Radial Velocities of T Dwarfs with FIRE Christine Nicholls, UCSD Adam Burgasser (USCD), John Bochanski (Penn State), Jackie Faherty (Chile), Rob Simcoe (MIT), Ben Burningham (Hertfordshire), Scott Sheppard (Carnegie)
Eccentric Ellipsoidal Red Giant Binaries periastron periastron are distorted close binaries. Photometric variations are caused by 10% the changing apparent surface area we observe as the pear-shaped primary of ellipsoidal variables have non-sinusoidal light curves indicative of eccentric orbits1. of the already ellipsoidal red giant. We measured radial velocity variations for a sample of ellipsoidals with nonsinusoidal light curves. Orbital Light maxima occur when we observe the broad sides of the ellipsoidal star, and light minima when we observe its narrower ends. Orbital motion dominates the radial velocity curve, so we can identify an ellipsoidal variable by its modelling with the Wilson-Devinney code3 confirms they are in eccentric orbits4. 2 light cycles per velocity cycle. OGLE I band light curve and our radial velocity curve (red points); orbital fits (blue curves).vertical black bar marks phase of periastron. We study ellipsoidal variables containing red giant stars with unseen main sequence companions. Their periods are typically hundreds of days long2. was selected from the OGLE LMC survey, on the basis of light curve shape. Because we know the distance to the LMC, and can find an ellipsoidal variable s orbital inclination via modelling, we can derive absolute masses and radii of these stars. This is possible because the amplitude of the photometric variation provides information on the fraction of its Roche lobe the red giant fills, and the radial velocity amplitude tells us the red giant s orbital semimajor axis5. These non-sinusoidal shapes are caused by INCREASED DISTORTION AT PERIASTRON ECCENTRIC ORBITS orbits its companion. Our sample MAJOR RESULT: Ellipsoidal variables MAJOR RESULT: Christine Nicholls, University of California San Diego cnicholls@physics.ucsd.edu But how is the distinctive light curve produced? Star ID Period (d) Eccentricity Inclination ( ) M1 (M ) M2 (M ) OGLE052013.51 452.47 0.42 60 1.41 1.11 66.29 OGLE052438.40 410.96 0.14 70 6.06 5.25 133.85 OGLE052812.41 258.70 0.24 90 1.39 0.89 55.64 OGLE052850.12 662.20 0.25 50 4.21 3.40 175.93 OGLE053033.55 390.17 0.21 60 5.72 3.20 122.76 OGLE053124.49 541.32 0.29 70 1.97 1.02 92.40 OGLE053159.96 501.10 0.39 60 4.79 2.10 124.72 The distinctive light curve shapes of the eccentric variables feature unequal maxima and minima, the brightest and dimmest of which are always close to periastron. This means the red giant s ellipsoidal distortion is enhanced during periastron passage, changing the apparent surface area we observe in that part of the star s orbit. periastron The WilsonDevinney code can produce pictures of each system throughout its orbit. Each red giant shows a significant increase in its ellipsoidal distortion apastron near periastron. Integrating over the apparent surface area of these images confirms that the apparent surface area changes almost completely account for the detected light variation. Evolved binaries should be in circular orbits, according to tidal theory6. Observations tell a different story. Although the current sample have higher eccentricities than most ellipsoidal variables, their eccentricity is comparable to some post-agb and M giant binaries in the literature. The question of what causes eccentric orbits in evolved binaries remains to be solved. Large Magellanic Cloud (Daniel Verschatse) [1] Soszynski I, et al. 2004, Acta Astronomica 54, 347 [2] Nicholls CP, et al. 2010, MNRAS 405, 1770 R1 (R ) [3] Wilson RE & Devinney EJ, 1971, ApJ 166, 605 [4] Nicholls CP & Wood PR, 2012, MNRAS 421, 2616 [5] WIlson RE, et al. 2009, ApJ 702, 403 [6] Zahn JP, 1977, A&A 57, 383
motivation Characterise T dwarfs as a population Medium-res spectroscopy 3D kinematics, ages, atmospheric properties Probe close T binary regime
FIRE Folded-port InfraRed Echellette Magellan Baade telescope, Las Campanas Echelle mode + 0.6 slit gives R~6000 0.82-2.51 μm Carnegie
radial velocities Cross correlation Spectra have most structure on red side standard of J peak => lowest errors xcor to RV standards and models model
radial velocities Typical precision 2 km/s for xcors to standards standards 1 km/s for xcors to models Standards or models better? models
model fits...
model fits...
T dwarf RV variables! late T with high amplitude RV variation tight T spectroscopic binary!
binary!(?)
binaries are cool because... components have common age RV technique probes close binaries: higher probability of eclipses mass/radius estimates => constrain evolution, atmospheric models
3D kinematics
summary/future work ongoing project characterise T dwarfs by kinematics, atmospheric properties describe T dwarf binaries watch this space!