arXiv subjects

Fiorenzo Stoppa

Publications and source records attributed to Fiorenzo Stoppa.

13 recordsLinked to original sources

arXiv2026

SNID-SAGE: A Modern Framework for Interactive Supernova Classification and Spectral Analysis

We present SNID-SAGE (SuperNova IDentification-Spectral Analysis and Guided Exploration), a framework for supernova spectral classification with both a fully interactive graphical interface and a scriptable command-line pipeline for large-scale processing. The pipeline combines deterministic spectral preprocessing, FFT-based cross-correlation against a curated template library, ranking of candidate matches using a composite quality metric, and consolidation of redshift and classification solutions into a single result with associated quality and confidence estimates. SNID-SAGE includes an upgradeable template library (about 6000 spectra), interactive line identification with velocity measurements, and optional natural-language summaries of classification results. We evaluate SNID-SAGE using two complementary tests: (i) leave-one-out cross-validation, in which each template spectrum is matched against the remainder of the library; and (ii) large-scale application to WISeREP spectra with valid coverage across the 4000-7000 A interval, irrespective of spectral type, comprising approximately 46 000 spectra, with redshift validation against known host-galaxy measurements where available. The full validation results and the SNID-SAGE framework are publicly available, supporting integration into spectroscopic survey workflows.

astro-ph.IM↗

arXiv2026

ATLAS100 -- I. A volume-limited sample of supernovae and related transients within 100 Mpc

We present ATLAS100 -- a sample of 1729 supernovae and other explosive optical transients within $\sim 100$ Mpc observed by the ATLAS survey over a span of 5.75 years from 2017 September 21 to 2023 June 21. The volume-limited sample includes transients associated with galaxies with a spectroscopic redshift of $z \leq 0.025$, and spectroscopically classified transients within this redshift threshold where a host redshift was not available in existing catalogues. Our host galaxy list is constructed from aggregating all available galaxy redshift and distance catalogues. We carefully select all transients within a projected radius of 50\,kpc of these hosts. The ATLAS100 transient sample has a host galaxy redshift completeness fraction of $83$ per cent, consistent with expectations for the redshift completeness of local galaxy catalogues. Within this volume, the spectroscopic classifications are 87 per cent complete and we reclassify many ambiguous transients with joint light curve and spectroscopic considerations. Here, we release the catalogue together with compiled, binned and cleaned ATLAS photometry for all transients. We fit the light curve data to derive peak luminosity and characteristic timescales. We explore the sample characteristics, demographics and discuss completeness and purity of the sample.

astro-ph.HE↗

arXiv2026

A Faint Progenitor System for the Faint Supernova 2024vjm

Type Ia Supernovae (SNe Ia) are well known for their role as standardizable cosmological candles. Their uniformity is credited to their single origin as thermonuclear explosions of White dwarf (WD) stars. Nevertheless, some SNe Ia break this regularity. Prominently, the Iax subclass are less energetic and remarkably diverse, raising questions about their progenitor systems. While no progenitor system of a normal SN Ia has ever been detected, a luminous blue star was identified in pre-explosion images of the site of the bright SN Iax SN 2012Z, suggested to be a helium giant companion star acting as a mass donor to a WD SN progenitor. This is in line with models of weak mass accretion of a WD from a binary companion, producing an explosion that does not fully disrupt the star. However, these models fail to explain the properties of the faintest Type Iax explosions, suggesting either they originate from other WD binary systems, or even from massive progenitor stars. Here, we present the faint SN Iax SN 2024vjm - possibly the faintest supernova observed to date. Using a deep pre-explosion image taken by the recently launched Euclid space mission, we show that its progenitor system must be fainter than the helium giant SN Iax progenitor candidate of SN 2012Z, as well as that of the luminous red companion or remnant of the faint SN 2008ha, and may require a subdwarf helium star as a mass donor. The deep image also provides strong arguments against a massive star origin for this faint supernova. Our observations argue that SN 2024vjm is a WD explosion, but we find that remarkably faint SNe Iax fade more slowly than bright ones, i.e., they evolve in an opposite manner from the famous Phillips relation that makes regular SNe Ia cosmological candles.

astro-ph.HE↗

arXiv2025

Textual interpretation of transient image classifications from large language models

