Mira stars in LAMOST DR4 catalogue.

This slides is for a presentation of my first undergraduate research program in astronomy -- I used optical spectra from the LAMOST telescope to study Mira stars, a type of old long period variables lying on the tip of asymptotic giant branch (AGB) !

Transcript

1. Mira Variable Stars from LAMOST DR4 Yuhan Yao October, 2017
2. Outline • Introduction: what are Mira stars • Background information • Body • Conclusions • Acknowledgements Spectral: emission line Evolutionary: AGB star Why use LAMOST spectra Why are Mira stars important How I assemble a sample of known Mira stars & Learn from these template spectra How I search for new candidates
3. Mira: late AGB variable star Evolutionary tracks of AGB stars (Marigo et al. 2008): Miras appear in the last evolutionary stage of low- and intermediate-mass stars. C-rich stars (red) are produced by the 3rd dredge up. Between O-rich and C-rich: S-type stars (C/O ~ 1, with prominent ZrO band). Spectra of an MS star (top) and a C star (bottom). Asymptotic Giant Branch Stars, Harm J. Habing
4. Mira: emission line spectra C-rich Miras: Balmer decrement (normal case). O-rich Miras: Balmer increment (i.e., FHδ >FHγ >FHβ >FHα ) Spectra vary with luminosity phase. Stellar Spectral Classification, Gray, R.O. Emission lines are excited by shocks. -- For O-rich Miras -- Balmer emissions: strongest at maximum, weakest after minimum luminosity. Metallic emissions: not in the same phase with Balmer lines.
5. Useful tracers! But problems… P-L relation in K band for LMC Miras. (Whitelock et al. 2008) -- Distances can be obtained. -- As tracers of the Milky Way & beyond: Kinematics from Maser, Metallicity from chemical abundance, Bright, be detected by Gaia. Period >80 days, large luminosity amplitudes Problems/Cons of current studies: à Long-term and infrared observations are required. à Spectroscopy has focused on only well-known targets. Have we understood all of Mira’s optical spectral characteristics? LAMOST can give us: à A chance to perform a comprehensive study of Mira variables. à Identification of more Mira candidates based on spectra.
6. Miras’ spectra in LAMOST To obtained a photometrically confirmed sample of Mira spectra from other catalogs... Status of current surveys (Noriyuki Matsunaga, 2016): Green: 783 KISOGP Miras; Red: ~15000 AAVSO Miras (~50% errors in distance); Cross-match with LAMOST: AAVSO (American Association of Variable Star Observers) à 238 KISOGP (Kiso Wide-Field Survey of the Galactic Plane) à 19 SIMBAD Database à another 34 In total: 291 sample spectra (258 O-rich, 33 C-rich)
7. Miras’ spectra in LAMOST From bottom to top: the M → MS → S → SC → C spectral sequence of Mira spectra. Solid lines -- emission lines (yellow: Balmer series; purple: NaI D doublets). Dashed lines -- molecular absorption bands (red: TiO; green: ZrO; cyan: C2; orange: CN). This sample covers the entire the evolutionary sequence of Mira variables.
8. O-rich Normal M giant templates, From Zhong et al. 2015 (M0-M6) & Flux 1994 (M7-M10). Mira spectra, From 258 O-rich sample.
9. O-rich Early-type spectra are very rare, But sth. new: FHδ < FHγ < FHβ ! Late-type spectra, Consistent with precious studies. Median of the Hδ/Hγ ratio for each subtype. Yellow histogram: number of spectra used. Why? – overlying absorption! As the temperature cools down from M0 to M10, the increasing TiO absorptions above the deepest shock- emitting regions strongly weaken the strength of lower-order Balmer lines.
10. O-rich Generally weaker Ca II λλ8498,8542,8662 in Miras. CaII λ8662: a fluorescence with Hε as the pump through CaII H line, Weak absorption in all of the triplets: Photoionization from Lyα photons.
11. O-rich 20 spectra have visible metallic emissions. Fe I λλ4202, 4308, 4376 and Mg I 4571 can be observed – only in late-type Miras !
12. Emission strength - phase Phase calculated from light curves (for KISOGP targets) or epoch & period (AAVSO targets). Large scatters due to: difference between individual stars, errors from AAVSO observations… But we still see a strong correlation ! The rise duration and maximum phase of Hδ and Hγ are not exactly the same: need more observation to be accurately determined. Blue: O-rich Red: C-rich
13. Pre-Selection Selection of Candidates Eye- Inspection 7,681,185 spectra à SNRi > 20 à Header subclass : M/Non/Carbon à + Carbon N-type stars from Ji et al. 2006 & Li et al. 2017 à 2MASS color criterion à 30,738 O-rich, C-rich candidates Giants 2MASS color criterion. Red: AAVSO Miras. Yellow: accepted dr4 targets. Gray: unaccepted ones. Steps Among O-rich candidates: leave only M-giants à 21,989 O-rich candidates. Emission line spectra Leave only spectra with metallic or hydrogen emissions à 951 O-rich, 263 C-rich candidates. Use our 291 sample as spectral templates for reference à 215 Mira spectra (194 targets) !
14. Selection of O-rich Giants TiO5 vs. CaH2+CaH3 plane. Red dots: 291 Mira sample Cyan: M dwarfs from Guo et al. 2015 Blue: M giants from Zhong et al. 2015 Cuts: Use our sample to find a criterion... And apply it to our candidates. Gray: accepted candidates, Cyan: rejected candidates. Magenta: separate them into early & late type.
15. Selection of Emission line spectra Fe/Mg: measure late-type O-rich candidates. Balmer emissions. C-rich: adopt strongest EW of Hα and Hβ O-rich: adopt strongest EW of Hγ and Hδ. Selected Spectra.
16. Conclusion • For O-rich Miras, Hδ/Hγ goes up as the star cools, which is likely driven by TiO absorption, and supports the theory of Joy et al. (1951). • Metallic lines (FeI and MgI) are mostly observed in late type stars. • The relation between Balmer emission strength and the pulsation phase is a population property. • 191 Mira candidates are newly selected from LAMOST DR4, photometry are required to confirm that they are real Mira stars. • 8 of the 111 O-rich candidates are of type M0-M2, which are rare cases. This work is published in ApJS.
17. Acknowledgement • This work is supported by the Hui-Chun Chin and Tsung-Dao Lee Chinese Undergraduate Research Endowment (CURE). • I want to express my deepest gratitude to my inspiring advisors, professor Richard de Grijs, Licai Deng and Chao Liu, who identified me this project, and make me determined to do good astronomical science in the future. • I would also acknowledge Noriyuki Matsunaga (The University of Tokyo) for his guidance throughout this research. • Special thanks to the whole star group in NAOC, and the star cluster group in KIAA-PKU. THANKS FOR YOUR LISTENING