Since the Mariner 9 and Viking era in the 1970s, the atmosphere of Mars has been repeatedly monitored by orbiters, landers and rovers. We now have a very good knowledge of the seasonal atmospheric evolution (chemical composition, thermal and cloud structure, dynamics, photochemistry). Global climatic models have been developed, especially at the Laboratoire de météorologie dynamique (LMD, Paris; Forget et al., 1999); they generally provide very good data fitting, at least below an altitude of about 50 km.

After the negative results of Viking regarding the presence of organics at the surface of Mars, the question was raised about the nature of the agent responsible for the oxidation of the surface. Hydrogen peroxide was suggested (Atreya and Gu, 1995; Clancy and Nair, 1996), although the amounts predicted by photochemical models appeared by far insufficient to destroy all organics (in particular those of meteoritical origin). Hydrogen peroxide was unsuccessfully searched for (Encrenaz et al., 2002; Krasnopolsky et al., 1997) until it was discovered through two ground-based experiments in 2003, first in the submillimeter range (Clancy et al., 2004), then by imaging spectroscopy (Encrenaz et al., 2004). The latter dataset is described below.

TEXES (Texas Echelon Cross-Échelle Spectrograph) is a high-resolution imaging spectrometer (Lacy et al., 2002), operating between 5 and 25 μm, that combines both very high spectral resolution (R = 80,000 at 8 μm in the high-resolution mode) and good spatial resolution (about 1 arcsec). At 8 μm, the 1.1 × 8 arcsec slit of the instrument, aligned along the north–south celestial axis, is moved from west to east with 0.5 arcsec steps. The size of Mars typically ranges between 6 and 15 arcsec. Two scans (north and south) are usually necessary to map the whole planet, which takes about 15 min. We choose an interval of about 7 cm⁻¹ of bandwidth, centered at 1240 cm⁻¹, which contains several transitions of the strong ν₆ H₂O₂ band. In June 2003, the areocentric longitude was 206° (southern spring), corresponding to a high H₂O₂ content according to photochemical models. All H₂O₂ transitions were detected, together with CO₂ transitions (both strong and weak) and a couple of HDO transitions.

In order to map the H₂O₂ mixing ratio on the Martian disk, we selected a H₂O₂ doublet which brackets a weak CO₂ line at 1241.6 cm⁻¹, and we simply compute the ratio of the line depths of the H₂O₂ transitions versus the CO₂ transitions. Radiative transfer calculations show that this line depth ratio is a very good indicator of the H₂O₂/CO₂ mixing ratio as it eliminates, to first-order, effects due to geometry and thermal structure. Fig. 1 shows the spectrum of the H₂O₂ doublet observed in June 2003 (Ls = 206°) at the time of the first detection, and the first H₂O₂ map recorded for this season.

Since 2003, we have been observing H₂O₂ in November 2005, June 2008, October 2009 March 2014, and July 2014, to monitor its behavior as a function of time and season (Encrenaz et al., 2005, 2008, 2012a,b). As an example, Fig. 2 shows the H₂O₂ map retrieved with this method in March 2014 (Ls = 96°), compared with the H₂O₂ map calculated using the LMD-GCM (Forget et al., 1999), including a photochemical model developed at LATMOS (Lefèvre et al., 2008). It can be seen that the H₂O₂ distribution over the disk is far from being uniform, and globally well fitted by the model. The disk-integrated H₂O₂ mixing ratio typically ranges between a few ppb and about 30 ppb. The conclusion of these observations is that the H₂O₂ behavior is well reproduced by the models. In particular, Fig. 3 shows the evolution of the H₂O₂ content as a function of Ls, compared with various photochemical models. It can be seen that the observations favor the LMD-IPSL model including heterogeneous chemistry on water ice grains (Encrenaz et al., 2012a,b, 2015a; Lefèvre et al., 2008). The same behavior is observed for the seasonal distribution of ozone on Mars, as well as in the terrestrial atmosphere, where the loss of polar ozone in the Earth's stratosphere has been explained by interactions between gaseous chemical species and ice cloud particles (Lefèvre et al., 2008).

Using weak HDO transitions in the 1237–1244 cm⁻¹ range, we have simultaneously retrieved maps of water vapor, using HDO as a tracer of H₂O. We made the assumption that D/H was constant over the Martian disk, with a value of 5.0 times the terrestrial value (Krasnopolsky et al., 1997). This is actually a first-order approximation, as theoretical calculations predict variations associated, in particular, with condensation (Montmessin et al., 2005). We assume a constant D/H value in the lack of precise measurements. The water vapor seasonal cycle is generally well understood on Mars since the Viking measurements and, more recently, the TES monitoring aboard the MGS orbiter (Smith, 2002, 2004), and maps of the water vapor content versus latitude and areocentric longitude are well represented by GCMs. However, no information has been acquired so far on two-dimensional maps of water vapor that would give, for any season, the H₂O distribution versus latitude and longitude, or versus latitude and local hour. This information requires imaging spectroscopic capabilities from some distance, as from the ground or Earth orbit, and has been obtained in the thermal regime for the first time with TEXES. In general, the agreement between the TEXES maps of HDO and H₂O maps predicted by the GCM is satisfactory. An example is shown in Fig. 4 for Ls = 80°, just before the northern summer solstice, when water vapor is expected to be maximum. There is a very good agreement between the data and the model (Encrenaz et al., 2010).

In summary, monitoring the Martian atmosphere using high-resolution ground-based imaging spectroscopy has allowed us to monitor the seasonal variations of hydrogen peroxide and water vapor, and to give indication for H₂O₂ photochemical models including heterogeneous chemistry. It has also confirmed that the GCM provides a very good understanding of the Martian climate in terms of chemical composition and thermal structure.

