New version page

Data Reduction Strategies and Challenges

This preview shows page 1-2-3-22-23-24-44-45-46 out of 46 pages.

View Full Document
View Full Document

End of preview. Want to read all 46 pages?

Upload your study docs or become a GradeBuddy member to access this document.

View Full Document
Unformatted text preview:

Data Reduction Strategies and Challenges: White, Pink and Red Noise; Data Trend Filtering & SystematicsFrédéric Pont Geneva ObservatoryMichelson Summer Workshop July 2007Michelson Summer Workshop July 2007The mass-radius relation for transiting EGP2Michelson Summer Workshop July 2007The challenge :Field transit searchs with CCD cameras Monitor 105-106 starsfor 101 - 102 nightsat 10-2 - 10-3 accuracyOHP 22-25 aout 2005 — 10th anniversary 51 Peg bMost transiting planets are near the detection thresholda transiting planet nothing at all!an eclipsing binarySystematics / red noise in surveysEXPLORE/OC!Sara SeagerMonitor! ! Suzanne AigrainSuperWASP! Andrew Cameron, ! ! Keith HorneHAT! ! Gaspar BakosTrES! ! Tim Brown! ! Hans DeegBEST! ! Heike Rauer,! ! Anders EriksonMagnitudeσMichelson Summer Workshop July 2007depthσ / √nSNR =Detection threshold:SNR > 7 - 10(105 - 108 stat. tests)Michelson Summer Workshop July 2007Field transit searchs with CCD camerasDream and reality Orders of magnitude :Number of pixels in camera 107 pixelsMean distance between targets >101 pixels󲰛 Number of targets 105Accuracy : better than 10-2 Accuracy over transit duration (2-4 hours): 10 -2 󲰛 hot Jupiter transits detected at 10 sigmasNumber of transiting hot Jupiters : 0.1 × 0.01 × 105 = 102Number of detected transiting planets : dozens (e.g. Horne 2003)7Actual detections : 1 -2 transiting hot Jupiter per year per season for major surveys0 for minor surveysMichelson Summer Workshop July 20078Michelson Summer Workshop July 20079Michelson Summer Workshop July 200710Michelson Summer Workshop July 200711Nature of OGLE planetary transit candidatesEclipsing binaries - grazing- low-mass companion- multiple systems and blends! False positives of the transit detection Transiting gas giant planets [5] [>20] [>100] [>50] [10-20]Colours of noisewhiteredpinkTransit detection signal-to-noise:SNR =Detection threshold:SNR > 7 - 10(105 - 108 stat. tests)Significance of transit detectionsFor a normal Hot Jupiter :σ ~ 5 mmag, depth ~ 1%, n ~ 3 transits x 10 points ⇒ SNR ~ 10 most HJ should be detectabledepthσ / √nSNR2 = Transit detection significance with real photometric noisedepth2σ2 / n 1/ n2 Σ cov(xi,xi)In the regime relevant to transit surveys, the red term dominates!white noise (photon+sky+scintillation)σ = 3-10 mmag, n=20-50, σ2 / n = 0.1 - 0.5 mmag 2red noise (systematics from seeing, weather, tracking)1/ n2 Σ cov(xi,xi) = (2-5 mmag)2 = 4 - 25 mmag 2 σ2 / n + 1/ n2 Σ cov(xi,xi)• Factor 3-5 in threshold!• Weaker dependence on magnitude• Steeper dependence on periodDetection threshold with red noise (systematics)Detection statistic with red noise(Pont, Zucker & Queloz 2006)Hot JupitersHot NeptunesDetected planets per 104 targetsDetection potential of surveysIndividual measurement dispersionWhite-noise σ / √n dispersion over transit-length interval Actual dispersion over transit-length interval Noise and detection in ground-based transit surveys0.60.40.20.01314 15MagnitudeTransit depth [%]Example: Red noise in SWASP data before detrending algorithmExample: Red noise in SWASP data after detrending algorithmSecond transiting planet around OGLE-TR-111 ?Known parameters decorrelation26Example from HST data on transiting HD189733b27Systematics Removal in large sets of light-curvesSysRem(Tamuz, Mazeh & Zucker 2005)Trend Filtering Algorithm(Kovacs, Bakos & Noyes 2005)28What kind of ‘systematics’?• Colour-dependent atmospheric extinction• ci – colour, aj – airmass• Contaminating light (moon, earth)• Position-dependent CCD response• etc…29Huge photometric datasets:30Example: SuperWASP 16h30+28 field300 stars, 2549 observations spanning 100 daysBeforeAfter31ri(j) = transit (?)+ white noise+ ci1 airmass (j)+ ci2 temperature(j)+ ci3 ???32Finding the ci and ai?Find ci and aj that minimize:33Assume aj are known,solve for ciNow ci are known,solve for ajStar no. i:Image no. j:34Applying Sys-Rem to OGLE data35Applying Sys-Rem to OGLE data36Applying Sys-Rem to OGLE data37Applying Sys-Rem to OGLE data3839OGLE-TR-132, Before…… and after4041D.J. Christian (Belfast) W.I. Clarkson (Open University) A. Collier Cameron (St Andrews) N.A. Evans (Keele) A. Fitzsimmons (Belfast) C.A. Haswell (Open University) C. Hellier (Keele) S.T. Hodgkin (Cambridge) K. Horne (St Andrews) S.R. Kane (St Andrews) F.P. Keenan (Belfast) T.A. Lister (St Andrews) A.J. Norton (Open University) D. Pollacco (Belfast) R. Ryans (Belfast) I. Skillen (ING) R.A. Street (Belfast) R.G. West (Leicester) P.J. Wheatley (Leicester) SuperWASPWide Angle Search for Planets4212h43+28 field: First four SysRem components (arbitrary units)Secondary extinction/dustTemperature dependent focus?A variable star among the standards?Mystery blip43Example: SuperWASP 16h30+28 field300 stars, 2549 observations spanning 100 daysBeforeAfter4445σr=5 mmagσr=3 mmagσr=2 mmagσr=1 mmagUnknown effects, unknown factorscross-terms Known effects, known factors ci × aiReductionKnown effects, unknown factors ?i × aiDetrendingUnknown effects, unknown factors ?i ×


Loading Unlocking...
Login

Join to view Data Reduction Strategies and Challenges and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Data Reduction Strategies and Challenges and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?