Astrometry of Dust Clouds
The space-based mission Gaia uses astrometry to find new planets and should provide its harvest in a few data release from now. About 20 000 new planets more massive than Jupiter with periods out to several au will be discovered within 500pc (Perryman et al. 2014). When doing astrometry on unresolved systems, we measure the displacement of the photo-centre of the system and not necessarily that of the star (if they are different). Indeed, if there is a bright source of light (e.g. a dust cloud) separated from the star but within the beam of the telescope, the photo-centre will not be on top of the star but rotate with the dust cloud. Thus, it could happen that instead of seeing the star wobbling because of the gravitational effect from a planet, it wobbles because of the presence of a bright dust cloud in the system (i.e. this is a purely photometric effect rather than gravitational and the barycentre of the system can still be almost on top of the star even if it wobbles).
In the paper that appears below, we study this artefact of the astrometric approach and show that dust clouds (or an inhomogeneity in a dust disc) can mimic the astrometric signal caused by planets. We show that the dust inhomogeneities required to mimic giant planets (such as the ones that Gaia is targeting) are large enough that mid-IR missions such as Spitzer or WISE could detect these dust clouds and thus confirm the non-planetary origin. However, a dust cloud (a cometary mass object is enough) that would mimic terrestrial-like planets would not be detectable with current facilities. This can also affect observations of very young planets forming in protoplanetary discs if astrometry is used. It is thus essential that future missions such as NEAT aiming to detect Earth-like planets using astrometry take account of this artefact. We provide some further tests (e.g. observations at different wavelengths as the signal from a planet should be achromatic) that could be done to disentangle if an astrometric signal is from a planet or not.
For Gaia planets, it means that some of the astrometric displacements measured will be due to this effect. However, for Gaia planets, a lack of an IR-excess rules out the possibility that an astrometric detection is dust but an IR-excess does not mean the planet is not real, since the dust could be axisymmetric and create no displacement of the photocentre. Therefore, for systems with IR-excesses, further tests such as proposed in our paper will be needed to confirm the planet.
This artefact can also be seen as a new method to detect dust clouds (or quantify the clumpiness of a disc) in inner debris discs that cannot be otherwise resolved.