Cepheids and the Extragalactic Distance Scale Vicky Scowcroft Carnegie Observatories DAO Seminar 8th June 2010 Overview DAO Seminar 8th June 2010 Overview • A brief history of Cepheids. DAO Seminar 8th June 2010 Overview • A brief history of Cepheids. • Cepheids as part of the distance scale. DAO Seminar 8th June 2010 Overview • A brief history of Cepheids. • Cepheids as part of the distance scale. • Current problems ☹ DAO Seminar 8th June 2010 Overview • A brief history of Cepheids. • Cepheids as part of the distance scale. • Current problems ☹ • Cepheid populations: • Optical - M33 • Infra-red - LMC DAO Seminar 8th June 2010 Overview • A brief history of Cepheids. • Cepheids as part of the distance scale. • Current problems ☹ • Cepheid populations: • Optical - M33 • Infra-red - LMC • Interesting finds ☺ DAO Seminar 8th June 2010 The discovery of Cepheid variables John Goodricke • Variability of Eta Aquilae discovered by Edward Pigott in 1784. • John Goodricke and EP later discovered variability of Delta Cephei. • Calculated period as 5.14 days - exceptionally close to today’s value (5.37 days) - time is easy to measure accurately☺ DAO Seminar 8th June 2010 The discovery of Cepheid variables John Goodricke • Variability of Eta Aquilae discovered by Edward Pigott in 1784. • John Goodricke and EP later discovered variability of Delta Cephei. • Calculated period as 5.14 days - exceptionally close to Goodricke’s gruelling today’s value (5.37 days) - observing schedule for time is easy to measure Delta Cep ended in accurately☺ pneumonia and death! DAO Seminar 8th June 2010 What is a Cepheid? • Young stars • M = few M⊙ 3 4 • L = 10 to 10 L⊙ • Live on the instability strip: • MW, LMC, SMC • Unstable to radial pulsations DAO Seminar 8th June 2010 Why do Cepheids pulsate? DAO Seminar 8th June 2010 Why do Cepheids pulsate? L He+ → He2+ DAO Seminar 8th June 2010 Why do Cepheids pulsate? κ↑, T↑, P↑, R↑ L L+ He+ → He2+ He2+ → He+ DAO Seminar 8th June 2010 Why do Cepheids pulsate? κ↑, T↑, P↑, R↑ T↓, κ↓, P↓, R ↓ L L+ L- He+ → He2+ He+ → He2+ He2+ → He+ DAO Seminar 8th June 2010 Why do Cepheids pulsate? κ↑, T↑, P↑, R↑ T↓, κ↓, P↓, R ↓ L L+ L- He+ → He2+ He+ → He2+ He2+ → He+ Single mode pulsation: 1 2 2R P assume HSE 4πR3 / Π= ∝ Π= v vs s ! ρ !3GM " DAO Seminar 8th June 2010 Why do Cepheids pulsate? κ↑, T↑, P↑, R↑ T↓, κ↓, P↓, R ↓ L L+ L- He+ → He2+ He+ → He2+ He2+ → He+ Single mode pulsation: 1 2 2R P assume HSE 4πR3 / Π= ∝ Π= v vs s ! ρ !3GM " 1 2 2 4 L3/2 / L =4πR σT Π ∝ ! T 6 " DAO Seminar 8th June 2010 Discovery of the PL relation Leavitt & Pickering, 1912 Pickering, & Leavitt Henrietta Leavitt “... the brighter variables have the longer periods...” Strong relationship between period and magnitude at maximum M = a log P + b and minimum DAO Seminar 8th June 2010 Cepheids as standard candles DAO Seminar 8th June 2010 Cepheids as standard candles Take a single population of Cepheids... Observed PL: m = a log P + b b Absolute PL: M = a log P + b0 DAO Seminar 8th June 2010 Cepheids as standard candles Take a single population of Cepheids... Observed PL: m = a log P + b b Absolute PL: M = a log P + b0 Distance can be found easily: μ = m - M = a log P + b - a log P - b0 = b - b0 DAO Seminar 8th June 2010 Cepheids as standard candles Take a single population of Cepheids... Observed PL: m = a log P + b b Absolute PL: M = a log P + b0 Distance can be found easily: Simples? μ = m - M = a log P + b - a log P - b0 = b - b0 DAO Seminar 8th June 2010 Not so simple... DAO Seminar 8th June 2010 Not so simple... M = a log P + b B V I DAO Seminar 8th June 2010 Not so simple... M = a log P + b B V I PL changes with bandpass DAO Seminar 8th June 2010 Not so simple... M = a log P + b B V I Extinction? DAO Seminar 8th June 2010 Not so simple... M = a log P + b Linearity? B V I DAO Seminar 8th June 2010 Not so simple... M = a log P + b Metallicity? B V I DAO Seminar 8th June 2010 Not so simple... M(P, λ,A,Z) = a(P,λ,Z) log P + b(A, λ,Z) B V I DAO Seminar 8th June 2010 Calibration of the PL relation • Reddening is easy(ish) to fix. • Some evidence that the PL changes slope around 10 Ngeow et al. (2009) days, but not conclusive. 0.0 0.5 log P 1.0 1.5 • Limiting samples to either Fouqué et al. (2007) above or below 10 days removes the effect. • Metallicity is a more complicated problem... DAO Seminar 8th June 2010 Why would metallicity matter? DAO Seminar 8th June 2010 Why would metallicity matter? • Compare two Cepheids with equal M, different Z... low Z high Z const. P IS DAO Seminar 8th June 2010 Why would metallicity matter? • Compare two Cepheids with equal M, different Z... low Z • Tracks show that for constant M and P, low Z high Z star will be brighter than high Z star. const. P IS DAO Seminar 8th June 2010 Why would metallicity matter? • Compare two Cepheids with equal M, different Z... low Z • Tracks show that for constant M and P, low Z high Z star will be brighter than high Z star. const. P • But PLC projection IS widens the IS, so does it really matter? DAO Seminar 8th June 2010 What are we