Detection of metal absorption lines in quasar spectra: a neural network approach using U-Net

概要

Current and future large spectroscopic surveys are significantly enhancing the volume and resolution of quasar spectra that are observed, which requires the creation of efficient and precise automated techniques to detect absorption features. This study focuses on the detection of metal absorption features using a novel U-Net model on WEAVE-like mock spectra in the quasar rest-frame wavelength interval $1230\,\mathring{\mathrm{A}} \leq λ_{\rm RF} \leq 3095\,\mathring{\mathrm{A}}$. We test the ne…