ECLIPSING BINARIES IN OPEN CLUSTERS John Southworth Dr Pierre Maxted Dr Barry Smalley Astrophysics Group Keele University
ECLIPSING BINARIESIN OPEN CLUSTERS
John SouthworthDr Pierre MaxtedDr Barry Smalley
Astrophysics Group Keele University
• EBs are good tests of theoretical stellar models– EBs in clusters have known age and metal abundance– EBs in clusters are even better tests of theoretical models
• EBs are good distance indicators– Find distance to cluster
without using MS fitting
Eclipsing binaries in open clusters
• Two EBs in one cluster:– four stars with same age
and chemical composition– excellent test of models– find metal and helium
abundance of cluster– 2004, MNRAS, 349, 547
HD 23642 in the Pleiades
• AO Vp (Si) + Am • Period 2.46 days • mV = 5.9 mag• Shallow eclipses discovered by Torres (2003)• Munari et al (2004) distance: 131.9 ± 2.1 pc
Distance to the Pleiades
• Possible solution: Pleiades is metal-poor– Castellani et al. (2002): Fit for Z = 0.012– But Boesgaard & Friel (1990): [Fe/H] = -0.03 ± 0.02
• Possible solution: Hipparcos parallaxes correlated– (Pinsonneault et al. 1998; Makarov 2002)
• ‘Long’ distance scale: 132 ± 3 pc– MS fitting (e.g., Percival et al. 2003)– HD 23642 (Munari et al. 2004) – Interferometric binary Atlas (Zwahlen et al. 2004)
• `Short’ distance scale: 120 ± 3 pc– Hipparcos (van Leeuwen et al. 2004)
HD 23642 light curves
• B and V light curves from Munari et al. (2004)– We analysed them using EBOP– Theoretical limb darkening and gravity darkening– Formal errors very optimistic
Monte Carlo analysis• Used Monte Carlo
simulations to find light curve uncertainties– Limb darkening
coefficients perturbed– rA = 0.151 ± 0.004
rB = 0.136 ± 0.007
• Problem: B and V solutions inaccurate and don’t agree well– Solution: spectroscopic
light ratio (Torres 2003)– rA = 0.154 ± 0.002
rB = 0.130 ± 0.004
Monte Carlo analysis results for HD 23642 without spectroscopic
light ratio
HD 23642 effective temperatures
• Compare observations to ATLAS9 spectra:– Temperatures: 9750 ± 250 K 7600 ± 400 K
• uvbyβ photometry + Moon & Dworetsky (1985) calibration:– 9200 K for system 9870 K for primary only
• Infrared Flux Method: 9620 ± 280 K 7510 ± 430 K
Pleiades is not metal-poor• HD 23642:
– MA = 2.19 ± 0.02
– MB = 1.55 ± 0.02
– RA = 1.83 ± 0.03
– RB = 1.55 ± 0.04
• Compare to Granada models:– Z ≈ 0.02
– Pleiades distance scales cannot be reconciled with low metal abundance
Granada theoretical models125 Myr Z = 0.01 0.02 0.03
Distance to the Pleiades• Distance from luminosity + bolometric correction:
– L = 4 π R2 σ Teff4 Mbol
– Mbol + BC + V MV + V distance
• Problems:– BCs depend on theoretical model atmospheres
– Fundamental effective temperatures are needed
– Consistent solar Mbol and luminosity values needed• Girardi et al. (2000) BCs: (V filter): 139.8 ± 5.3 pc
(K filter): 138.8 ± 3.3 pc– Bessell et al. (1998) BCs give same results– BCs better in the infrared: reddening less important
metallicity less important
BCs less dependent on Teff
Distance from surface brightness• Calibrations of surface brightness vs. colour index
– SV = surface brightness in V filter
– Φ = angular diameter (mas)
– R = linear radius of star (R)
– SV = mV - 5 log Φ
– distance = 9.3048 (R / Φ) parsecs • Distance to HD23642: 138 ± 19 pc
– Use Di Benedetto (1998) calibration of SV against (B - V)
• Problems:– HD 23642 B and V light ratios are inaccurate– B filter is sensitive to metallicity– (B - V) is not very sensitive to surface brightness– Reddening is important
Surface brightness from temperature• Use zeroth-magnitude angular diameter Φ(m=0)
