X-shooter helps to distinguish red clump stars from red giant branch stars

By Kirsten Banks
[email protected]>

One of the biggest questions in the field of astrophysics is, how do galaxies form and evolve? The ways in which spiral galaxies form and evolve leave imprints in the distribution of stellar ages, kinematics and abundances. Studying those observable properties in spiral galaxies allows scientists to piece together their past and understand their history. By looking at stars in our Milky Way galaxy, we can gain even more insights, examine specific places where different elements are abundant, and study how galaxies change over time. Using stellar standard candles is an effective way to do this, and one such example of standard candles are red clump stars.

Red clump stars are a post-red giant branch phase of core helium burning, and make very effective standard candles. This is because all low-mass (0.8-2.0 solar masses) red giant branch stars ignite helium fusion at a characteristic core mass of 0.47 solar masses, independent of the original stellar mass, thus resulting in very small luminosity differences. However, red clump stars have very similar surface features, such as effective temperature and surface gravity, as red giant branch stars, which are not standard candles (Fig. 1 below). Therefore, separating standard candle red clump stars from non-standard candle red giant branch stars with photometry alone results in significant contamination.

Keil and HR diagrams
Figure 1: Kiel (left) and Hertzsprung-Russell (right) diagrams showing the similarity of surface features between red clump (red circles) and red giant branch (black circles) stars. The grey dots represent the overall distribution of stars from the GALAH DR3 catalogue

Red clump and red giant branch stars can be effectively characterised using asteroseismology, i.e. by studying their unique oscillations as a result of their differing internal structures. However, the study of optical and infrared spectra of red giant stars with data-driven analyses has revealed that this asteroseismic information is imprinted in said spectra. Thus, data-driven methods using red giant spectra can effectively infer the physical properties necessary to distinguish the red clump from the red giant branch when the “gold standard” asteroseismic data are unavailable.

In our study recently published in MNRAS, we expanded on previous investigations of the spectro-seismic connection of red giant stars over a tenfold wavelength range (0.33-2.5 microns) using the ESO VLT/X-Shooter spectrograph. This broader wavelength range allowed us to probe more spectral features to determine which atomic or molecular species held the most significance in the classification of red clump and red giant branch stars (Fig. 2 below). Our analysis of 49 stars with asteroseismic classifications from the K2 mission confirms that CN, CO, and CH features are indeed the primary carriers of spectroscopic information on the evolutionary stages of red giant stars.

CN & CO spectra
Figure 2: Two spectral windows about 0.923 microns and 1.574 microns showing in the top panel the spectra of one red clump (red line) and one red giant branch (black line) star from our dataset with similar surface features. Identified spectral features are also labelled. The bottom panels show the significance each wavelength pixel has in the prediction of red giant evolutionary phase. Pixels with values near the solid line about zero indicate little to no significance, whereas, pixels above or below the top and bottom dashed lines about ~0.033 (and highlighted in red) indicate the greatest significance in predicting red giant evolutionary state. These pixels correspond primarily with CN, CO and CH molecular features.


Michael Murphy is the Australian representative on the ESO Science Technical Committee. Contact: [email protected]

Sarah Sweet is the Australian representative on the ESO Users Committee. Contact: [email protected]

Stuart Ryder is a Program Manager with AAL. Contact: [email protected]

Guest posts are also welcome – please submit these to [email protected]