Corot Week 9 ESTEC 5-9 Dec 2005 Frédéric Pont Geneva Observatory Lessons from the OGLE planetary transit survey Francois Bouchy (Marseille/OHP), Nuno Santos (Lisbon), Claudio Melo (ESO), Didier Queloz, Shay Zucker, Stephane Udry, Michel Mayor (Geneva), Claire Moutou (Marseille)
Corot Week 9 ESTEC 5-9 Dec 2005 FLAMES : 7-fiber link to high-resolution spectro UVES on the VLT R ~ 45’ mn on mv=17 S/N ~ 8 ~30 m/s FLAMES follow-up of OGLE candidates Field of view 25’x 25’ 137 shallow transit candidates published by OGLE 60 followed in radial velocity by our team with FLAMES (15<V<18) 5 transiting planets detected
Corot Week 9 ESTEC 5-9 Dec 2005 Nature of planetary transit mimics Eclipsing binaries - grazing - low-mass companion - multiple systems and blends False positive of the transit detection deep transits shallow transits
Corot Week 9 ESTEC 5-9 Dec 2005 Light curve can be undistinguishable from planetary transit can be impossible to discriminate from light curve easy with high-resolution spectroscopy II: small stellar companion Planetary transit mimics OGLE-TR-122b, Pont et al. 2005
Corot Week 9 ESTEC 5-9 Dec 2005 Wide range of possible scenarios Requires global analysis, light curve + radial velocity Often more than one component visible in the spectrum (but not always) Triple system can also mimic planetary RV variation! ranges from easy to very tricky to discriminate III: multiple system or blend OGLE-TR-33 Torres et al (GSC Mandushev et al. 2005) Planetary transit mimics RV Light curveCMD
Corot Week 9 ESTEC 5-9 Dec 2005 RV follow-up essential part of photometric transit surveys required to establish planetary nature of Gas Giant candidates Implications of planetary transit mimic scenarios
Corot Week 9 ESTEC 5-9 Dec 2005 What are the contaminants for super-Earth transits ? Transiting M-dwarfs like OGLE-TR-122 will not be a problem BUT Gravitationally bound triple system will still be a source of mimics (how good mimics? To be studied) Background eclipsing binaries will be very numerous. “CorotLux” simulator (Guillot et al.) --> ~ per field Most of these will be easy to discriminate
Corot Week 9 ESTEC 5-9 Dec 2005 Detection threshold of transit signal
Corot Week 9 ESTEC 5-9 Dec
Corot Week 9 ESTEC 5-9 Dec 2005 — magnitude dependence with white noise — magnitude dependence with red noise OGLE transit detection threshold
Corot Week 9 ESTEC 5-9 Dec 2005 Colours of noise white red pink
Corot Week 9 ESTEC 5-9 Dec 2005
Factor 3-5 in threshold! Weaker dependence on magnitude Steeper dependence on period Detection threshold with red noise (systematics)
Corot Week 9 ESTEC 5-9 Dec 2005 Transit detection signal-to-noise: SNR = Detection threshold: SNR > ( stat. tests) Real detection treshold and potential of transit surveys? depth / n OGLE-TR-131OGLE-TR-132
Corot Week 9 ESTEC 5-9 Dec 2005 SNR 2 = Transit detection significance with real photometric noise depth 2 In the regime relevant to transit surveys, the red term dominates! white noise (photon+sky+scintillation) = 3-10 mmag, n=20-50, 2 / n = mmag 2 red noise (systematics from seeing, weather, tracking) 1/ n 2 cov(x i,x i ) = (3-5 mmag) 2 = mmag 2 2 / n + 1/ n 2 cov(x i,x i )
Corot Week 9 ESTEC 5-9 Dec 2005 Factor 3-5 in threshold! Weaker dependence on magnitude Steeper dependence on period Detection threshold with red noise (systematics)
Corot Week 9 ESTEC 5-9 Dec 2005 Second transiting planet around OGLE-TR-111 ?
Corot Week 9 ESTEC 5-9 Dec 2005 Application of revised yield estimates to some surveys affects also Gillon et al. 2005
Corot Week 9 ESTEC 5-9 Dec site / optics /detector / photometric reduction - covariance removal (SYS-REM, Tamuz et al. 2004) - modify tradeoff systematics vs random error There may be a practical limit for ground-based wide-field photometry (due to number of comparison star compared to number of correction parameters and total variation of flux) We find r 3 mmag for all large ongoing surveys The hour-scale systematics of ground-based surveys: does the atmosphere set a hard limit ? ! Detection of transiting Neptunes from ground prevented by red noise.
Corot Week 9 ESTEC 5-9 Dec 2005 COROT r = 0.7 mmag according to Corot instrument team r = 2.2 mmag according to realistic blind-test simulations [Moutou et al. 2005] High predicted yield of transiting Hot Jupiters due to large time coverage (150 x 24 hrs) Sensitivity to Super-Earth will depend on level of actual red noise Implications for Corot ?