1 Introduction

Two space projects, Darwin (Léger et al. 1996) and TPF (Beichman et al. 1999) are presently under study to search for terrestrial exoplanets and signatures of life in their atmosphere. Molecular oxygen ${\rm O}_{2}$ and the closely related species ozone ${\rm O}_{3}$ will be looked for as tracers of oxygenic photosynthesis, by analogy with Earth. We address here the question of the real meaning of ${\rm O}_{2}$ and ${\rm O}_{3}$ detections on a planet, and the possibility of "false positive detections'' (i.e. cases where abiotic (photochemical) processes might mimic the biogenic production of ${\rm O}_{2}$ and ${\rm O}_{3}$). Our purpose is to answer the following questions:

-

What level of atmospheric ${\rm O}_{2}$ can be built up in the absence of life from photochemical processes?

-

Can abiotic ${\rm O}_{2}$ results in a detectable signature of ${\rm O}_{3}$ in the mid-infrared range (5-20 $\mu$m)?

-

Can we unambiguously identify in those cases the abiotic origin of ${\rm O}_{2}$ and ${\rm O}_{3}$?

To investigate the risk due to abiotically produced ${\rm O}_{2}$, we have selected three types of atmospheres, which present one or both of the main photochemical processes for its production: photodissociation of ${\rm CO}_{2}$ and ${\rm H}_{2}$O. These types are humid ${\rm CO}_{2}$ dominated atmospheres (cases A-D), dry ${\rm CO}_{2}$ atmospheres, (E, F) and a humid atmosphere with permanent water supply and hydrogen escape (G).

The first type we will consider is ${\rm CO}_{2}$ dominated atmospheres with water vapor rising from a moist saturated surface. This very important case is believed to correspond to the early stages of the atmosphere of all the three terrestrial planets Venus, Earth and Mars. It also includes Mars at the present time, and a possible future, terraformed state of this planet. We will study 4 cases with $P_{{\rm CO}_{2}}$ from 6 to 3200 mbar. Sensitivity of the model to the parameters and processes included was checked in case A (Mars).

The second type is a dry ${\rm CO}_{2}$ atmosphere. We consider this type because, due to the lack of chemical destruction path using the OH radical, it is expected to maximize the abiotic ${\rm O}_{2}$ and ${\rm O}_{3}$ production as compared to the humid ${\rm CO}_{2}$ atmosphere. This case is the most favorable for abiotic ${\rm O}_{2}$ production. The 2 cases considered are for $P_{{\rm CO}_{2}}$ = 4000 (large ${\rm O}_{2}$ content) and 50 mbar (compromise between ${\rm O}_{2}$ production and ${\rm O}_{3}$ detectability).

The third case, suggested by J. Schneider, is an atmosphere with a permanent ${\rm H}_{2}$O influx at high altitudes, where water photolysis can occur. When this delivery, provided by comets or small ice particles of cometary origin, is associated with hydrogen escape, this can lead to an ${\rm O}_{2}$ accumulation.

After reviewing the cases of abiotic ${\rm O}_{2}$ and ${\rm O}_{3}$ production in the solar system today, we present the ${\rm CO}_{2}$ and ${\rm H}_{2}$O photodissociation processes responsible for it, and the sources of ${\rm CO}_{2}$ and ${\rm H}_{2}$O in the present and the past of the terrestrial planets (Sect. 3.3). The numerical tools developed to study these present and past atmospheres of terrestrial planets in the solar system and around other stars are described in Sect. 4: the photochemical model PHOEBE and the infrared spectrum computation tool LWT. The results of our simulations are presented Sect. 5.

Two processes not taken into account in previous works prove to be important: the retroaction of chemistry on the vertical temperature profile, and the disappearance of the ozone mid-IR signature for high $P_{{\rm CO}_{2}}$ (above $\sim$50 mbar). These results are discussed in Sect. 6 with respect to the remote detection of life signatures through Darwin or TPF. We conclude from these results that the real risk of "false positive'' detection of ecosystems appears very limited, which makes the triple ${\rm O}_{3}$- ${\rm H}_{2}$O- ${\rm CO}_{2}$ IR signature a robust indicator of life.