The Evolution of the Martian Climate

Aaron P Zent. American Scientist. Volume 84, Issue 5. Sep/Oct 1996. 

In spite of the 20-year hiatus since a satellite from Earth last orbited the Red Planet, Mars continues to exert a powerful attraction on the human psyche-not only on the popular imagination but perhaps even more so on the curiosity of those who are carrying out its exploration. Consider the surprising parallel between our current incomplete understanding of the planet and the even more sparse comprehension of scientists and the public 100 years ago. Like the imaginative and well-publicized astronomers and fiction writers of the past century, we are once again prepared to be convinced that Mars once supported life.

During the 20th century, sophisticated telescopic instruments, as well as spacecraft from the United States and the former Soviet Union, have injected a dose of reality into such speculations. Our present readiness to admit to the possibility of life on Mars derives not only from a continuing string of discoveries that demonstrate the tenacity of terrestrial life, but also from an appreciation that the martian climate was once different from today’s frigid, desiccated regime. At the time life appeared on Earth, the martian climate apparently was more clement and perhaps conducive to the same prebiotic chemistry that led to life on Earth. Much of current Mars-related research focuses on the evolution of the martian climate and the processes that have brought us the Mars we see today.

Historical Exploration of Mars

At the turn of the century, Mars was already recognized as the most Earthlike of the planets. Telescopes had revealed familiar features such as polar caps and the change of seasons. Although early “observations” of canals were debunked early in this century, the scientific community believed strongly until the early 1960s that life was possible there. When the first flyby spacecraft reached Mars, astronomy textbooks required a sudden and serious revision. The handful (by today’s standards) of images revealed a surface strikingly similar to the Moon-with craters and craggy inter-crater plains that gave no reason to hope that Mars was, or ever had been, much different from our hostile satellite.

Through sheer perversity, the cameras of Mariners 4, 6 and 7 had passed over some of the oldest and most lunarlike terrain on the planet. It was not until the orbital missions of Mariner 9 in 1971 and Viking in 1976 that a more complete picture of Mars emerged. Textbook writers, who had duly reported Mars to be boring and lunar-like, once again found themselves hung out to dry. The new images revealed an incredibly complex surface. For the first time, geologists had the information they needed to get to work sorting out the geological history of the planet.

One of the first tasks was to map the geological units on the surface and assign relative ages. The procedure for doing this is based on the relative density of impact craters. From our understanding of lunar-crater densities, along with age dating of lunar samples, we know that the flux of impacting objects was quite high early in solar-system history, a period labeled “heavy bombardment.” The cratering flux dropped rapidly after about 3.8 billion years ago, as the small bodies left over from the birth of the solar system were swept up. Since then, a small but steady flux of objects has been gravitationally stirred into planet-crossing orbits by close encounters with more massive objects. Young surfaces show few impact craters, either because the craters have been removed by erosion or because they have been buried by subsequent deposits. The surface of Earth, for example, has a very low crater density because its continuing volcanic history and intense hydrologic cycle constantly renew its surface.

By counting the density of craters on different geologic terrains, a rough .. stratigraphic system has been worked out for Mars. The oldest surfaces, those with the highest crater densities, are referred to as Noachian in age. Hesperian age units are next and date from perhaps the middle half of geologic history. Finally, the Amazonian surfaces are the youngest. The absolute ages of these surfaces must still be guessed at; until samples from known surface units can be retrieved, there is no way to calibrate the crater-frequency curves. Models of cratering flux tied to the lunar example are used to estimate the cratering flux elsewhere in the solar system. Figure 2 shows two published estimates of the absolute ages of these stratigraphic units. The crudeness of our understanding is immediately apparent.

The Current Martian Climate

By contrast with its geologic history, the current martian climate is well known, although what we know often poses puzzling questions. The temperature at the surface has been measured repeatedly, and we can write numerical models that predict surface temperatures accurately, a sign that we understand most of the physical processes that control them. Average temperatures are around -0 degrees Celsius, with summer afternoon highs of 20 degrees and polar nights as cold as -120 degrees.

