Atelier national Services et bases de données en spectroscopie stellaire. Amphithéâtre de l Institut d'astrophysique de Paris.

Transcription

1 Atelier national Services et bases de données en spectroscopie stellaire Amphithéâtre de l Institut d'astrophysique de Paris 8-9 Mars 2016 France Allard Directrice de Recherche (DR2), CNRS Centre de Recherche Astrophysique de Lyon Site ENS-Lyon

2 PHOENIX Created in 1994 in Phoenix, AZ Peter Hauschildt, France Allard & Eddie Baron 1D, static, Radiative Transfer OS/ALI : w spherical symmetry with adaptive angular resolution w restraint relativity effects (solution in comoving frame) w 3D Hydrostatic Equilibrium (stars, brown dwarfs, planets), or Velocity field in relativistic expansion (novae, supernovae) Layer-dependant velocity up to speed of light (novae, supernovae) Convection: Mixing Length Theory Atomic diffusion Non-LTE (rate-operator splitting) for atoms and CO Chemical Equilibrium with NLCE for certain species (CO, CH 4, NH 3 ) 26 ionization levels, 85 elements (Th, U), 600 molecules, >1000 grain types Dynamical (no pre-tabulation) Opacity Sampling Database of atomic and molecular transitions Extinction cross-sections for 64 types of grains Cloud Model based upon Rossow (1978) timescales (sedimentation, condensation) Supersaturation computed from chemical equilibrium tables. Mixing from Radiative HydroDynamic (RHD)

3 Allard & Hauschildt (1995) Spectral Energy Distribution (SED) of a typical Very Low Mass star (VLMs) of effective temperature (T eff ) similar to that of a young brown dwarf (few million years). The SED of de dm8e VB10 (full line) is compared to a model (AH95, dotted lines) based on band model opacities. The spectrum where all discrete opacities (atomic & molecular line transitions) are omitted (dot-dashed), and the blackbody SED of same T eff indicates the importance of non-grey opacities.

4 The Infrared Problem remained despite the Progress on Water Opacities Dusty/Cond 2001 NextGen 1999 Test 1994 Base 1995

5 Search for planets around M dwarfs Easier to detect by RV, transit - Smaller planets - 05 R and 0.1 M -SPIRou -Carmenes -PLATO 2.0 -GAIA

6 The WISE satellite Fg1, ApJS, vol. 406, p. 142 WISE executes a sky survey from 3 to 25 m with a sensibility 500,000 times that of COBE/DIRBE and hundreds of times that of IRAS. The survey helps understand the origin of planets, stars, and galaxies, and will create an infrared atlas that will become an incomparable heritage for decades to come.

7 Clouds in brown dwarfs Fergley & Lodders, Astrophysics Update 2, edited by John W. Mason. ISBN X. Published by Springer Verlag, Heidelberg, Germany, 2006, p.1

8 Stars and brown dwarfs later than M3 are fully convective. The ωα dynamo cannot operate. ü ω 2 ü turbulent ü or both (rotation, stratified) And for dwarfs later than M8, activity decreases due to: ü neutrality of the atmosphere (low fractional ionization) ü high density: neutral particles kick ions off field lines

9 Surface convection in late M dwarf Ludwig, Allard, Hauschildt, (A&A 2002) RHD simulation T eff =2800K, logg=5.0 Vertically: timescale=100 sec velocities=240 m/s Horizontally: cell size = 80 km contrast 1.1%

10 Solar A Abundances factor of two revised less by oxygen RHD!!! simus Shifting to the blue (Doppler) due to the relative surface of the ascending convective cells Simulations by Ludwig et al. (2006) of the atmosphere of red dwarf (M type) with T eff = 3000K, logg= 5.0, [M/H]= 0.0 Figure 3 of Asplund, Grevesse, Sauval & Scott (2009, ARAA 47, 181). The profile predits a spectral line typical of Fe I computed with a HDR 3D model of the Sun (solide line) compared to the observations (grey losanges). This agreement is very clearly satisfying, and is due to the Doppler displacement coming from the convective mouvements, computed self-consistantly, and which enlarge, displace, and distort the theoretical profile. For comparison, the profile predicted by a 1D model atmosphere (red line; Howeger & Müller 1974) is also shown; the 1D profile has been computed with a microturbulence of 1 km s -1 and a macroturbulence adapted to reproduce the general profile in terms of its widnest. Note even with these two free parameters, the 1D profile reproduce neither the displacement,nor the asymmetry of the line.