Modern astronomical surveys deliver immense volumes of transient detections, yet distinguishing real astrophysical signals (for example, explosive events) from bogus imaging artefacts remains a challenge. Convolutional neural networks are effectively used for real versus bogus classification; however, their reliance on opaque latent representations hinders interpretability. Here we show that large language models (LLMs) can approach the performance level of a convolutional neural network on three optical transient survey datasets (Pan-STARRS, MeerLICHT and ATLAS) while simultaneously producing direct, human-readable descriptions for every candidate. Using only 15 examples and concise instructions, Google's LLM, Gemini, achieves a 93% average accuracy across datasets that span a range of resolution and pixel scales. We also show that a second LLM can assess the coherence of the output of the first model, enabling iterative refinement by identifying problematic cases. This framework allows users to define the desired classification behaviour through natural language and examples, bypassing traditional training pipelines. Furthermore, by generating textual descriptions of observed features, LLMs enable users to query classifications as if navigating an annotated catalogue, rather than deciphering abstract latent spaces. As next-generation telescopes and surveys further increase the amount of data available, LLM-based classification could help bridge the gap between automated detection and transparent, human-level understanding.

astro-ph.IM↗

arXiv2025

BGRem: A background noise remover for astronomical images based on a diffusion model

Context: Astronomical imaging aims to maximize signal capture while minimizing noise. Enhancing the signal-to-noise ratio directly on detectors is difficult and expensive, leading to extensive research in advanced post-processing techniques. Aims: Removing background noise from images is a valuable pre-processing step catalog-building tasks. We introduce BGRem, a machine learning (ML) based tool to remove background noise from astronomical images. Methods: BGRem uses a diffusion-based model with an attention U-Net as backbone, trained on simulated images for optical and gamma ({\gamma})-ray data from the MeerLICHT and Fermi-LAT telescopes. In a supervised manner, BGRem learns to denoise astronomical images over several diffusion steps. Results: BGRem performance was compared with a widely used tool for cataloging astronomical sources, SourceExtractor (SExtractor). It was shown that the amount of true positive sources using SExtractor increased by about 7% for MeerLICHT data when BGRem was used as a pre-processing step. We also show the generalizability of BGRem by testing it with optical images from different telescopes and also on simulated {\gamma}-ray data representative of the Fermi-LAT telescope. We show that in both cases, BGRem improves the source detection efficiency. Conclusions: BGRem can improve the accuracy in source detection of traditional pixel-based methods by removing complex background noise. Using zero-shot approach, BGRem can generalize well to a wide range of optical images. The successful application of BGRem to simulated {\gamma}-ray images, alongside optical data, demonstrates its adaptability to distinct noise characteristics and observational domains. This cross-wavelength performance highlights its potential as a general-purpose background removal framework for multi-wavelength astronomical surveys.

astro-ph.IM↗

arXiv2025

Towards a foundation model for astrophysical source detection: An End-to-End Gamma-Ray Data Analysis Pipeline Using Deep Learning

The increasing volume of gamma-ray data demands new analysis approaches that can handle large-scale datasets while providing robustness for source detection. We present a Deep Learning (DL) based pipeline for detection, localization, and characterization of gamma-ray sources. We extend our AutoSourceID (ASID) method, initially tested with \textit{Fermi}-LAT simulated data and optical data (MeerLICHT), to Cherenkov Telescope Array Observatory (CTAO) simulated data. This end-to-end pipeline demonstrates a versatile framework for future application to other surveys and potentially serves as a building block for a foundational model for astrophysical source detection.

astro-ph.IM↗

arXiv2025

A catalog to unite them all: REGALADE, a revised galaxy compilation for the advanced detector era

