Mode Angular Degree Identification in Subgiant Stars with Convolutional Neural Networks based on Power Spectrum

TL;DR

This paper presents a convolutional neural network approach to accurately identify the angular degrees of oscillation modes in subgiant stars, aiding asteroseismology analysis.

Contribution

It introduces a CNN-based method trained on simulated data and fine-tuned on real data for mode identification in subgiant stars, improving accuracy.

Findings

Achieved 95% accuracy on Kepler data.

Successfully distinguished radial and non-radial modes.

Enhanced peakbagging analysis in asteroseismology.

Abstract

Identifying the angular degrees $l$ of oscillation modes is essential for asteroseismology and depends on visual tagging before fitting power spectra in a so-called peakbagging analysis. In oscillating subgiants, radial ($l$= 0) mode frequencies distributed linearly in frequency, while non-radial ($l$ >= 1) modes are p-g mixed modes that having a complex distribution in frequency, which increased the difficulty of identifying $l$. In this study, we trained a 1D convolutional neural network to perform this task using smoothed oscillation spectra. By training simulation data and fine-tuning the pre-trained network, we achieved a 95 per cent accuracy on Kepler data.