The principal atmospheric constituent is carbon dioxide (CO2), and the average atmospheric pressure is only 8 millibars, or eight one-thousandths that of the terrestrial atmosphere. Semiannually, some of the atmosphere freezes out at the winter pole, forming large seasonal CO2 caps. Although temperatures are occasionally high enough to melt ice, liquid water never forms because the vapor pressure of the H2O in the atmosphere is so low. Ice that is heated by the sun sublimates directly from ice to vapor without passing through the liquid phase. A glass of water magically transported to Mars and poured onto the surface during a summer afternoon would boil away instantly.

In spite of the instability of water at the martian surface, there is evidence that things were once different. Across the Noachian terrains on Mars are valley systems that are reminiscent of terrestrial river systems. These systems are referred to as valley networks. They are common on the oldest terrains but virtually absent on younger terrains. If water is not stable on the martian surface today, how could river valleys have formed billions of years ago? For that matter, are the valley networks really river valleys?

In addition to the valley networks there are what are called outflow channels. Unlike the networks of coalescing valleys, these are single channels that apparently record the eruption of enormous amounts of water directly from the martian subsurface. Also by contrast with the smaller valley networks, they are not uniformly old but are spread throughout martian history, concentrating in the Hesperian. The volume of these outflow channels is astounding, dwarfing any flow of water known anywhere else in the solar system. How could such enormous floods have occurred? What were the climatic effects of emplacing small seas virtually instantaneously on the surface?

We also see that the polar caps are surrounded by enormous expanses of what is referred to as layered terrain. These are huge deposits of sediments that show layering down to the limit of our photographic resolution, and apparently record the cyclic deposition and erosion of materials. Unlike either the valley networks or the outflow channels, the absence of craters on these deposits demonstrates that they have continued to form up to the very recent past and may still be forming today. What drives the cycle of erosion and deposition?

By combing these observations with an understanding of geology, climatology, astronomy, physics and chemistry, planetary scientists have evolved theories that may explain the progress of Mars from its early infancy to the cold and barren planet we see today We are now on the verge of a new era of exploration of Mars, with no less than eight spacecraft set to orbit and land on Mars before the close of the decade. The experiments on these spacecraft have been carefully designed to test the following hypotheses that we have spent much of the past 20 years debating.

Epochal Climate Change

What do we make of the riverlike valley networks that dissect only the oldest of the martian surfaces? These valleys have flat floors and steep walls. Often their upward reaches begin in amphitheater-like heads that are reminiscent of groundwater sapping channels on Earth. The valley systems are typically hundreds of kilometers in length and from 1 to 10 kilometers wide. They uniformly run downhill and often empty into local depressions such as impact craters. A key observation is that the drainage density, the total length of valleys in a given area, is much lower than in terrestrial river valleys. Whatever process carved these channels, it did not operate as do the processes that lead to river valleys on Earth.

There is other more subtle, yet equally compelling, evidence of different conditions, and it also has to do with erosion. Most of the martian surface shows very low erosion rates; craters appear to be almost fresh, even at densities that make it clear that the surfaces are very old. Yet the very oldest terrains have been severely weathered. The small craters are missing altogether, and larger craters are badly degraded. Their upturned rims have been eroded away, and their floors appear to have received a covering of sediments. The very oldest valleys, too, have a severely degraded appearance, even though only slightly younger valleys look almost pristine.

Clearly there was a period in martian history when erosion took place much more efficiently than it does today. That period ended about the time that the period of heavy bombardment ended, and the period of valley network formation ended shortly thereafter. The word “shortly” to a planetary scientist can mean up to a few hundred million years, which seems like a long time to most people, but which represents no more than several percent of the age of the planet. Nonetheless, the correspondence between the falloff in heavy bombardment and the dramatic decrease in global erosion rates is probably not coincidental.

The first attempts to explain the differences in erosion rates centered on the concept of the atmospheric greenhouse. If there was a period when water flowed across the surface of Mars, producing valley networks and deeply eroded craters, so the thinking went, then conditions were obviously wetter and warmer than they are today. Therefore, something must have kept the martian surface warm early in solar-system history.

