Research interests and main results

Cloud of the most frequent words appearing in my papers. Done with ADS Bumblebee

Characterisation of transiting exoplanets

Artist Impression of a multiplanetary system (credits: A.Santerne/ESO)
The search for planets outside the solar system revealed that planets are common in the Milky Way. They also present a large range of physical properties. Characterizing new exoplanets led us to understand the formation, migration and evolution of planetary systems. Some exoplanets have an orbital plane nearly aligned with the Earth that allow them to pass in front of their host star. For such planets, called transiting planets, it is possible to characterize their radius and mass, and thus their bulk density. The density is then used to constrain the internal structure of the planet. Since the planet is transiting its host star, it is also possible to probe its atmospheric properties (like albedo, heat distribution) and composition.

Radius of exoplanets can be measured throught high-precision photometry, from the ground or using space-based telescope likeKepler or CoRoT. To measure their mass, one technique consists in searching for the reflex motion of the star orbited by the planet using the Doppler effect, so-called radial velocity. This requires high-resolution spectrographs like SOPHIE, HARPS, HARPS-N.

Since 2010, I am leading a large spectroscopic programme with the SOPHIE spectrograph at the Haute-Provence Observatory (France) to characterise giant exoplanets and brown dwarfs found in transit by the Kepler / K2 mission. Here is a sample of characterised systems:

The solution against astrophysical false positives: PASTIS

Several configurations of stellar systems can mimic the transit of a plane and are limiting the interpretation of large photometric surveys like Kepler. The planetary nature of transiting candidates must be confirmed. This can be done e.g. by measuring the mass of the transiting object to determine if it is in the planet’s mass range. Following up transiting candidates with radial velocity instruments is thus in important step in the discovery and characterization of a new transiting exoplanet.

(a)    (b)   (c)   (d)
Different astrophysical scenarios that can mimic a planetary transit. (a) a transit of a planet ; (b) an eclipse of a low-mass star or a brown dwarf with radius similar to that of Jupiter ; (c) an eclipse of a binary in the foreground/background, aligned with the target star, as seen from the Earth (so-called background eclipsing binary) or physically bound with the target star (so-called triple system) ; (d) a transit of a planet on a star aligned with the target star, as seen from the Earth, in the foreground/background (so-called background transiting planet) or physically bound with the target star (so-called companion transiting planet).

Current spectrographs are not sensitive enough to detect the mass of most of the small-planet candidates discovered by Kepler and CoRoT. To establish their planetary nature, the planet-validation technique consists in evaluating the probability of each scenario to mimic the observed data (photometry, radial velocity, spectroscopy). If the planet scenario has a significantly higher probability than all other false-positive scenarios, then the candidate is considered as validated. The PASTIS tool (Planet Analysis and Small Transit Investigation Software) has been developped, mainly at the Laboratoire d’Astrophysique de Marseille, to perform such statistical comparison of the various aforementionned scenarios and then, to validate planets as small as the Earth.
The PASTIS tool is fully described in the papers:PASTIS logo So far, PASTIS was used to statistically validate the following exoplanets:
Being a cutting-edge data-analysis software for transiting exoplanets, PASTIS was also used for the detailled characterisation of several tens of planetary systems. PASTIS was the name of my project for the  Marie Curie fellowship hosted at the Centro de Astrofísica da Universidade do Porto (Portugal) and funded by the european commission, with the project ID 627202.

Evaluation of Kepler's false-positive rate

The main objective of the Kepler prime mission was to measure the occurrence rate of exoplanets (i.e. the fraction of stars with planets), down to Earth-size planets in the habitable zone (i.e. the distance from the star where liquid water might be at the surface of the planet) of their host stars. To derive robust occurrence rates, it is mandatory to evaluate the reliability of the planetary candidate sample, or the false-positive probability/rate (FPP / FPR), i.e. the probability that a transit candidate is not a planet. While the FPP was initially estimated to be very low, the observations with the SOPHIE spectrograph that I led on a complete sample of Kepler's short-period giant-planet candidates revealed that the false-positive rate in this planet population was much higher, with an observed value of 34.8 ± 6.5%. I then revised this value on the sample of Kepler's giant-planet candidates transiting up to an orbital period of 400 days to a value of 54.6 ± 6.5 %. This last value includes the detections based on the entiere data set collected by Kepler (Q1 - Q17).

I led a study to evaluate how eccentric binairies for which only the secondary eclipse could be seen from the Earth might mimic exoplanetary transits. This study also estimated how many of these cases should be in the exoplanet candidate list of Kepler.

My papers on the Kepler false-positive rate:

Statistics of the giant planet populations in the Kepler field

Using the giant-planet candidates detected in the Kepler data (Q1 - Q17) for which all false positives were screened out with the SOPHIE spectrograph (over a 6-year observing programme), I led a paper deriving the occurrence rates of giant planets, in different ranges of orbital periods, from a day up to 400 days. By identifying one by one the false positives in the candidate sample, the derived statistics are very robust. The main surprise was the detection of the valley in the period distribution of giant exoplanets. This valley was first clearly identifiied since the beginning of spectroscopic surveys. This valley was first unconfirmed in the previous occurrence rate studies based on the Kepler data because of not-perfect correction for false positives. Hence, this valley that reveals two different formation procesess for giant planets was for the first time confirmed in this paper.

This study also confirmed the depression of hot jupiters in the Kepler field compared to other photometric and spectroscopic surveys. This is not observed for giant planets with longer orbital periods, although few photometric surveys have explored this domain. The reason for the Kepler field to be hot-jupiter-poor is still not understood.

Last update: August 2017