measuring? Metallicity effect can be added into the PL relation: M = a log P + b - γΔZ change in measured μ δ (m − M) γ = 0 δlogZ change in metallicity DAO Seminar 8th June 2010 Observational tests Theory says that metallicity does make a difference, but is it observable? DAO Seminar 8th June 2010 Observational tests Freedman & Madore (1990): No (statistically significant) change in μ with Z DAO Seminar 8th June 2010 Observational tests Sasselov et al. (1997): SMC Cepheids too faint when compared with LMC Cepheids. dμ = 0.44 log (Z / ZLMC) DAO Seminar 8th June 2010 Observational tests HST Key Project result: γ = − 0.24 ± 0.16 mag dex-1 Kennicutt et al. (1998) DAO Seminar 8th June 2010 Observational tests Romaniello et al. (2008) DAO Seminar 8th June 2010 Observational tests This is the opposite direction to the theoretical suggestions! DAO Seminar 8th June 2010 M33 to the rescue! Many previous studies have used selections of • ★ Single reddening law for the whole sample galaxies. (?) ★ Range of distances, reddenings and M33 is ideal! metallicities can introduce uncertainties. • ★ Well studied - distance is well known ★ Difficult to ensure you’re always on the independently. same zero-point. ★ Well defined metallicity gradient. • Using a single galaxy gets rid of these issues. ★ Small inclination angle - we can assume all ★ All Cepheids can be assumed to have the Cepheids have the same true value of μ. same μ. DAO Seminar 8th June 2010 M33 to the rescue! Many previous studies have used selections of • ★ Single reddening law for the whole sample galaxies. (?) ★ Range of distances, reddenings and M33 is ideal! metallicities can introduce uncertainties. • ★ Well studied - distance is well known ★ Difficult to ensure you’re always on the independently. same zero-point. ★ Well defined metallicity gradient. • Using a single galaxy gets rid of these issues. ★ Small inclination angle - we can assume all ★ All Cepheids can be assumed to have the Cepheids have the same true value of μ. same μ. DAO Seminar 8th June 2010 First test: WIYN survey of M33 DAO Seminar 8th June 2010 First test: WIYN survey of M33 DAO Seminar 8th June 2010 Observations • Four fields in two regions: ★ 1, 2: <R> = 0.52 kpc, <12 + log O/H> = 8.852 ★ 3, 4: <R> = 5.11 kpc, <12 + log O/H> = 8.296 • Assuming double-component Z gradient from Magrini et al. (2007) • 25 epochs of BVIC observations using the WIYN 3.5 m Imager (first 20 epochs) and Mini-Mosaic (last 5 epochs) cameras. • Overlapping fields of view mean some stars have up to 38 observations. Original copyright NOAO DAO Seminar 8th June 2010 Photometry DAO Seminar 8th June 2010 Photometry • Photometry process automated using DAOPHOT and ALLFRAME. • ALLFRAME used to get consistent object list and deeper detections. • Imager photometry calibrated using Landolt standards, Mini-Mosaic calibrated using secondary standards. • Astrometry used USNO-B for outer field, Massey et al. (2006) catalogues for inner field. • Massey et al. also used for photometry comparison. DAO Seminar 8th June 2010 Photometry comparisons Scowcroft et al., 2009 al., et Scowcroft Inner field Outer field Excellent agreement between this work and Massey et al. (2006). Around 5% of our stars appeared in both catalogues. Average ΔV = −0.009 (inner), −0.014 (outer) DAO Seminar 8th June 2010 Cepheid selection • Cepheids identified using Fourier Parameters (J.B. Marquette for the CFHT M33 survey). • Magnitudes determined using 3rd order Fourier fit to light curves. • Bad Cepheids: weird amplitude ratios, bad data points, incorrect periods. FINAL SAMPLE: 91 inner, 28 outer DAO Seminar 8th June 2010 BVI PL relations PL relations for BVI fit using slope of LMC B V IC • relations in Fouqué et al. (2007), zero-point fit using an iterative weighted-least-squares method. • Differences found in zero-point between inner and outer field for all bands. • Several possibilities: 1. Different reddening/extinction. B V IC 2. Blending affecting inner field. 3. Different P distribution - nonlinearity of PL. 4. Metallicity effects. Scowcroft et al., 2009 DAO Seminar 8th June 2010 Getting rid of reddening Reddening can be removed from the problem using the reddening-free Wesenheit index: Wvi = V - RVI (V - I) RVI = AV / E(V - I) Wvi=[V0 + RVI (V−I)0]+[AV−RVIE(V−I)] DAO Seminar 8th June 2010 Getting rid of reddening Reddening can be removed from the problem using the reddening-free Wesenheit index: Wvi = V - RVI (V - I) RVI = AV / E(V - I) Wvi=[V0 + RVI (V−I)0]+[AV−RVIE(V−I)] DAO Seminar 8th June 2010 Getting rid of reddening Reddening can be removed from the problem using the reddening-free Wesenheit index: Wvi = V - RVI (V - I) RVI = AV / E(V - I) Wvi=[V0 + RVI (V−I)0]+[AV−RVIE(V−I)] All reddening terms cancel, apparent index equal to absolute index ➔ reddening free! DAO Seminar 8th June 2010 Reddening-free PL relations Scowcroft et al., 2009 DAO Seminar 8th June 2010 Reddening-free PL relations • Reddening-free PL relations still show an offset ➔ different zero- points can’t be caused be differences in the reddening.