Many applications in transient science, gravitational wave follow-up, and galaxy population studies require all-sky galaxy catalogs with reliable distances, extents, and stellar masses. However, existing catalogs often lack completeness beyond $\sim 100$ Mpc, suffer from stellar contamination, or do not provide homogeneous stellar mass estimates and size information. Our goal is to build a high-purity, high-completeness, all-sky galaxy catalog out to 2,000 Mpc, specifically designed to support time-domain and multi-messenger astrophysics. We combined major galaxy catalogs and deep imaging surveys -- including the Legacy Surveys, Pan-STARRS, DELVE, and SDSS -- and added spectroscopic, photometric, and redshift-independent distances. We cleaned the sample using the Gaia catalog to remove stars and visually inspected all ambiguous cases below 100 Mpc through a classification platform that gathered 27,000 expert votes. Stellar masses were estimated using optical and mid-infrared profile-fit photometry, and we improved the accuracy of photometric distances by combining multiple independent estimates. The resulting catalog, REGALADE, includes nearly 80 million galaxies with distances under 2,000 Mpc. It provides stellar masses for 88% of the sample and ellipse fits for 80%. REGALADE is more than 90% complete for galaxies contributing 50% of the total $r$-band luminosity out to 360 Mpc. In science tests, it recovers 60% more known supernova hosts, doubles the number of low-luminosity transient hosts, and identifies more reliable hosts for ultraluminous and hyper-luminous X-ray sources. REGALADE is one of the most complete and reliable all-sky galaxy catalog to date for the nearby Universe, built for real-world applications in transient and multi-messenger astrophysics. The full dataset, visual classifications, and code will be released to support broad community use.

astro-ph.GA↗

arXiv2024

Two waves of massive stars running away from the young cluster R136

Massive stars are predominantly born in stellar associations or clusters. Their radiation fields, stellar winds, and supernovae strongly impact their local environment. In the first few million years of a cluster's life, massive stars are dynamically ejected running away from the cluster at high speed. However, the production rate of dynamically ejected runaways is poorly constrained. Here we report on a sample of 55 massive runaway stars ejected from the young cluster R136 in the Large Magellanic Cloud. Astrometric analysis with Gaia reveals two channels of dynamically ejected runaways. The first channel ejects massive stars in all directions and is consistent with dynamical interactions during and after the birth of R136. The second channel launches stars in a preferred direction and may be related to a cluster interaction. We find that 23-33% of the most luminous stars initially born in R136 are runaways. Model predictions have significantly underestimated the dynamical escape fraction of massive stars. Consequently, their role in shaping and heating the interstellar and galactic medium, along with their role in driving galactic outflows, is far more important than previously thought.

astro-ph.SR↗

arXiv2023

Consistency Tests for Comparing Astrophysical Models and Observations

In astronomy, there is an opportunity to enhance the practice of validating models through statistical techniques, specifically to account for measurement error uncertainties. While models are commonly used to describe observations, there are instances where there is a lack of agreement between the two. This can occur when models are derived from incomplete theories, when a better-fitting model is not available or when measurement uncertainties are not correctly considered. However, with the application of specific tests that assess the consistency between observations and astrophysical models in a model-independent way, it is possible to address this issue. The consistency tests (ConTESTs) developed in this paper use a combination of non-parametric methods and distance measures to obtain a test statistic that evaluates the closeness of the astrophysical model to the observations. To draw conclusions on the consistency hypothesis, a simulation-based methodology is performed. In particular, we built two tests for density models and two for regression models to be used depending on the case at hand and the power of the test needed. We used ConTEST to examine synthetic examples in order to determine the effectiveness of the tests and provide guidance on using them while building a model. We also applied ConTEST to various astronomy cases, identifying which models were consistent and, if not, identifying the probable causes of rejection.

astro-ph.IM↗

arXiv2022

AutoSourceID-Light. Fast Optical Source Localization via U-Net and Laplacian of Gaussian

$\textbf{Aims}$. With the ever-increasing survey speed of optical wide-field telescopes and the importance of discovering transients when they are still young, rapid and reliable source localization is paramount. We present AutoSourceID-Light (ASID-L), an innovative framework that uses computer vision techniques that can naturally deal with large amounts of data and rapidly localize sources in optical images. $\textbf{Methods}$. We show that the AutoSourceID-Light algorithm based on U-shaped networks and enhanced with a Laplacian of Gaussian filter (Chen et al. 1987) enables outstanding performances in the localization of sources. A U-Net (Ronneberger et al. 2015) network discerns the sources in the images from many different artifacts and passes the result to a Laplacian of Gaussian filter that then estimates the exact location. $\textbf{Results}$. Application on optical images of the MeerLICHT telescope demonstrates the great speed and localization power of the method. We compare the results with the widely used SExtractor (Bertin & Arnouts 1996) and show the out-performances of our method. AutoSourceID-Light rapidly detects more sources not only in low and mid crowded fields, but particularly in areas with more than 150 sources per square arcminute.

