| Name: Rajarshi Barman |
| Affiliation: Tata Institute of Fundamental Research (TIFR), Mumbai |
| Conference ID: ASI2026_487 |
| Title: Potential of Gaia XP Spectra in Red Giant Asteroseismology: A Deep Learning Approach |
| Abstract Type: Poster |
| Abstract Category: Stars, Interstellar Medium, and Astrochemistry in Milky Way |
| Author(s) and Co-Author(s) with Affiliation: Rajarshi Barman(Indian Institute of Science, Bengaluru (1) ; Tata Institute of Fundamental Research, Mumbai (2)), Shatanik Bhattacharya(Tata Institute of Fundamental Research, Mumbai), Shravan Hanasoge(Tata Institute of Fundamental Research, Mumbai), Siddharth Dhanpal(Tata Institute of Fundamental Research, Mumbai) |
| Abstract: Red giants are key tracers of stellar evolution and Galactic structure, and their asteroseismic properties, particularly the large frequency separation (∆ν), the frequency of maximum oscillation power (ν_max), and the dipole-mode period spacing (∆Π_1), provide direct insight into their internal structure, masses, and evolutionary states. Until now, seismic inference for large stellar samples has relied primarily on high-quality photometric light curves from missions such as Kepler and TESS, or on moderate-resolution spectroscopy from surveys such as LAMOST (R ∼ 1,800) and APOGEE (R ∼ 22,500) that preserves information correlated with these seismic quantities. With Gaia XP spectrophotometry (R ∼ 15–85), the possibility arises to extend asteroseismic measurements to orders of magnitude more stars despite the much lower spectral resolution. In this work, we assess whether XP spectra retain sufficient information to enable reliable seismic inference for red giants. We develop hybrid convolutional neural network–long short-term memory (CNN–LSTM) models trained on red giants with seismic parameters measured from Kepler photometry. The networks learn subtle spectral signatures, imprinted through global stellar properties, that correlate with ∆ν, ν_max, and ∆Π_1. We find that all three global asteroseismic parameters can be recovered from Gaia XP spectra with accuracies comparable to those obtained from moderate-resolution spectroscopic surveys, demonstrating that even low-resolution spectrophotometry contains sufficient information for seismic prediction. Saliency analysis reveals wavelength regions most strongly associated with seismic sensitivity and highlights physically distinct spectral behaviour between red giant branch and red clump stars. Applying our models to Gaia DR3, we obtain seismic predictions for more than 2.5 million bright red giants, enabling population-level asteroseismic studies on an unprecedented scale, and identify a small subset of low-∆ν red clump candidates with unusual spectral–seismic correlations that may offer new insights into evolved stellar populations. |