The obvious solution is, unfortunately, not very likely. Our understanding of stellar evolution, backed by years of observation, is that stars are not brighter and hotter when young. In fact, stellar evolution theory only digs us a deeper hole: Young stars are known to be dimmer and cooler. The young sun would have been no more than 70 percent as bright as today Therefore, the martian surface, and Earth’s surface for that matter, should have been cooler than the freezing point of water, making rivers and oceans impossible. This problem, known as the faint-young-sun paradox, has long been a puzzle to terrestrial climatologists. We can go to Earth’s geologic record and clearly see that not only were there oceans on Earth during this period, but life had already appeared in those oceans. If the sun was too dim and faint to warm either Earth or Mars to 0 degrees Celsius, how is it that both show unambiguous evidence of liquid water in their geologic records?

There is no answer to this question yet. A great deal of Earth’s geologic record from this period is missing, as a result of erosion and plate tectonics. The crust of the Earth operates like a giant conveyor belt: New crust forms at midocean ridges, travels laterally and is subducted at ocean trenches. The continents, which are made up of the lightest rocks, bob along on the Earth’s crust like so much pond scum and are seldom subducted. But because of the vigorous hydrologic cycle the Earth has enjoyed for the past several billion years, most of the continental rocks dating from 3.8 billion years ago have been eroded and can no longer be examined. Mars, on the other hand, has neither plate tectonics nor a hydrologic cycle. Therefore, the geologic evidence that will one day allow us to resolve the faint-young-sun paradox is available and awaiting our exploration.

If the sun itself could not have warmed the martian surface adequately to permit water to exist in a liquid state, perhaps the atmosphere made the most of the available sunlight through the greenhouse effect. When sunlight strikes a surface, some is reflected away immediately, and some is absorbed by the surface, causing it to grow warmer. If the temperature of the surface is to return to equilibrium with its surroundings, the absorbed energy must eventually be re-emitted as infrared light. The energy of light corresponds to its wavelength: Lower-energy light has a longer wavelength. The wavelength of photons emitted by a surface is determined by the temperature of the surface; hotter surfaces emit higher-energy, shorter-wavelength photons.

In this application, the shorter-wavelength light is the visible light from the sun, and the longer-wavelength, lowerenergy light is infrared radiation. The trick to atmospheric greenhousing is that different gases absorb different wavelengths of light. Therefore, although some gases, such as CO2, do not absorb visible light, they are practically opaque to the infrared radiation emitted by the surface. A photon of visible light travels unimpeded to the martian surface, where it is absorbed and warms the surface. The surface then emits a lower-energy infrared photon, which tries to make its way back out through the atmosphere. If, however, there are constituent gases in the atmosphere, such as CO2 , that can absorb the photon, they will do so, and the temperature of the atmosphere will rise. The atmosphere then emits more energy itself, and both the atmosphere and the surface grow warmer. Finally, the planet warms to the point that equilibrium is established, and the wavelength of photons emitted will tell an observer the temperature of the planet. Gases that are transparent to visible light, but which absorb infrared radiation, are referred to as greenhouse gases. The higher the abundance of greenhouse gases, the greater the greenhouse warming.

The warm, wet Mars hypothesis holds that a heavy atmosphere of some greenhouse gas maintained the surface temperature above the melting point of water early in the planet’s history. Radiative transfer calculations by workers including Jim Pollack at NASA Ames Research Center in California and Susan Postawko, now at the University of Oklahoma, suggested that an atmospheric presence of CO2 on early Mars equivalent to 5 bars (that is, five times the Earth’s current atmosphere) would warm the surface above zero degrees. Carbon dioxide is the favorite greenhouse gas for Mars because, unlike other candidates, it is not rapidly broken down by sunlight and is already known to be present in the martian atmosphere.

This hypothesis, too, has big problems. For starters, it is probably not possible to maintain a heavy CO2 atmosphere above a surface covered with liquid water and rocks. The carbon dioxide would dissolve into the water, react with ions dissolved from the rocks and precipitate out as carbonate deposits such as limestone. This process is so efficient and rapid that it is hard to find a scenario in which a CO2 atmosphere could survive even to 3.8 billion years ago. Furthermore, massive deposits of carbonates have yet to be found.