11 CO5BOLD R(M)HD simulations Bernd Freytag, ENS-Lyon, now Uppsala H. G. Ludwig, W. Schaffenberger, S. Wedemeyer-Böhm, S. Höfner Matthias Steffen, IAP Potsdam, Germany General: 2D/3D Cartesian box, parallelized with OpenMP Magneto-Hydrodynamics (compressible): ü HYD module: approximate Riemann solver (Roe type) ü MHD module: HLLE solver Radiation transport: ü Module for global "Star-in-a-Box" models (central potential) ü Module for local "Box-in-a-Star" models (constant gravity) ü Non-local transport, grey/non-grey opacity scheme ü opacities from ATLAS, MARCS, PHOENIX Molecules, dust; additional densities: ü Dust; 2-bin: monomers + grains, one size per grid cell, forsterite (Mg 2 SiO 4 ) ü Dust; multi-bin: monomers + several grain sizes, forsterite (Mg 2 SiO 4 ) ü Dust; 4-moment method: amorphous carbon ümolecules: network for CO Rotation: Coriolis and Centrifugal Forces in each grid cell Impinging radiation from a parent star

12 The CO 5 BOLD GRID Kucinskas et al Freytag & Hoefner 2008 Freytag et al Chiavassas et al Freytag et al Caffau et al Wende et al Freytag et al Tremblay et al

13 2D HDR simulations of dust cloud formation in brown dwarfs atmospheres Gravity Waves!!! W350 x H80 km 2 during 36 hrs CO5BOLD simulations (Freytag et al. 2010) of the gas and forsterite (Mg 2 SiO 4 ) dust based on Phoenix opacities, on a cloud model (dust size bon distribution), and on the nucleation, condensation, coagulation, and sedimentation rates by Rossow (1978). Shown, in red, the dust grain mass density, and, in green, the entropy indicates the convection.

14 M-L transition: Comparing models from different authors Estimated Teffs of M dwarfs by Cassagrande, Flynn & Bessell mag uncertainty due to solar abundances! Oxygen solar abundance reduced by 22% (Caffau et al. 11) 34% (Asplund et al. 09) 39% (Grevesse et al. 07) compared to Grevesse et al. 93

15 BT2-Settl or MUSE/GAIA grid

16 Rajpurohit et al. (2013) La Silla, R=10.4Å vs Phoenix synthetic spectra

17 New evolution tracks 17

18 New BT-Settl Interior Models Quadruple system LkCa 3 (Torres et al. 2013) d = 127 pc and extinction E(B V) = 0.10 as in Torres et al. (2013) Incertainties on TiO opacities remain: AMES TiO less accurate but more complete vs Plez 2008 TiO more accurate but not complete enough ML calibration improve the models for young stars, BDs & Exoplanets 18

19 Local High Spatial Resolution 3D «box-in-a-star» simulations of M L T dwarfs Freytag, Allard, Homeier (2016) In preparation 3D Box-in-a-star simulations 4000 T eff 1400K with high spatial resolution x points x km/s 19

20 New BT-Settl Interior Models 20

21 UVES R = 40,000 Rajpurohit et al. (2014)

22 Bessell et al. (in preparation)

23 Phoenix synthetic spectra (black) vs observed spectra (color curves) Grid currently extends from : T eff = K Logg= [M/H]= Project: extend this grid to 15000K, to [M/H]= -4.0, [C/O] and less then 100K Includes disequilibrium chemistry for N 2, NH 3, CO, CO 2, CH 4 23

24 Web Simulator ONLINE! Offers synthetic spectra and thermal structures of published model grids and the relevant publications. Computes synthetic spectra, with/without irradiation by a parent star, and photometry for: ü stars ü brown dwarfs (1 Myrs - 10 Gyrs) ü irradiated stars or planets ü telluric exoplanets Computes isochrones and finds the parameters of a star by chi-square fitting of colors and/or mags to the isochrones. Rosseland/Planck as well as monochromatic opacity tables calculations. 24

25 Web Simulator In Test Mode! Offers synthetic spectra and thermal structures of published model grids and the relevant publications. Computes synthetic spectra, with/without irradiation by a parent star, and photometry for: ü stars ü brown dwarfs (1 Myrs - 10 Gyrs) ü irradiated star or planet ü telluric exoplanets Computes isochrones and finds the parameters of a star by chi-square fitting of colors and/or mags to the isochrones. Rosseland/Planck as well as monochromatic opacity tables calculations. 25

26 Global RHD simulations Freytag, Schaffenberger & Allard 2014 (in prep.) T eff = 2200K, logg= 3.5, solar, P=8 Hr Jupiter è However radius scaled by a factor 20! Improvement expected with MPI