Info

3 transits detected

Current

CCfcCe.NMKOl'tit 0 Jr r,

Fig. 8.7 Artist's impression of the CoRoT space mission (CNES)

CCfcCe.NMKOl'tit 0 Jr r,

Fig. 8.7 Artist's impression of the CoRoT space mission (CNES)

As far as instruments are concerned, CoRoT consists of an afocal off-axis telescope of about f/4, with two parabolic mirrors having a collecting surface equivalent to that of a telescope 27 cm in diameter. This images a star field 7 square degrees in area, half of which (for the exoplanet survey) is imaged by 2 CCDs, with 2048 x 2048 pixels, cooled passively to -40°C, and working in photon-counting mode in the visible and near infrared (passband 400-900 nm).

CoRoT was launched on 27 December 2006 by a Soyuz rocket, and placed in a polar orbit at an altitude 896 km. Such an orbit allows the same field to be observed continuously for a maximum of 5 months, after which a rotation by 180° is necessary to avoid being blinded by the Sun. Another field is then observed for the next 5 months. Between long-duration observations, small, shorter programmes are carried out. Taking the chosen orbit and the constraints imposed by reversing the satellite because of the relative positions of the Earth and Sun into account, CoRoT will observe in two principal directions ('CoRoT's eyes') directed towards the galactic centre and the anticentre (Fig. 8.8).

CoRoT should be able to observe 12000 targets simultaneously, with magnitudes between 11 and 16 per field for 5 months (150 days), and at least 5 fields during the mission's overall duration (about 3 years), making a total of more than 60 000 targets. The photometry of each star is measured every 8.5 min (512 s) with a photometric precision of a few times 10-4 in 1 h. Such accuracy should enable the detection of terrestrial-type objects a few Earth radii in size. Given the observational time for each field, and the necessity of detecting several transits to determine the period of the object, only 'hot' planets (close to their stars) are likely to be detected in sufficient quantity to be of use statistically. The first exoplanet was detected on 3 May 2007.

Fig. 8.8 'CoRoT's eyes': the regions of the sky towards the galactic centre (Right Ascension = 104.5°) and the anticentre (Right Ascension = 284.5°) which CoRoT will observe continuously for 5 months

The specific feature of the CoRoT instrument is that it provides photometry of 80 per cent of objects (the brightest) in three bands (Fig. 8.9). To do this, a system of prismatic dispersers is located between the objective and the focal plane. This gives each object an image in the form of a small spectrum. To avoid transferring the whole contents of a CCD each time it is read (every 32 s to avoid saturation), each CCD is partially read (over about 5-10 per cent of the surface), thanks to the use of photometric masks for each object (Fig. 8.9). These masks have a shape appropriate to each object. They are calculated by optimization of criteria based on a maximum signal-to-noise ratio and also taking the following into account:

• the magnitude of the object

• its temperature (its spectral type and thus its colour)

• its position on the CCD (the image spot for any one object varies across the field)

• the environment of each star (i.e., any possible contamination)

BlUe Green Red

Fig. 8.9 The principles involved in CoRoT's colour photometry. For each star, the image is a small spectrum (top left, where blue is to the left and red at the right), which is observed through a photometric mask (centre left), the three zones of which define three colours (bottom left)

The colour photometry of each object is obtained by selecting three distinct zones in the readout mask for the mini-spectra. So this is not spectral photometry in the classical sense of the term (with filters whose passband is known), but rather relative colour information about each object.

Colour information is of interest on two counts:

• It allows us to distinguish a stellar event from a planetary transit in the object's photometric curve. To a first approximation, a transit is achromatic (the relative variation in the luminosity is independent of the wavelength of observation), whereas a stellar fluctuation caused by a spot, a facula, or any other event is mainly chromatic. A sunspot, for example - where we can crudely model the effects such as a local cooling of the surface - is extremely chromatic: the relative variations in intensity vary by a factor of two between the blue and the red.

• It enables us to explore the image spot spatially. Bearing the resolution of the instrument in mind (the image spot covers about 300 arcsec2), we should be able to distinguish a transit across a target star in an eclipsing binary if the components are spatially separate, by seeing if the effects are equally present in the three photometric channels.

After correction for the various sources of instrumental bias (sky background, electronic gain, periodic perturbations, fluctuations in the satellite's guidance, etc.) it should be possible to obtain a photometric curve where the residual noise is less than 1.4 times the photon noise of the target over a period of 8.5 min. To obtain such a performance, it is necessary to eliminate light scattered by the Earth. To do this, CoRoT's telescope has a shield that contains several optical baffles that block scattered light. The baffling reduction transmission of scattered light to less than 10-13.

The photometric curve, if planetary transits are present, shows regular extinctions over the period of the transit (Fig. 8.10). The data processing computers are able to identify transits and detect objects down to a size of about two Earth radii. For each candidate transit, it is possible to determine the period (and thus distance from the star), the duration and the intensity of the transit (and thus the size of the object). Several methods are used by the computers for detecting and analyzing the transits: variable-pitch Dirac comb, least-squares fit, morphological analysis, wavelet analy-

Fig. 8.10 Photometric curve corrected for various errors for a star of solar type and of magnitude 14, representing the transit of an object 0.5 Jupiter diameters, with a period of 10.54 days

sis, etc. The results from the different data-processing algorithms for investigating transits primarily depend on:

• the depth of the transits

• the stability of the stellar photometry

The various algorithms are currently being evaluated in the CoRoT community. Finally, we should also mention that a comprehensive, ground-based programme, following the transit candidates, will be necessary to confirm the planetary nature of the events. The follow-up programme, using various methods (high-angular-resolution photometry, and mainly radial-velocity measurements) will allow us to eliminate from the list of candidate transits 90 per cent of the alerts consisting of:

• transits of eclipsing binaries, of greater magnitude and in the background to the principal target, which would mimic a transit of the main target

• transits of grazing binaries where the principal target is an eclipsing binary.

CoRoT should enable us to identify several tens of hot Jupiters, or other objects with sizes comparable to Uranus and Neptune. Equally, CoRoT should enable us to carry out the first statistical analyses of objects whose size is equal to a few times that of the Earth, provided several tens of such objects are detected. An accurate estimate of the number of detections of this type of object is difficult to obtain, because their existence is not yet unequivocally confirmed.

Fig. 8.11 Artist's impression of the Kepler mission (NASA)

Was this article helpful?

0 0

Post a comment