html version of Poster displayed at JENAM-95 meeting in Catania, Italy 9/95

The TEP Network - Searching for Transits of Extrasolar Planets

H.J. Deeg, E.L. Martin (Inst. Astrofisica Canarias), J. Schneider, M. Chevreton (Meudon Obsv.), L.R. Doyle (SETI), J.M. Jenkins (NASA-Ames), M. Palaiologou (Skinakas Obsv., Crete), W.B. Lee (Korea Astr. Obsv.)

What do we want to do?

Detect the presence/absence of terrestrial planets around the eclipsing main sequence Binary CM Draconis


By observing the small decrease in light from the binary if a planet transits across the components. The transits are observed performing differential photometry from a network of telescopes, providing continuous coverage as much as possible. A planetary transit will result in unique lightcurves, depending on the phase of the binary and the planetary period (Fig. 1).

Why CM Dra?

CM Dra is unique in its small size (R=0.235 and 0.252 Rsol), its low temperature (spectral class M4.2-M4.4 for both components), its nearly edge-on inclination of 89.8 degrees, and its proximity of 16.9 pc. These properties are advantageous for photometric observations for several reasons: Due to the small size of CM Dra (total surface area: 12% of the sun), the transit of an earth sized planet across one binary component would cause a brightness drop of 0.1%; a Neptune sized body would cause a drop of 1.2%. These drops are an order of magnitude larger than the transits of similar planets across stars of solar size. Second, in systems of low luminosities and temperatures it is expected, that terrestrial planets will form in a region much closer to the central star. If planets comparable to those of the solar system have formed around CM Dra at distances of similar flux, then the orbital period of these planets are as given in the table:

Table 1. Planetary period of terrestrial planet equivalents around CM Dra

 *Note: The minimum stable period around CM Dra is 7.62 days.

What are the limits to the method and what is the detection probability?

The network of 1m Telescopes used for observations in 1994 and 1995 obtains data with typically 0.6% precision in relative flux. This corresponds to the ability to detect Planets with a radius of 2.6 REarth (which cause transits with 0.6% flux decrease) directly, and for continuous observation of more than one planetary period the detection probability would be 100%. In practice, continuos observation can be achieved from space only - for long discontinuous observing runs (covering several planetary periods), the detection probability will be close to 100% too. An example for the phase coverage obatined by relatively short runs is given in Fig. 2. Smaller Planets whose transit signatures may be hidden in the photometric noise (with S/N ~ 0.4 - 1) may be retrieved using a cross correlation with transit event models (Fig. 3). The probabilities of a correct detection will be less than 100%, though, see Table 2). This table also shows the dependence between the light collecting power of a telescope and the sizes of planets that can be detected directly or through cross correlation. If the duty cycle (time that a telescope collects light / total time) is constant, the detectable radius of a planet depends by: RPlanet ~ 1/(Area of Telescope).

Table 2: Detection probabilities averaged for Planets with 8-30 day periods at several telescope sizes and observation durations (from Jenkins et al, 1995):

*For non-continuos coverage of short (<2 months) observing runs, where only a few transits of longer period planets are expected (such as 1 transit of a 30 day planet in a 1 month run), planets may be missed and the probabilities are less predictable. In particular, the rightmost column indicating direct detection (no cross correlation) will be < 100%.

What has been done so far?

Observations of CM Dra in the frame of the TEP network began in spring 1994 at several 1m class telescopes equipped with CCDs (see separate list of observatories). Through July 1995, about 630h coverage in 16000 individual CCD frames has been obtained. Aperture photometry with variable (seeing-dependent) apertures is performed on the CCD frames, using a custom developed suite of IRAF tasks, 'vaphot'. The precision of the differential photometry obtained so far with the 1m class telescopes is about 0.6 % in flux. A composite lightcurve containing most 1994 observations is shown in Fig. 4. The phase coverage is shown in Fig. 5. In both lightcurves, the off-eclipse magnitude of Cm Dra has been set to 0. The lightcurves have been scanned by eye for dips that may be planetary transits, comparison with the phase of CM Dra gives a first rejection, if these dips may be from planetary transits. A second step will then be the application of the cross correlation program, whose outcome is strongly influenced, if these dips are repeated in a fashion compatible with the presence of a planet.

Other things that can be done with the data?

A second method to detect the presence of objects of Jupiter size or larger is given by the observation of slow periodic changes in the period (or Epoch) of CM Dra, from the motion of the Binary - Large Planet system around its barycenter. A Jupiter sized planet sized at a distance of 5.2 AU would cause a periodic change in Epoch of about 11 sec over 8 years. From the reduction of a precisely timed eclipse in May 1995, there is currently a 50 sec discrepancy against the Epoch and period given by Lacy (1977). The lightcurve data provide also unprecedented coverage (for enlargement of primary phase, see Fig. 6), of Cm Dra. Already, several flares have been observed and an improved derivation of the orbital elements of CM Dra will be possible, allowing spin off results like a determination of the helium abundance of this very old system.


Here a link to a more complete set of references related to TEP

Doyle, L.R., Dunham, E.T., Deeg, H.J., Blue, J.E., Jenkins, J.M. (1996), Detectability of Terrestrial Planets: U.C. Lick Observations of CM Dra, J. of Geophys. Res. - Planets, in print

Jenkins, J.M., Doyle, L.R, Cullers, D.K. (1996), A Matched Filter Method for Ground Based Sub-Noise Detection of Terrestrial Planets in Eclisping Binaries: Application to Cm Dra, Icarus, 119,244

Lacy, C.H. (1977), Absolute Dimensions and Masses of the Remarkable Spotted Binary Star CM Dra, ApJ, 218, 444

Schneider, J. and Doyle, L.R. (1995),"Ground Based Detection of Terrestrial Planets by Photometry: The Case for CM Draconis", Earth Moon and Planets, 71,153

If you would like to receive current and future preprints, please send email to

To the TEP home page

Last update: 23/Oct/1997. For comments, send email to Hans-Jörg Deeg at