2 Observational search for exoplanets and signatures of life

The existence of other planetary systems, discussed already by ancient greek philosophers, has only been proved very recently, first around pulsars (Wolszczan & Frail 1992), then around solar type stars (Mayor & Queloz 1995). In this latter case, the discovered planets are massive (of the order of a Jupiter mass) and likely to be giant gaseous planets, which at least proved to be the case for HD 209458b (Charbonneau et al. 2000). Up to now, 76 planets with masses above 0.2 Jupiter mass have been found around 68 main-sequence stars. Recently, the discovery of 15 "free floating'' massive planets (i.e. not orbiting around a star) has been announced in a star forming region (Zapatero Osorio et al. 2000).

The search for smaller planets of terrestrial type rises considerable scientific and philosophical interest, but is technically much more difficult. Among the ambitious projects aimed at their detection, the space observatories under study by ESA (Darwin; Léger et al. 1996) and by NASA (TPF; Beichman et al. 1999) are based on direct detection in the near infrared through a "nulling interferometer'' (Bracewell 1978). Other concepts for indirect search of terrestrial planets are the space projects Corot (Rouan et al. 1998), Eddington (Deeg et al. 2000) and Kepler (Borucki et al. 1997).

As soon as photons coming from the planet can be distinguished from those coming from the star, a spectral analysis is feasible within the available signal-to-noise ratio. The physical and chemical properties of the planets and their atmosphere can be studied. The recent detection of the atmosphere of a "hot Jupiter'' in absorption during a transit is the first spectral information gained on the atmosphere of an exoplanet (Charbonneau et al. 2001). As life on Earth has strongly modified the planet (atmosphere, oceans, surface), can we use this fact to distinguish spectroscopically the presence of similar life on another planet? In the particular case of Earth, ${\rm O}_{2}$ is fully produced by the biosphere: less than 1 ppm coming from abiotic processes (Walker 1977). In a famous paper, Sagan et al. (1993) analyzed a spectrum of the Earth taken by the Galileo probe, searching for signatures of life. They conclude that the large amount of ${\rm O}_{2}$ and the simultaneous presence of ${\rm CH}_{4}$ traces is suggestive of biology. Moreover, the detection of a widespread red-absorbing pigment with no plausible mineral origin supports the hypothesis of bio-photosynthesis. Owen (1980) suggested to search for ${\rm O}_{2}$ whereas Angel (1986) has suggested to consider rather ${\rm O}_{3}$ and its signature in the mid-infrared range around 10 $\mu$m. Paetzold (1962) and Léger et al. (1993) have investigated the feature of ${\rm O}_{3}$ as a tracer of ${\rm O}_{2}$ in the atmosphere of exoplanets. This concept is at the root of the spectroscopic design of Darwin, as proposed to ESA by Léger et al. (1996), and of TPF (Terrestrial Planet Finder, Beichman et al. 1999).

For Darwin, all the volatile ingredients and products of oxygenic photosynthesis lead, directly or indirectly, to infrared spectral signatures in the instrument's window:

$$\begin{eqnarray*}2{\rm H}_{2}{\rm O}^{*} + \mbox{${\rm CO}_{2}$ }\ + {\rm photon... ...}{\rm H}_{2}{\rm O} + \mbox{${\rm H}_{2}$ O}\ + {\rm O}_{2}^{*}. \end{eqnarray*}$$

(As pointed out by the asterisks, photosynthetic ${\rm O}_{2}$ is made of oxygen atoms extracted from water.)

The absorption lines of ${\rm H}_{2}$O, ${\rm CO}_{2}$ and, as a tracer of the oxygen, ${\rm O}_{3}$, will be simultaneously searched for: this is what we call the triple signature. For the project TPF, an alternative option is to search for ${\rm O}_{2}$ in the visible range (Nisenson & Papaliolios 2001).

To validate the concept of Darwin and TPF with respect to the search for extraterrestrial life, it is compulsory to understand both the risk of a "false positive'' result (where an abiotic process is mistakenly interpreted as a signature of life) and of a "false negative'' result (when a planet contains life forms but does not produce any characteristic signature). When using the criterion of ${\rm O}_{2}$ or ${\rm O}_{3}$, false negative cases are clearly possible: on Earth itself, almost 2 Gyr are thought to separate the first traces of life -3.7 Gyr BP (before present) (Rosing 1999) and maybe earlier (Mojzsis et al. 1996) - from the epoch where ${\rm O}_{2}$ has reached detectable levels in the atmosphere - 2.3-1.9 Gyr BP (Holland 1994). Furthermore, even oxygen-rich atmospheres can fail to lead to a detectable signature of ozone in the Darwin spectra (Selsis 2000a, 2002).