Ages inferred for red giant stars via machine learning are generally insensitive to hyperparameters and architecture but somewhat sensitive to training set choice, especially for the oldest, coolest, and lowest-metallicity stars.
TESS Asteroseismology of Red Giants in the Old Metal-Rich Open Clusters NGC 188 & NGC 6791
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
Open clusters are fundamental laboratories for investigating stellar and Galactic evolution, and serve as important benchmarks for asteroseismic analyses. Using a boutique method to analyze TESS photometry, we study red giants in two old metal-rich open clusters: NGC 188 & NGC 6791. By comparing Kepler and TESS observations for NGC 6791, similar oscillation mode frequencies are recovered, however we find a systematic offset of 2.2% with a scatter of 9% in the $\nu_{\text{max}}$ measurements. We attribute this discrepancy to the lower signal-to-noise of the TESS data for these relatively faint stars. For the brighter cluster NGC 188, we present new seismic measurements in 17 red giants. We estimate average seismic masses for the RGB of $M_{\text{RGB,NGC188}} = 1.13\pm0.04$(rand)$^{+0.12}_{-0.19}$(sys) $M_{\odot}$ and RC of $M_{\text{RC,NGC188}} = 1.11\pm0.01$(rand)$^{+0.11}_{-0.19}$(sys) $M_{\odot}$, consistent with independent mass estimates for this cluster and with similar precision to previous Kepler studies. From the difference between the average evolutionary phase masses, we estimate an integrated RGB mass loss of $\Delta M = 0.02 \pm 0.04$(rand)$\pm0.01$(sys) $M_{\odot}$, supporting the evidence for lower mass loss at higher metallicities. Using asteroseismology and chemical abundances, we identify three binary interaction candidates: two under-massive stars and one over-massive star potentially exhibiting dipole-mode suppression. Finally, we derive an average seismic cluster age of $7.0\pm0.9$ Gyrs, in good agreement with previous literature ages. Our analysis demonstrates the strong potential of TESS asteroseismology for open clusters, and motivates extending this investigation to other TESS clusters that span a wider range of ages and metallicities.
fields
astro-ph.SR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Evaluating the Sensitivity of the Age Inferences of Red Giant Stars to Machine Learning Methodology
Ages inferred for red giant stars via machine learning are generally insensitive to hyperparameters and architecture but somewhat sensitive to training set choice, especially for the oldest, coolest, and lowest-metallicity stars.