The VMC Survey

Abstract

Context. Variability is a key property of stars on the asymptotic giant branch (AGB). Their pulsation period is related to the luminosity and mass-loss rate of the star. Long-period variables (LPVs) and Mira variables are the most prominent of all types of variability of evolved stars. The reddest, most obscured AGB stars are too faint in the optical and have eluded large variability surveys. Aims. We obtained a sample of LPVs by analysing K-band light curves (LCs) of a large number of sources in the direction of the Magellanic Clouds with the colours expected for red AGB stars ((J − K) > 3 mag or equivalent in other colour combinations). Methods. Selection criteria were derived based on colour-colour and colour-magnitude diagrams from the combination of the VISTA Magellanic Cloud (VMC) survey, Spitzer IRAC and AllWISE data. After eliminating LPVs with known periods shorter than 450 days, a sample of 1299 candidate obscured AGB stars was selected. K-band LCs were constructed by combining the epoch photometry available in the VMC survey with literature data, were analysed for variability, and fitted with a single period sine curve to derive mean magnitudes, amplitudes, and periods. A subset of 254 stars are either new variables, known variables where the period we find is better determined than the literature value, or variables with periods longer than 1000 days. The spectral energy distributions (SEDs) of these stars were fitted to a large number of templates. For this purpose the SEDs and Spitzer IRS spectra of some non-AGB stars (Be stars, HII regions and young stellar objects – YSOs) were also fitted to have templates of the most likely contaminants in the sample. Results. A sample of 217 likely LPVs is found. Thirty-four stars have periods longer than 1000 days, although some of them have alternative shorter periods. The longest period of a known Mira in the Magellanic Clouds from Optical Gravitational Lensing Experiment data (with P = 1810 d) is derived to have a period of 2075 d based on its infrared LC. Two stars are found to have longer periods, but both have lower luminosities and smaller pulsation amplitudes than expected for Miras. Mass-loss rates and luminosities are estimated from the template fitting. Period-luminosity relations are presented for carbon (C-) and oxygen (O-) rich Miras that appear to be extensions of relations derived in the literature for shorter periods. The fit for the C stars is particularly well defined (with 182 objects) and reads Mbol = (−2.27 ± 0.20) ⋅ log P + (1.45 ± 0.54) mag with an rms of 0.41 mag. Thirty-four stars show pulsation properties typical of Miras while the SEDs indicate that they are not. Overall, the results of the LC fitting are presented for over 200 stars that are associated with YSOs.

New results for FBS late-type stars using Gaia EDR3 data

“…Figure 5 nicely confirms the ability of the Gaia-2MASS-diagram to distinguish between M- and C-stars, although there are two objects which, according to our low-resolution spectra, show a spectral type different from the one assumed by its location in that diagram. This is in agreement with findings by Groenewegen et al (2020), who detected a few high-mass-loss M-stars in the extreme C-stars region as well. The mass-loss leads to a circumstellar shell, so that M- and C-stars cannot be distinguished reliably.…”

Asteroseismology of luminous red giants with Kepler. II. Dependence of mass loss on pulsations and radiation

“…A strong positive correlation between the mass-loss rate and luminosity in long-period AGB stars ( 300 days) has been recently reported by Danilovich et al (2015) and Groenewegen et al (2020). This suggests that the effect of radiation on dust is effective in driving superwinds in these stars.…”

Discovery of a mid-infrared protostellar outburst of exceptional amplitude

“…More importantly, the time domain behaviour (see Figure 1) does not resemble a pulsating AGB star. The longest period Mira variables have periods, P < 6 yr (Groenewegen et al 2020;Navarro Molina et al 2016), typically with smooth, sinusoidal light curves. Unbiased searches of the near infrared variable sky with UKIDSS and VVV find that their peak to trough amplitudes very rarely exceed ∆K s = 4 mag (Contreras Peña et al 2017a;Lucas et al 2017, Lucas et al, in prep.…”