Detection of metal absorption lines in quasar spectra: a neural network approach using U-Net
Abstract
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 mathringA ≤ λRF ≤ 3095 mathringA. We test the network performance for absorption detection both on ideal data and after simulating the continuum fitting step as applied on real data. The performance of these architectures is evaluated by the completeness, purity, and F1 score reached in bins of signal-to-noise ($\mathrm{S}/ \mathrm{N}_\mathrm{line} $) for the absorption lines and with the absolute fractional flux error for the continuum. The ability to recover the correct line centers is also studied. The U-Net reaches scores of ≈ 90% for all metrics (completeness, purity, and F1 score) at S/ Nline ≈ 4. All false positive detections with S/ Nline ≥ 5 fall in the tails of the broad Lyα absorbers distribution of damped Lyα systems. The combination of continuum fitting and line detections has negligible effects on the detection performance at S/ Nline ≥ 4. Our proposed U-Net architecture offers a competitive tool for the analysis of absorption lines in current and upcoming large spectroscopic surveys and is well-suited to the identification of any absorption line feature.
Related
- Radiation pressure powers quasar broad absorption line winds but fails to drive galaxy feedback
- The Wavelet Detection Filter: a real time unmodeled pipeline for gravitational wave transients, ranking coincidences with a graph neural network
- Detection of episodic Ultra-fast Outflows in the Low-luminosity quasar Mrk 205 using XMM-Newton data
- First-principles calculations of far-wing absorption in He-broadened CO2 infrared spectra
- Probing multi-state dark matter via optical absorption lines in DESI spectra
BibTeX
@misc{mangola2026detection,
title = {{Detection of metal absorption lines in quasar spectra: a neural network approach using U-Net}},
author = {Elena Sofia Mangola and Francesco Pistis and Michele Fumagalli and Matteo Fossati and Ting-Yun Cheng and Ryan J. Cooke and Rajeshwari Dutta and Ignasi Pérez-Ràfols and Matthew Pieri and Emanuel Gafton},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.28072},
eprint = {2609.28072},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.28072},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Detection of metal absorption lines in quasar spectra: a neural network approach using U-Net
AU - Elena Sofia Mangola
AU - Francesco Pistis
AU - Michele Fumagalli
AU - Matteo Fossati
AU - Ting-Yun Cheng
AU - Ryan J. Cooke
AU - Rajeshwari Dutta
AU - Ignasi Pérez-Ràfols
AU - Matthew Pieri
AU - Emanuel Gafton
PY - 2026
DA - 2026/09/23
PB - arXiv
DO - 10.48550/arxiv.2609.28072
UR - https://arxiv.org/abs/2609.28072
AB - 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 mathringA ≤ λ_RF ≤ 3095 mathringA. We test the network performance for absorption detection both on ideal data and after simulating the continuum fitting step as applied on real data. The performance of these architectures is evaluated by the completeness, purity, and F1 score reached in bins of signal-to-noise ($\mathrm{S}/ \mathrm{N}_\mathrm{line} $) for the absorption lines and with the absolute fractional flux error for the continuum. The ability to recover the correct line centers is also studied. The U-Net reaches scores of ≈ 90% for all metrics (completeness, purity, and F1 score) at S/ N_line ≈ 4. All false positive detections with S/ N_line ≥ 5 fall in the tails of the broad Lyα absorbers distribution of damped Lyα systems. The combination of continuum fitting and line detections has negligible effects on the detection performance at S/ N_line ≥ 4. Our proposed U-Net architecture offers a competitive tool for the analysis of absorption lines in current and upcoming large spectroscopic surveys and is well-suited to the identification of any absorption line feature.
ER -