The very low mass multiple system LHS 1070. A testbed for model atmospheres for the lower end of the main sequence^⋆,⋆⋆

¹ Université de Franche Comté, Institut UTINAM CNRS 6213, Observatoire des Sciences de l’Univers THETA de Franche-Comté-Bourgogne, Observatoire de Besançon, BP 1615 25010, Besançon Cedex, France
e-mail: celine@obs-besancon.fr
² Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany
³ CRAL, UMR 5574, CNRS, Université de Lyon, École Normale Supérieure de Lyon, 46 Allée d’Italie, 69364 Lyon Cedex 07, France
⁴ Universitäts-Sternwarte München, Ludwig-Maximilians-Universität, Scheinerstr. 1, 81679 München, Germany
⁵ Hamburger Sternwarte, Gojenbergsweg 112, 21029 Hamburg, Germany
⁶ Lund Observatory, Lund University, Department of Astronomy and Theoretical Physics, Box 43, 221 00 Lund, Sweden
⁷ Max-Planck-Institut für Astrophysik, Karl-Schwarzschild-Str. 1, 85748 Garching, Germany

Received: 13 February 2012
Accepted: 1 August 2012

Abstract

Context. LHS 1070 is a nearby multiple system of low mass stars. It is an important source of information for probing the low mass end of the main sequence, down to the hydrogen-burning limit. The primary of the system is a mid-M dwarf and two components are late-M to early L dwarfs, at the star-brown dwarf transition. Hence LHS 1070 is a valuable object to understand the onset of dust formation in cool stellar atmospheres.

Aims. This work aims at determining the fundamental stellar parameters of LHS 1070 and to test recent model atmospheres: BT-Dusty, BT-Settl, DRIFT, and MARCS models.

Methods. Unlike in previous studies, we have performed a χ²-minimization comparing well calibrated optical and infrared (IR) spectra with recent cool star synthetic spectra leading to the determination of the physical stellar parameters T_eff, radius, and log    g for each of the three components of LHS 1070.

Results. With exception of the MARCS models which do not include dust formation, the models are able to reproduce the observations and describe the main features of the visible to IR spectra. This is consistent with the fact that dust formation prevails in the B and C component atmospheres. The parameters obtained with the DRIFT models confirm the values determined in earlier studies. But important differences between models are observed, where the MARCS model is too bright in the H and K bands, and the BT-Settl and BT-Dusty models systematically yield up to 100 K higher T_eff in the case of the B and C components. This confirms a trend for models without, or with less efficient cloud formation, to predict higher T_eff than models richer in dust (DRIFT). Even models including cloud physics however still produce slightly too bright J band flux, showing as too blue J − K colors. The onset of dust formation remains therefore a particularly challenging regime to understand.