Even worse, recent calculations by Jim Keating at Pennsylvania State University show that carbon dioxide condenses high in the martian atmosphere. Such CO2 clouds would reflect sunlight back into space and decrease the sunlight reaching the surface. When the atmospheric temperature profile is corrected for the heat of condensation of the CO2 clouds, the surface temperature drops below the freezing point. Thus it appears impossible to raise the surface temperature above zero with a CO2 atmosphere alone. The presence of other greenhouse gases, such as ammonia or methane, might increase the warming, but they are broken down very rapidly by sunlight, and there is no known recycling mechanism.

An alternative explanation that has received relatively little critical attention suggests that high heat flow from the planet’s interior played a major role in establishing the conditions that led to high erosion rates and valley networks. A higher geothermal heat flow would steepen the subsurface temperature gradient, conceivably bringing the zero-degree isotherm close enough to the martian surface to melt near-surface ground ice, which could then flow onto the surface. Nonetheless, even if high heat flow could provide subsurface water, it is not immediately apparent how it could have cut the valley networks.

Mike Carr, at the U.S. Geological Survey in Menlo Park, California, has suggested that the channels form by mass wasting. In mass wasting, geologic materials lose cohesive strength and flow downhill. It is often facilitated by the presence of ice or liquid water in the pore spaces of the material. There are some characteristics of valley networks that suggest an origin by mass wasting, as opposed to groundwater sapping. There is almost never evidence of an actual channel within the valleys; instead they exhibit flat floors that show no evidence of bedforms that indicate running water. In addition, the valleys sometimes show longitudinal ridges running parallel to the valley. It appears likely that the material that moved down these valleys was the valley fill itself, rather than liquid water Perhaps this flow was lubricated by basal groundwater flow brought about by high heat flow.

Although the mass-wasting hypothesis may explain the presence of channels on the oldest terrain, it does not explain the high erosion rates in the Noachian. Perhaps the eroded terrain represents an even earlier episode of global mass wasting. Could the rims of even the largest craters have been removed, and the smaller craters completely erased, by mass wasting? Substantial research must be done to sort out the nature of the earliest martian climate, to understand the physical processes that operated to produce the surfaces we see today, and to explain how the climate evolved from its earliest and most dynamic period to the cold and barren planet we see today.

Transient Climate Excursions

Not only has the martian climate evolved over billions of years, but there has been some argument that Mars has undergone brief climatic excursions that resulted in an atmospheric hydrologic cycle similar to Earth’s.

The chain of observational evidence and climatic inference begins with the outflow channels, which have formed periodically throughout martian history. The amounts of water involved in the formation of these outflow channels is truly staggering. By assuming that the channels were carved by flowing water, which must have carried away the eroded debris, it is possible to estimate the discharge of water that carved the channels. The estimate proceeds by knowing the width and slope of the channels, which can easily be measured. The depth of the flow is subject to more guesswork, but it must have been substantially greater than the height of the bedforms on the bottoms of these channels. Since the bedforms are often tens of meters high, a flow depth of 100 meters is often used. The discharge rates that fall out from these assumptions are on the order of 100 million to a billion cubic meters per second. By contrast, the Mississippi River discharges water from its mouth at about 20,000 cubic meters per second. The torrents that cut the martian channels were then perhaps 50,000 times greater than the Mississippi. Even the channeled scablands in eastern Washington-the result of the largest discharges in Earth’s history-are still 100 times smaller than the discharge rates of the martian outflow channels.

These channels start in disturbed areas, where the ground has apparently collapsed, and flow for hundreds of kilometers across the surface. There are both young and old outflow channels, and, despite water’s instability in the martian atmosphere, the discharges were so enormous that the water would have neither frozen nor evaporated before it had run its course and ponded in small local areas. Mike Carr has suggested that the enormous hydrostatic head necessary to initiate these floods was provided by a layer of ground ice that grew progressively thicker as the early high heat flow from the planet decreased. In time, only the most enormous hydrostatic pressure could disrupt this layer of ice, which may explain why the floods date from the period after the geothermal heat flow had begun to fall off.

