中文

Blind Line Search System: BLiSS

天体物理仪器与方法 2026-07-08 v1

摘要

The increasing sensitivity and spectral resolution of current and forthcoming X-ray observatories, including \textit{XRISM} and \textit{NewAthena}, are expected to reveal increasing numbers of weak and blended emission lines, motivating reproducible tools for their systematic identification. Existing workflows often rely on manual inspection or source-specific analysis pipelines, making homogeneous analyses of large datasets difficult. To address this need, we present BLiSS (Blind Line Search System), an open-source Python package for the fast, blind detection and characterization of emission-line candidates in one-dimensional X-ray spectra without requiring a prior physical continuum model. BLiSS is intended as an exploratory analysis tool that complements subsequent physical spectral modelling. The package estimates an empirical baseline directly from the observed spectrum, identifies positive excesses, groups them into candidate regions, and characterizes them with Gaussian models. Candidate reliability is estimated by comparison with synthetic spectra using a Gaussian Mixture Model classifier. Finally, optional routines perform a simultaneous multi-Gaussian fit and associate detected features with compatible atomic transitions. The methodology implemented in BLiSS has already enabled published spectroscopic studies and is presented here as a documented, modular, and publicly available software package. Its performance is demonstrated using \textit{Chandra}/HETGS and \textit{XRISM}/Resolve observations of the high-mass X-ray binary Vela X-1, one of the best-studied X-ray sources. BLiSS recovers the principal emission features reported in previous studies while providing a fast, reproducible, and instrument-independent workflow for exploratory line searches.

引用

@article{arxiv.2607.07783,
  title  = {Blind Line Search System: BLiSS},
  author = {Luis Abalo and Graciela Sanjurjo-Ferrín and Jessica Planelles-Villalva and José Joaquín Rodes-Roca and José Miguel Torrejón},
  journal= {arXiv preprint arXiv:2607.07783},
  year   = {2026}
}