While both Mars and Venus have in common a CO₂-dominated atmosphere, with N₂, Ar, CO, H₂O as minor species, the atmosphere of Venus presents, as compared to the Martian atmosphere, an extreme case with very high surface pressure (90 bar) and temperature (730 K). In addition, sulfur (most likely outgassed from the interior) plays a major role in the photochemistry and dynamics of Venus. The surface of Venus is hidden by an opaque cloud deck, rich in sulfuric acid, extending to altitudes of about 50–65 km.

In contrast with the case of Mars, the atmosphere of Venus is far from being understood in terms of vertical transport and dynamics. Even the mass budget of sulfur is not understood. Indeed, both SO₂ and H₂O are relatively abundant (150 ppm and 30 ppm respectively) in the lower troposphere, below the cloud deck (Bézard and de Bergh, 2007). Above the clouds, these abundances drop down to about 100 ppb and 1 ppm, respectively (Marcq et al., 2011; Zasova et al., 1993). Assuming that most of water combines with SO₂ to form H₂SO₄, only one fifth of the total SO₂ content is used in this process. So in which form is the remaining sulfur? Most likely in the form of aerosols, but their exact nature – and thus their optical and radiative properties – is not known. Above the cloud deck, another surprising result has been found. In the lower mesosphere, the SO₂ mixing ratio is known to decrease with altitude, down to a few ppb above the clouds (Belyaev et al., 2012). Then, it increases again in the upper mesosphere, at a level of about 90 km, where SO and SO₂ have both been detected in the submillimeter range, at a level of a few tens of ppb (Sandor et al., 2010, 2012). Thus, a second reservoir, probably also made of sulfur aerosols, must be present at these altitudes.

Since 2011, we have been observing water and sulfur species on Venus on different occasions and at different altitudes. In November 2011, we have used the Atacama Large Millimeter Array (ALMA) facility to obtain submillimeter maps of minor species in the upper atmosphere, at an altitude of about 90 km. In January and October 2012, we have used TEXES at IRTF to study the spatio-temporal variations of SO₂ and H₂O (through its proxy HDO) at the cloud-top and within the clouds. These observations have shown evidence of strong spatial and temporal variations of the sulfur species, both at the cloud-top and in the upper mesosphere. These variations are presently not understood by the photochemical or dynamical models.

3.1 ALMA observations the upper mesosphere

Fig. 5 shows the variations of SO, SO₂ and HDO on 14 and 15 November. The SO map is of best quality because, although SO is less abundant than SO₂, the SO transition is intrinsically much stronger. All three species show significant variations over the disk and over a timescale of a day. In the case of HDO, the variations could be associated with fluctuations of the mesospheric temperature; in the case of the sulfur species, the origin of these variations is unknown.

3.2 TEXES observations of the lower mesosphere

Fig. 6 summarizes the observations of SO₂ and HDO obtained in January 2012 over a timescale of 3 days, at the cloud-top of Venus (z = 65 km), using high-resolution spectra around 1350 cm⁻¹ (λ = 7.4 μm). It can be seen that SO₂ shows strong spatial variations over the disk, and significant changes within a day; in contrast, the water distribution is much more uniform and constant with time. The maximum mixing ratio of SO₂ is about 100 ppb, and the H₂O mixing ratio, constant over the disk, is about 1.5 ppm (Encrenaz et al., 2012).

In July 2016, the Juno space mission, launched by NASA in August 2011, will encounter Jupiter. One of its main objectives is the measurement of the O/H ratio in the deep interior of the planet, a key measurement for understanding the nature of the planetesimals that formed the planet (Atreya et al., 1999). This measurement will be made in the radio range where water has a major contribution to the continuum spectrum (Janssen et al., 2005). However, other minor species (NH₃, PH₃, H₂S…) also have contributions in this spectral range; in particular, ammonia is a major radio absorber.

Ammonia and phosphine are both tracers of Jovian dynamics. PH₃ is a disequilibrium species that is not expected to be observed in the Jovian spectra, as it should react with H₂O in the deep troposphere to form P₄O₆. Its presence in the upper troposphere is due to vertical motions that carry the molecule upward on timescales shorter than the destruction lifetime of the molecule (Lewis, 1997). The presence of NH₃, in contrast, is predicted by thermo-chemical models. Its abundance is constrained by reactions with H₂O and H₂S to form NH₄OH and NH₄SH clouds at about 2 bar, then by NH₃ condensation above the 0.5 bar level (Atreya, 1986). Both ammonia and phosphine are thus tracers of Jovian dynamics; it is important to measure their abundances in the troposphere of Jupiter, and to monitor their possible temporal variations.

In February 2014, we have started an observing program to monitor the abundances of NH₃ and PH₃ at different atmospheric levels of the Jovian troposphere (with pressure levels ranging from 0.1 bar to a few bars). Using the same technique as that developed for our Mars and Venus observations, we have mapped the Jovian disk in three different spectral ranges probing different pressure levels: 4.64 μm (P = 3–5 bar); 8.83 μm (P = 0.3–0.5 bar); 10.47 μm (P = 0.1–0.3 bar). Preliminary results are given in Fig. 7, which shows maps of the line depths of two specific transitions of NH₃ and PH₃ at the 0.3–0.5 bar pressure level. It can be seen that both maps are distinctly different, illustrating that both molecules are sensitive to different dynamical mechanisms; in addition, the PH₃ abundance seems to be enhanced at high northern latitudes, a trend that is also observed at other pressure levels (Encrenaz et al., 2015b). This work will be pursued in 2015 and beyond, with the objective of mapping the abundances of these two species as a function of time, longitude, and latitude at the time of the Juno mission.