– SV = V0 - 5 log Φ so Φ(m=0) = Φ 10(0.2 m) = 0.2 SV
– Kervella et al (2004) give Φ(m=0) - log Teff calibrations
• Use 2MASS JHK photometry: IR relations better– Distance : 139.1 ± 3.5 pc– Individual uncertainties:
• Effective temperatures: 0.7 pc 1.4 pc• Stellar radii: 1.4 pc 1.5 pc• Apparent K magnitude: 1.9 pc• `Cosmic’ scatter in calibration: 1.4 pc
The Pleiades distance is ....? • Long distance scale: 132 ± 3 pc
– main sequence fitting– study of astrometric binary Atlas
• Short distance scale: 120 ± 3 pc– Hipparcos parallaxes
• Distance to HD 23642: 139 ± 4 pc– only weakly dependent on temperatures and radii
• The Pleiades is not metal-poor– from comparison between the masses and radii
and theoretical evolutionary models
Southworth, Maxted & Smalley, astro-ph/0409507
W W Aurigae
• A4 m + A5 m• Period 2.52 days • mV = 5.9 mag• Discovered by Solviev (1918) and Schwab (1918)• Hipparcos distance: 84.3 ± 7.3 pc
WW Aur spectral characteristics
• Both components are Am stars– spectra show strong lines of both components
WW Aur spectroscopic orbit• TODCOR: two-dimensional cross-correlation
– Cross-correlate against many observed template spectra– Fit spectroscopic orbits using SBOP– Choose which sets of spectra give good orbits– Average good orbits to find best orbit
• RV semiamplitudes:KA = 116.81 ± 0.23 km/sKB = 126.49 ± 0.32 km/s
– Uncertainty is standard deviation of the results from each good orbit
– SBOP uncertainties agree very well
WW Aur light curves 1
• UBV light curves from Kiyokawa & Kitamura (1975)– 3037 datapoints scanned from paper
WW Aur light curves 2
• uvby light curves from Etzel (1975) Master’s Thesis– 3748 datapoints on a nine-track magnetic tape
WW Aur light curve analysis• UBV and uvby light curves fitted using EBOP
– Limb darkening coefficients adjusted
• Uncertainties from Monte Carlo analysis– Good agreement with
variation between the seven light curves:
– rA = 0.1586 ± 0.0009
– rB = 0.1515 ± 0.0009
HD 23642
WW Aur
WW Aur effective temperatures
• Am stars so spectral analysis unreliable• Hipparcos parallax gives distance 84.3 ± 7.3 pc• Get bolometric flux
– UV fluxes from TD-1 satellite– UBVRI magnitudes– 2MASS JHK magnitudes
• Convert to separate fluxes using V light ratio • Temperatures:
– Teff (A) = 7960 ± 420 K
– Teff (B) = 7670 ± 410 K
– almost no dependence on model atmospheres
WW Aur results
• Masses from cross-correlation against observed spectra: MA = 1.964 ± 0.007 M
MB = 1.814 ± 0.007 M
• Radii from EBOP geometrical analysis:– Gravity darkening unimportant
– Limb darkening fitted RA = 1.927 ± 0.011 R
RB = 1.841 ± 0.011 R
• Effective temperatures from Hipparcos parallax and UV-optical-IR fluxes: Teff (A) = 7960 ± 420 K
Teff (B) = 7670 ± 410 K
• Assume common age and chemical composition for both stars in WW Aur
• Problem: no published theoretical stellar models fit the masses and the radii
Comparison with theoretical models
Solution: Z = 0.06
Claret (2004) models fit for Z = 0.06 age 77107 Myr
Metallic-lined eclipsing binaries
Conclusions• EBs are excellent distance indicators
– HD 23642 gives Pleiades distance 139 ± 4 pc– Agrees with MS fitting but not Hipparcos
• Distance from surface brightness is good– Avoids bolometric corrections from model atmospheres
– Best in the infrared (reddening, Teff dependence)
• Eclipsing binaries in open clusters are very useful
• WW Aur seems to be very metal-rich– Masses and radii found to accuracies of 0.4%, 0.6%
– Teff s from Hipparcos parallax and UV-optical-IR fluxes
– Metal abundance of Z ≈ 0.06 not connected to Am spectra
John Southworth ([email protected]) Keele University, UK