The detection of extra-solar planetary systems is one of the great challenges to contemporary astronomy. Adapted to the ambitious task, practically all the main observational techniques seem in principle to be capable to contribute to the aim. Different planet-detection techniques have however different "discovery spaces" which define the subsets of extrasolar planets with the best chances of discovery. Observations from the ground and from space, even from the stratosphere were proposed. The status of proposals and methods as of the end of 1992 was summarized in the proceedings of the conference on Planetary Systems: Formation, Evolution, and Detection (Burke et al. 1994). Recent reports on the subject were given by Schneider (1996a, b), e.g. .
The proposals refer to a broad wavelength range from optical and infrared to millimetre bands. In cases were the planetary body itself is to be observed, the contrast to the parent star is smaller at longer wavelengths.
Direct imaging using sophisticated coronographic apodization methods, high-resolution IR arrays, speckle methods, and aperture synthesis in the radio region aim to unveil planet-like companions close to carefully selected target stars. Other methods are directed to measure and model the morphology and physical parameters of circumstellar structures related to planetary systems in their earliest phases. Eclipsing binaries offer the chance that the lines of sight are sufficiently close to the orbital plane of the suspected planets. The transits of inner planetary bodies might be detected from monitoring photometry as a measurable modulation of the stellar light. Doyle et al. (1996) reported on a dozen candidate transit events in the CM Dra system, observed within their Transit of Extrasolar Planets programme. A very promising search technique is offered by gravitational lensing. Planets (down to ) passing near the line of sight to distant stars will cause a brief brightening of the stellar light. Further potentials are in the spectrophotometry. Respective methods aim at the detection of characteristic atmospheric bands, the A band of O2 at 760 nm and the O3 band at 9.6 µm, e.g. In principle one can try to observe the bands directly in the planet's spectrum separated from the stellar one, or as a transient phenomenon in the stellar spectrum when the planet occults the star and stellar radiation traces the planetary atmosphere. The most promising and meanwhile successful methods are given by accelerometry and astrometry. These indirect methods are making use of the gravitational interaction between the central star and its companion(s), which causes a motion of the components around the common barycentre. This should be reflected in a varying radial velocity and a positional variation of the observable star. In the past, periodic terms in the proper motion were observed for several stars, but the accuracy of classical methods is not sufficient to detect companions of planet-like masses. Here interferometric methods are demanded. High-resolution spectroscopy was successful, however, to definitely detect companions of masses of the order of the Jupiter mass in the case of a few nearby stars (Mayor & Queloz 1995). Orbital motion is responsible for the characteristic pattern in the arrival times of pulses from the PSR 1257+12 pulsar observed by Wolszczan & Frail (1992) as well as for the drift in the timing of eclipsing binaries (see Doyle et al. 1996).
This paper is concerned with the detection of spectroscopic features assumed to be typical of the atmospheres of Earth- and Jupiter-like planets in general. Especially in the infrared spectral region with the enhanced contrast to the emission of the central star, this might be a powerful tool to detect extrasolar planets. IR spectra of solar system planets (see Burke 1992) show prominent absorption and emission features due to atmospheric constituents. Besides pointing to the existence of an atmosphere, these features could help to find out whether atmospheres of extrasolar planets can be classified as CO2 - or CH4 -dominant ones, as it is the case with Earth- and Jupiter-like planets, resp. The detection of ozone would be a strong indicator of Earth-like biological activity. Methane in coexistence with ozone would give us an additional argument for life as a "chemical reactor". Without biological activity methane would be oxidized and thus disappear.
Even in relatively late phases when planet-like bodies have already formed, the disk structure surrounding a star may still contain such an amount of diffuse material that the detection of certain atmospheric features of the extrasolar planet is severely hampered. The features might be washed out by emission from circumstellar material or are hidden in the noise from radiation scattered or reemitted by circumstellar dust particles along the line of sight to the planets. To evaluate the amount of dust which does not significantly influence the atmospheric bands, Monte Carlo simulations of radiative transfer were performed in configurations consisting of an Earth- and a Jupiter-like planet embedded in a debris dust disk around a solar-type central star. We restricted our investigation to the O3 and bands at m and the CO2 band at m (see Fig 3), which interfere with circumstellar silicate and methanol features at m and m. The calculations were done for two different grain models, total dust masses and for different viewing angles. In one of the models the presence of gaps in the disk due to perturbations from the orbiting planets is assumed.
In Sect. 2 the physical properties of our models are described and the results are presented and discussed in Sects. 3 and 4, resp.
© European Southern Observatory (ESO) 1997
Online publication: April 28, 1998