Long before the arrival of the Mariner and Viking spacecraft, astronomers knew from Earth-based spectroscopy that the Martian atmosphere was quite thin and composed primarily of carbon dioxide. In 1964, Mariner 4 confirmed these results, finding that the atmospheric pressure is only about 1/150 the pressure of Earth's atmosphere at sea level and that carbon dioxide makes up at least 95 percent of the total atmosphere. With the arrival of Viking, more detailed measurements of the Martian atmosphere could be made. Its composition is now known to be 95.3 percent carbon dioxide, 2.7 percent nitrogen, 1.6 percent argon, 0.13 percent oxygen, 0.07 percent carbon monoxide, and about 0.03 percent water vapor. The level of water vapor is quite variable. Weather conditions encountered by Mars Pathfinder were quite similar to those found by Viking 1.
As the Viking landers descended to the surface they made measurements of the temperature and pressure at various heights. The results are shown in Figure 10.15. The Martian atmosphere contains a troposphere (the lowest-lying atmospheric zone, where convection and "weather" occur), which varies both from place to place and from season to season. 
(Sec. 7.2) The variability of the troposphere arises from the variability of the Martian surface temperature. At noon in the summertime, surface temperatures may reach 300 K. Atmospheric convection is strong, and the top of the troposphere can reach an altitude of 30 km. At night, the atmosphere retains little heat, and the temperature can drop by as much as 100 K. Convection then ceases and the troposphere vanishes.
Figure 10.15 Structure of the Martian atmosphere, as determined by Viking. The troposphere rises to an altitude of about 30 km in the daytime. It occasionally contains clouds of water ice or, more frequently, dust during the planetwide dust storms that occur each year. Above the troposphere lies the stratosphere. Note the absence of a higher-temperature zone in the stratosphere, indicating the absence of an ozone layer.
On average, surface temperatures on Mars are about 50 K cooler than on Earth. The low early-morning temperatures often produce water-ice "fog" in the Martian canyons (Figure 10.16). Temperatures in the stratosphere are low enough for carbon dioxide to solidify, giving rise to a high-level layer of carbon dioxide clouds and haze.
Figure 10.16 Fog in the Martian canyons, photographed by the Viking orbiter. As the Sun's light reaches and heats the canyon floor, it drives water vapor from the surface. When this vapor comes in contact with the colder air above the surface, it condenses again, and a temporary water—ice "fog" results.
For most of the year, there is little day-to-day variation in the Martian weather: The Sun rises, the surface warms up, and light winds blow until sunset, when the temperature drops again. Only in the southern summer does the daily routine change. Strong surface winds (without rain or snow) sweep up the dry dust, carry it high into the stratosphere, and eventually deposit it elsewhere on the planet. At its greatest fury, a Martian storm floods the atmosphere with dust, making the worst storm we could imagine on Earth's Sahara Desert seem inconsequential by comparison. The dust can remain airborne for months at a time. The blown dust forms systems of sand dunes similar in appearance to those found on Earth.
Although there is some superficial similarity in composition between the atmospheres of Mars and Venus, the two planets obviously have quite different atmospheric histories—Mars's "air" is over 10,000 times thinner than that on Venus. As with the other planets we have studied, we can ask why the Martian atmosphere is as it is.
Presumably, Mars acquired a secondary outgassed atmosphere quite early in its history, just as the other terrestrial worlds did. Around 4 billion years ago, as indicated by the runoff channels in the highlands, Mars may have had a fairly dense atmosphere, complete with blue skies and rain. Despite Mars's distance from the Sun, the greenhouse effect would have kept conditions fairly comfortable, and a surface temperature of over 0° C seems quite possible. Sometime during the next billion years, most of the Martian atmosphere disappeared. Possibly some of it was lost because of impacts with other large (Moon-sized) bodies in the early solar system. More likely, the Martian atmosphere became unstable, in a kind of reverse runaway greenhouse effect.
On Venus, as we have seen, the familiar greenhouse effect ran away to high temperatures and pressures. (Sec. 9.4) Planetary scientists theorize that the early Martian atmosphere also ran away, but in the opposite direction. The presence of liquid water would have caused much of the atmospheric carbon dioxide to dissolve in Martian rivers and lakes (and oceans, if any), ultimately to combine with Martian surface rocks. Recall that on Earth most carbon dioxide is presently found in surface rocks. Calculations show that much of the Martian atmospheric carbon dioxide could have been depleted in this way in a relatively short period of time, in perhaps as little as a few hundred million years, although some of it might have been replenished by volcanic activity, possibly extending the "comfortable" lifetime of the planet to a half-billion years or so.
As the level of carbon dioxide declined and the greenhouse-heating effect diminished, the planet cooled. The water froze out of the atmosphere, lowering still further the level of atmospheric greenhouse gases, and accelerating the cooling. (Recall from Section 9.4 that water vapor also contributes to the greenhouse effect.)
Since those early times the overall density of the Martian atmosphere has continued to decline through the steady loss of carbon dioxide, nitrogen, and water vapor as solar ultraviolet radiation in the upper atmosphere splits the molecules of these gases into their component atoms, providing some with enough energy to escape. The present level of water vapor in the Martian atmosphere is the maximum possible, given the atmosphere's present density and temperature. Estimates of the total amount of water stored as permafrost or in the polar caps are quite uncertain, but it is likely that if all the water on Mars were to become liquid, it would cover the surface to a depth of several meters.
