar medium and track its thermal history.
Galaxy physical properties-such as star formation rate (SFR), stellar mass, and gas-phase metallicity-are essential for population studies and evolutionary analyses. Deriving these quantities for billions of galaxies in modern imaging surveys presents significant challenges due to limited spectroscopy and the computational costs associated with traditional spectral energy distribution fitting. As a result, many galaxies in large photometric surveys still lack homogeneous property estimates. This study introduces a multimodal deep learning model that integrates optical imaging with photometric catalog features to estimate SFR, stellar mass, and oxygen abundance in low-redshift galaxies. The model incorporates a ResNet-based convolutional neural network to extract spatial information from multiband images and a multilayer perceptron that processes catalog-level photometric features, leveraging complementary constraints from morphology, surface brightness, and broadband colors. Trained on reference measurements from the MPA-JHU DR8 catalog, the model is optimized for efficient large-scale estimation. When applied to the DESI Legacy Imaging Surveys (LS) DR10, the model generates a value-added catalog containing physical property estimates for approximately 547 million galaxies with redshifts z <= 0.5. Validation through comparisons with independent catalogs and exploration of key scaling relations demonstrates that while the derived properties are not intended for precision measurements of individual objects, they effectively capture the dominant astrophysical trends necessary for ensemble studies. This catalog represents the first homogeneous set of photometry-based SFR, stellar mass, and metallicity estimates for DESI LS DR10, providing a vital resource for statistical studies of galaxies in the local Universe and facilitating comparisons with current and future spectroscopic surveys.
Recent results from the Dark Energy Spectroscopic Instrument (DESI) provide evidence for a dynamical dark-energy component, whose equation of state appears to have recently crossed the phantom divide. In this Letter, we present an interacting dark-energy model, grounded in field theory, that naturally accommodates such a double crossing. In our framework, fermionic dark matter is coupled via a Yukawa interaction to a tachyonic scalar field governed by Born-Infeld dynamics. The phantom crossing arises at the level of the effective dark-energy equation of state, while the underlying scalar-field dynamics remains nonphantom and well bounded. We confront our model with data including BAO from the DESI (DR2) survey, CMB distance priors from Planck 2018, and the latest Type Ia supernovae compilations, obtaining robust constraints across the different data combinations and reconstructing a recent double crossing of the phantom divide. Furthermore, under naturalness assumptions, the model expects an ultralight fermionic dark matter mass of order $1.9\times10^{-3}\,\mathrm{eV}$, suggesting a possible connection with new light particles in the dark sector and motivating future tests with cosmological perturbations.
The radio galaxy NGC 1275 is the Brightest Cluster Galaxy in the Perseus cluster. It is well-studied across all wavebands, including Very High Energy (VHE; E>100GeV gamma-rays, and with radio observations over the last 20 years tracking an unusual radio component, "C3". NGC 1275 was observed in an exceptional VHE flaring state between 2016 December 31 and 2017 January 3. The flare peak reached ~1.5 Crab units as measured by the MAGIC observatory. We report on the observations of NGC~1275 conducted by VERITAS and multi-wavelength data collected during this flaring state, and for context, data taken between 2009 and 2017 inclusive. VERITAS detected the declining state of the flare on 2017 January 2 (MJD 57755) and 3 (MJD 57756) at an average flux state of 0.5 Crab units. VERITAS spectra show an overall long-term trend of harder-when-brighter. During the flare, the gamma-ray spectrum obtained from the combined Fermi-LAT, MAGIC, and VERITAS observations, changes from a power law with an exponential cut-off on January 1 to a log-parabola on January 2. To study the evolution of the flare in more detail, multi-band spectral energy distributions (SEDs) were constructed for the nights of 2017 January 1 and 2 corresponding to the shift from the peak to the decline of the flare. A blob-in-jet modeling of the SEDs results in support for a two-component model with a jet angle of 10 degrees to the line of sight and the gamma-ray emission zone located in the vicinity of the C3 radio component.
this https URL for more details
this https URL Catalogs will be accessible long-term at the IRSA-served DOI link: this https URL