NMMA-Astro-COLIBRI: An Automated Light-Curve Supernovae Classification Service in the Multi-Survey Era

Abstract

The surge in publicly available photometric alerts from wide-field surveys requires automated tools for real-time transient classification. We present NMMA--Astro-COLIBRI, an on-demand Bayesian classification service that couples the Nuclear-physics and Multi-Messenger Astrophysics (NMMA) inference framework to the Astro-COLIBRI real-time multi-messenger platform. After the detection of an optical transient, if photometry is available, it is quality-filtered. The filtered photometry is fitted by nested sampling against a user-selected model from a library of eleven supernova templates; results are delivered to every user within minutes. Applying two or more models on the same optical transient, the service reports the corresponding log Bayes factors as a quantitative ranking of competing subtypes. We demonstrate the workflow on SN 2021ugl (ZTF21abotose), a Type IIb supernova initially mistaken for a kilonova candidate by automated real-time pipelines, comparing competing supernova and kilonova models. In an early-time configuration using only the first ~ 6 days of photometry in two bands (ZTF g and r), so ten days before spectroscopic confirmation, the empirical Type IIb template recovers the correct classification, favored over both the kilonova template and the kilonova-mimicking shock-cooling model. In the full 47-day, three-band baseline, it again achieves the highest evidence over every competing supernova and kilonova template. These results highlight the importance of a comprehensive supernova template library for kilonova discrimination in the multi-survey era.