A comparison between CoRoT and Dome C highprecision photometry

1
A comparison between CoRoT and Dome C
high-precision photometry

• A. F. Lanza, S. Messina, G. Cutispoto, G. Leto,
  I. Pagano
• INAF-Osservatorio Astrofisico di Catania, Italy
• M. Auvergne, A. Baglin
• LESIA, CNRS UMR 8109, Observatoire de Paris,
  France
• P. Barge
• Laboratoire dAstrophysique de Marseille, CNRS
  UMR 6110,
• Université de Provence, France

2
The CoRoT satellite
A space project developed and operated by CNES,
with the contribution of Austria, Belgium,
Brasil,ESA/RSSD, Germany and Spain Primary
science goals a) asteroseismology b) search for
planetary transits.

• Launched on 27 December 2006
• Low-Earth orbit (orbital period 6184 s)
• Telescope aperture 27 cm
• Observational runs up to 150 days
• Duty cycle 90-95
• Photometric accuracy of 85 ppm for a V12.5 star
  with an integration time of 6184 s in white light
  (passband 300-1100 nm). see Auvergne et al.
  2008 for details

3
The light curve of CoRoT-Exo-2a

• A main-sequence G7 star (V12.57), accompanied by
  a hot Jupiter with an orbital period of 1.743 d
  (Alonso et al. 2008)

4
Out-of-transit light curve analysis

• The original chromatic N2-level data are summed
  up to get white light fluxes
• the effects of hot pixels, pointing jitter and
  outliers are removed (see Lanza et al. 2008)
• Transits are removed by means of the ephemeris of
  Alonso et al. (2008)
• The out-of-transit light curve is binned in
  1-orbital period (6184 s) intervals, obtaining a
  time series consisting of 1945 normal points
  along 142.006 days

5
Search for the rotation period

• The out-of-transit light curve is binned in
  1-orbital period intervals, obtaining a time
  series of 142.006 d consisting of 1945 normal
  points
• we apply the Lomb-Scargle periodogram to extract
  the rotation period from the light modulation
• we find Prot 4.52 0.14 d.

6
Spot modelling

• We adopted the Maximum Entropy spot model of
  Lanza et al. (2007, AA 464, 741)
• The surface of the star is divided into pixels of
  18 x 18 and a mixture of spotted and unspotted
  photosphere is assumed within each pixel
• The fraction of the area covered by the spots is
  the pixel filling factor fi
• A single map is derived out of virtually infinite
  solutions by means of the ME regularization
• Q ?2(f) ? ? S(f).

7
Results of the Maximum Entropy spot modelling

• Our main interest is in studying stellar rotation
  and photospheric magnetic activity as traced by
  starspots
• We fit the CoRoT light curve with the model of
  Lanza et al. (2007) and find
• Longitude distribution of the spotted area vs.
  time
• Variation of the total spotted area vs. time.
• (see Lanza et al. 2008, Cool Star 15,
  arXiv0809.0187)

8
Best fit of the light curve
9
Spot area vs. longitude and time
10
Variation of the spot area vs. time(a possible
Rieger cycle)
Spots only (solid line) Pcyc 28.9 4.8 d
Spots and faculae (dot-dashed line) Pcyc 29.5
4.8 d
11
Estimating the relative spot area variation from
the integrated flux deficit
Funspotted is unknown, but for the purpose of
measuring relative variations, we can assume the
brightest observed flux.
12
Comparison of the two methods
Spot area from the flux deficit integrated along
successive rotation cycles
Spot area from ME spot modelling (with and
without solar-like faculae)
13
Simulating Dome C observations

• The light curve of CoRoT-Exo-2a can be used to
  simulate time series photometric observations
  from Dome C
• First, we extend the light curve up to 278 days,
  by mirroring the time series beyond the final
  endpoint.

14
The extended time series
15
Duty cycle of Dome C observations

• Photometric observations are possible if
• a) The Sun is at least 8 deg below the horizon
• b) The Moon is at least 30 deg away from the
  target
• and its phase is less than 0.90 (1.0 is full
  Moon)
• c) The airmass is lower than 2.0
• d) There are no clouds
• we assume that the statistics of cloudless
  intervals follow a gamma distribution (see the
  talk by Fruth, this meeting) with ?0.84 and
  ?6.8 days
• (Kenyon Storey 2006 Rauer et al. 2008)

16
Duty cycle of Dome C observations

• Photometric observations are possible if
• a) The Sun is at least 8 deg below the horizon
• b) The Moon is at least 30 deg away from the
  target
• and its phase is less than 0.90 (1.0 is full
  Moon)
• c) The airmass is lower than 2.0
• d) There are no clouds
• we assume that the duration of single cloudless
  periods follows a gamma distribution of the form
• p(t) (4t/?2) exp(-2t/?),
• where p(t) is the probability of a cloudless
  interval of length t, and ? is the mean duration
  of a cloudless period
• (Kenyon Storey 2006 Rauer et al. 2008)

17
Observing time distribution

• For simplicity, we assume a target at declination
  ? -90
• The observing season starts on 2008 March 1st (as
  in Rauer et al. 2008)

18

• Cloud model with a fraction of clear sky ? of 84
  percent and ? 6.8 days (? is the mean duration
  of clear periods)

19
Photon shot noise and scintillation

• We simulate photometric time series for two
  levels of photon shot noise
• 100 ppm (corresponding to V 14 for D60 cm,
  under ideal conditions)
• 1000 ppm (probably more realistic because of red
  noise, variable extinction and background
  effects)
• The integration time is assumed to be 6184 s
• We simulate 500 light curves for each of the two
  noise levels above, each light curve
  corresponding to a different realization of the
  photon noise and cloud coverage.

20
Distribution of the rotation periods

• Lomb-Scargle periodograms of the time series give
  the distribution of the rotation periods (FAP lt
  0.01 in all the cases)

21
Distribution of the rotation periods (close-up
view of the central portions of the histograms)
22
Distribution of the periods of the short-term
spot cycle

• The integrated flux method is applied to estimate
  the spotted area variations
• For a noise level of 100 ppm, we find a
  significant periodicity detection
• (i.e., FAP lt 0.01) for 180 out of 500
  light curves using Lomb-Scargle periodogram
• For a noise level of 1000 ppm, for 168 out of 500
  light curves

23
Distribution of the periods of the short-term
spot cycle

• For a noise level of 100 ppm, we find a
  significant detection (i.e., FAP lt 0.01) for 180
  out of 500 light curves
• For a noise level of 1000 ppm, for 168 out of 500
  light curves

24
Conclusions

• Our results are preliminary, being based on the
  light curve of just one target
• For a target at high southern declination (? lt
  -55), Dome C allows us a longer seasonal
  coverage than CoRoT, but with an average duty
  cycle of ? 45-50 (cf. Rauer et al. 2008 for
  details)
• For an active solar-like star with V14-15
  (observed with D60 cm), this allows us to
  measure the rotation period (for an amplitude of
  0.04 mag and Prot 4-5 d)
• Short-term spot cycles, with a period of the
  order of a month, are detectable in 30-35 of
  the cases, as a result of the reduced duty cycle
  with respect to CoRoT (90-95).