Search for exoplanets with the radial-velocity technique: quantitative diagnostics of stellar activity^*

¹ Laboratoire d'Astrophysique de l'Observatoire de Grenoble, Université Joseph Fourier, BP 53, 38041 Grenoble, France e-mail: Morgan.Desort@obs.ujf-grenoble.fr
² Observatoire de Genève, Université de Genève, 51 Ch. des Maillettes, 1290 Sauverny, Switzerland

Received: 22 June 2007
Accepted: 27 July 2007

Abstract

Aims.It is known that stellar activity may complicate the analysis of high-precision radial-velocity spectroscopic data when looking for exoplanets signatures. We aim at quantifying the impact of stellar spots on stars with various spectral types and rotational velocities and comparing the simulations with data obtained with the HARPS spectrograph.

Methods.We have developed detailed simulations of stellar spots and estimated their effects on a number of observables commonly used in the analysis of radial-velocity data when looking for extrasolar planets, such as radial-velocity curves, cross-correlation functions, bisector velocity spans, and photometric curves. Stellar and spot properties are taken into account, as well as the characteristics of the spectrograph used (generally HARPS). The computed stellar spectra are then analyzed in the same way as when searching for exoplanets.

Results.1) A first grid of simulation results (radial-velocity amplitudes, bisector velocity-span amplitudes and shapes, and photometry) is built for F-K type stars, with different stellar and spot properties. 2) It is shown quantitatively that star spots with typical sizes of 1% can mimic both radial-velocity curves and the bisector behavior of short-period giant planets around G-K type stars with a  lower than the spectrograph resolution. For stars with intermediate , smaller spots may produce similar features. Such spots may complicate the search for low-mass planets on short-period orbits. In these cases, additional observables (e.g., photometry, spectroscopic diagnostics) are mandatory to confirm the presence of short-period planets. We discuss these possibilities and show that, in some cases, photometric variations may not be enough to clearly rule out spots as explanations of the observed radial-velocity variations. This is particularly important when searching for super-Earth planets. 3) It is also stressed that quantitative values obtained for radial-velocity and bisector velocity-span amplitudes depend strongly on the detailed star properties, on the spectrograph used, on the line or set of lines used, and on the way they are measured. High-resolution spectrographs will help in distinguishing between spots and planets.