Frdric Pont

1
Lessons from the OGLE planetary transit survey

• Frédéric Pont
• Geneva Observatory

Francois Bouchy (Marseille/OHP), Nuno Santos
(Lisbon), Claudio Melo (ESO), Didier Queloz, Shay
Zucker, Stephane Udry, Michel Mayor (Geneva),
Claire Moutou (Marseille)
2
FLAMES follow-up of OGLE candidates
137 shallow transit candidates published by
OGLE 60 followed in radial velocity by our team
with FLAMES (15ltVlt18)

• FLAMES 7-fiber link to high-resolution spectro
  UVES on the VLT
• R 45000
• 45 mn on mv17
• S/N 8 ? 30 m/s

Field of view 25x 25
? 5 transiting planets detected
3
Nature of planetary transit mimics
deep transits
Eclipsing binaries
- grazing- low-mass companion- multiple systems
and blends
False positive of the transit detection
shallow transits
4
Planetary transit mimics
II small stellar companion
Light curve can be undistinguishable
from planetary transit
OGLE-TR-122b, Pont et al. 2005
? can be impossible to discriminate from light
curve ? easy with high-resolution spectroscopy
5
Planetary transit mimics
III multiple system or blend
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!
Light curve
CMD
OGLE-TR-33 Torres et al. 2004
RV
(GSC 01944-02289 Mandushev et al. 2005)
? ranges from easy to very tricky to
discriminate
6
Implications of planetary transit mimic scenarios

• RV follow-up essential part of photometric
  transit surveys
• required to establish planetary nature of Gas
  Giant candidates

7
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.) --gt 100-200 per
field Most of these will be easy to discriminate
8
Detection threshold of transit signal
9
0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 5 0 0 0 0 0
10
OGLE transit detection threshold

• magnitude dependence with white noise
• magnitude dependence with red noise

11
Colours of noise
white
red
pink
12
(No Transcript)
13
Detection threshold with red noise (systematics)

• Factor 3-5 in threshold!
• Weaker dependence on magnitude
• Steeper dependence on period

14
Real detection treshold and potential of transit
surveys?

• Transit detection signal-to-noise
• SNR
• Detection threshold
• SNR gt 7 - 10
• (105 - 108 stat. tests)

depth
? / ?n
OGLE-TR-131
OGLE-TR-132
15
Transit detection significance with real
photometric noise
depth2

• SNR2

?2 / n 1/ n2 ? cov(xi,xi)
In the regime relevant to transit surveys, the
red term dominates! white noise
(photonskyscintillation) ? 3-10 mmag,
n20-50, ?2 / n 0.1 - 0.5 mmag 2 red noise
(systematics from seeing, weather, tracking) 1/
n2 ? cov(xi,xi) (3-5 mmag)2 9 - 25 mmag 2
16
Detection threshold with red noise (systematics)

• Factor 3-5 in threshold!
• Weaker dependence on magnitude
• Steeper dependence on period

17
Second transiting planet around OGLE-TR-111 ?
18
Application of revised yield estimates to some
surveys
affects also Gillon et al. 2005
19
The hour-scale systematics of ground-based
surveys does the atmosphere set a hard limit ?

• - 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

! Detection of transiting Neptunes from ground
prevented by red noise.
20
Implications for Corot ?

• 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