Period-luminosity relations in evolved red giants explained by solar-like oscillations

Abstract

Context. Solar-like oscillations in red giants have been investigated with the space-borne missions CoRoT and Kepler, while pulsations in more evolved M giants have been studied with ground-based microlensing surveys. After 3.1 years of observation with Kepler, it is now possible to link these different observations of semi-regular variables. Aims. We aim to identify period-luminosity sequences in evolved red giants identified as semi-regular variables and to interpret them in terms of solar-like oscillations. Then, we investigate the consequences of the comparison of ground-based and space-borne observations. Methods. We first measured global oscillation parameters of evolved red giants observed with Kepler with the envelope autocorrelation function method. We then used an extended form of the universal red giant oscillation pattern, extrapolated to very low frequency, to fully identify their oscillations. The comparison with ground-based results was then used to express the period-luminosity relation as a relation between the large frequency separation and the stellar luminosity. Results. From the link between red giant oscillations observed by Kepler and period-luminosity sequences, we have identified these relations in evolved red giants as radial and non-radial solar-like oscillations. We were able to expand scaling relations at very low frequency (periods as long as 100 days and large frequency separation less than 0.05 μHz). This helped us identify the different sequences of period-luminosity relations, and allowed us to propose a calibration of the K magnitude with the observed large frequency separation. Conclusions. Interpreting period-luminosity relations in red giants in terms of solar-like oscillations allows us to investigate the time series obtained from ground-based microlensing surveys with a firm physical basis. This can be done with an analytical expression that describes the low-frequency oscillation spectra. The different behavior of oscillations at low frequency, with frequency separations scaling only approximately with the square root of the mean stellar density, can be used to precisely address the physics of the semi-regular variables. This will allow improved distance measurements and opens the way to extragalactic asteroseismology with the observations of M giants in the Magellanic Clouds.

The variability behaviour of CoRoT M-giant stars

“…Afterwards, Mosser et al (2013) studied global oscillation parameters of Kepler red giants, concluding that the main excitation mechanism in M giant SR variables are solar-like oscillations, thus confirming the findings of Dziembowski & Soszyński (2010), but they were unable to disentangle RGB from AGB stars. In this regard, Stello et al (2014) detected non-radial modes in early M giants, as previously found by Mosser et al (2013) for less luminous giants. Bányai et al (2013) performed a detailed study of the variability of M-type giants using Kepler data as well.…”

The connection between stellar granulation and oscillation as seen by theKeplermission

“…Compared to the previous analysis of Mathur et al (2011) we extended the sample towards sub-giant branch and main-sequence stars and used time series for the red giants that are about three times as long. On the other hand, M giants that show solar-like oscillations (Mosser et al 2013b) are excluded from our analysis as their granulation signal has such long periods that we cannot derive precise parameters even with the long Kepler time series. The total sample now covers more than 3 order of magnitudes in ν max .…”

Detection of solar-like oscillations in relics of the Milky Way: asteroseismology of K giants in M4 using data from the NASA K2 mission

“…The asteroseismic analysis code was then used to extract from the detected oscillation spectra estimates of two commonly used global or average asteroseismic parameters: ∆ν, the average frequency separation between consecutive overtones of modes having the same angular degree; and ν max , the frequency at which the oscillations present their strongest observed amplitudes (see Elsworth et al, in preparation for details). To compare the detected power with expectations we used the relations in Mosser et al (2012) and Kallinger et al (2014) to describe the power envelope due to the oscillations and the power spectrum of the granulation, to which we then added the contribution due to shot noise estimated from the mean power close to the Nyquist frequency (see Fig. 2).…”