astro-ph.IM↗

arXiv2022

The GW-Universe Toolbox II: constraining the binary black hole population with second and third generation detectors

We employ the method used by the GW-Universe Toolbox to generate a synthetic catalogue of detection of stellar mass binary black hole (BBH) mergers. We study advanced LIGO (aLIGO) and Einstein Telescope (ET) as two representatives for the 2nd and 3rd generation GW observatories, and study how GW observations of BBHs can be used to constrain the merger rate as function of redshift and masses. We also simulate the observations from a detector that is half as sensitive as the ET at design which represents an early phase of ET. Two methods are used to obtain the constraints on the source population properties from the catalogues: 1. parametric differential merger rate model and applies a Bayesian inference on the parameters; and 2. non-parametric and uses weighted Kernel density estimators. The results show the overwhelming advantages of the 3rd generation detector over the 2nd generation for the study of BBH population properties, especially at a redshifts higher than ~2, where the merger rate is believed to peak. With the simulated aLIGO catalogue, the parametric Bayesian method can still give some constraints on the merger rate density and mass function beyond its detecting horizon, while the non-parametric method lose the constraining ability completely there. We also find that, despite the numbers of detection of the half-ET can be easily compatible with full ET after a longer observation duration, the catalogue from the full ET can still give much better constraints on the population properties, due to its smaller uncertainties on the physical parameters of the GW events.

astro-ph.HE↗

arXiv2022

The Gravitational Wave Universe Toolbox: A software package to simulate observation of the Gravitational Wave Universe with different detectors

Context. As the importance of Gravitational Wave (GW) Astrophysics increases rapidly, astronomers in different fields and with different backgrounds can have the need to get a quick idea of which GW source populations can be detected by which detectors and with what measurement uncertainties. Aims. The GW-Toolbox is an easy-to-use, flexible tool to simulate observations on the GW universe with different detectors, including ground-based interferometers (advanced LIGO, advanced VIRGO, KAGRA, Einstein Telescope, and also customised designs), space-borne interferometers (LISA and a customised design), pulsar timing arrays mimicking the current working ones (EPTA, PPTA, NANOGrav, IPTA) and future ones. We include a broad range of sources such as mergers of stellar mass compact objects, namely black holes, neutron stars and black hole-neutron stars; and supermassive black hole binaries mergers and inspirals, Galactic double white dwarfs in ultra-compact orbit, extreme mass ratio inspirals and Stochastic GW backgrounds. Methods. We collect methods to simulate source populations and determine their detectability with the various detectors. The paper aims at giving a comprehensive description on the algorithm and functionality of the GW-Toolbox. Results. The GW-Toolbox produces results that are consistent with more detailed calculations of the different source classes and can be accessed with a website interface (gw-universe.org) or as a python package (https://bitbucket.org/radboudradiolab/gwtoolbox). In the future, it will be upgraded with more functionality.

astro-ph.HE↗

arXiv2021

MeerCRAB: MeerLICHT Classification of Real and Bogus Transients using Deep Learning

Astronomers require efficient automated detection and classification pipelines when conducting large-scale surveys of the (optical) sky for variable and transient sources. Such pipelines are fundamentally important, as they permit rapid follow-up and analysis of those detections most likely to be of scientific value. We therefore present a deep learning pipeline based on the convolutional neural network architecture called $\texttt{MeerCRAB}$. It is designed to filter out the so called 'bogus' detections from true astrophysical sources in the transient detection pipeline of the MeerLICHT telescope. Optical candidates are described using a variety of 2D images and numerical features extracted from those images. The relationship between the input images and the target classes is unclear, since the ground truth is poorly defined and often the subject of debate. This makes it difficult to determine which source of information should be used to train a classification algorithm. We therefore used two methods for labelling our data (i) thresholding and (ii) latent class model approaches. We deployed variants of $\texttt{MeerCRAB}$ that employed different network architectures trained using different combinations of input images and training set choices, based on classification labels provided by volunteers. The deepest network worked best with an accuracy of 99.5$\%$ and Matthews correlation coefficient (MCC) value of 0.989. The best model was integrated to the MeerLICHT transient vetting pipeline, enabling the accurate and efficient classification of detected transients that allows researchers to select the most promising candidates for their research goals.

astro-ph.IM↗