The subsequent fate of that water is the subject of the transient-greenhouse hypothesis, advocated by Vic Baker and others at the University of Arizona in Tucson. They contend that the deposition of so much water at the surface, and the amount of heat it represents, would have been enough to tip the global climate into an entirely different regime from the one we see today. They suggest that much of the water from these outflow channels ponded in the northern lowlands of Mars, forming an Oceanus Borealus. As evidence of climatic variations, they point to possible relic shorelines, described by Tim Parker at the Jet Propulsion Lab in Pasadena, California, and to a suite of geologic features that they argue are of glacial origin, including eskers and moraines.

Eskers are sinuous ridges that result from subglacial stream flow, which deposits gravel along the length of its flow. When the glaciers melt away, the ridges are left standing above the surrounding plain. Moraines are the rubble pushed in front of and alongside the glacier. They likewise are left standing after the glaciers are gone. This interpretation has been championed by Jeff Kargel at the U.S. Geological Survey in Flagstaff, Arizona, and it is being hotly debated.

Because glaciers can only form as a consequence of a global atmospheric hydrologic cycle involving precipitation onto the head regions of the glacier, this interpretation has significant implications. It absolutely requires the kind of warm and wet conditions that have been suggested as representative of the earliest martian climate. The mechanism that Baker and colleagues propose for establishing warm, wet conditions on Hesperian Mars involves the release of considerable amounts of carbon dioxide and the establishment of an atmospheric greenhouse. It is their suggestion that the flood of water from the outflow channels had two effects that allowed geologically brief climatic excursions to warm and wet conditions.

The first effect was melting of the polar ice caps. Although the surface of the caps is water, it is possible, as Bruce Jakosky at the University of Colorado has pointed out, that there is a considerable amount of carbon dioxide trapped in polar ice. That CO2 would have been released into the atmosphere, increasing greenhouse warming. Recent calculations by Mike Mellon at NASA Ames suggest, however, that no more than about 0.1 bar of carbon dioxide could be stored in this way. This amount of COz is well short of the amount needed to produce the required climatic warming.

In addition to carbon dioxide from the caps, COz adsorbate on the surfaces of martian soil particles would be released into the atmosphere when wetted. Adsorption is the result of unbalanced charges on a surface, and the consequence of this imbalance is an accumulation of gas molecules at the surface. We have measured the adsorption of carbon dioxide onto a variety of minerals at martian conditions in our laboratory and believe we understand well how much adsorbed C02 a teaspoon of soil holds on Mars. What we do not know is how much soil is actually present on Mars. Without that knowledge, we cannot predict the total amount of carbon dioxide that would be released into the atmosphere by flooding at the surface. We do know, however, that there is only about 10 meters of finely ground material on the Moon. Even if the depth of martian fines is 100 meters, the total amount of C02 available in that reservoir is insufficient, in combination with the 0.1 bar from the caps, to achieve the required amount of greenhouse heating. The possibility of periodic and transient climate excursion to a warm, wet condition remains to be tested in upcoming Mars missions.

Quasiperiodic Climate Change

For at least much of the past billion years, there is evidence that yet another climatechange mechanism may have been operating on Mars. Surrounding both the north and south polar caps are deposits that have been termed layered terrain. The layering in these deposits is evident in alternating bands of light and dark material, the thickness of which is at least 30 meters in the highest-resolution Viking imagery. Jeff Plaut at the Jet Propulsion Laboratory has estimated, based on crater counts, that the age of the layered deposits is on the order of several hundred million years.

The composition of the layered terrains has been hotly debated. The obvious inference, because of the light layers and their restriction to the polar regions, is that the layers represent compositional differences or even variations in grain size.

Another indication that the martian climate has varied with time is the asymmetry of the polar caps. The north pole has a residual polar cap (that is, a cap left at the height of summer) composed of water at about -70 degrees Celsius. For reasons not at all understood, the residual summer cap in the southern hemisphere is composed of CO2. Since CO2 is the principal atmospheric component, with a pressure of about 8 millibars, and since the temperature of the cap must be controlled by equilibrium between the cap and the atmosphere, we know that the temperature of the south polar cap, at least during the most recent years, has been around -120 degrees. Therefore, atmospheric water will condense on the south polar cap, because of its colder temperature, and the north polar cap will supply more water molecules to the atmosphere to maintain the equilibrium vapor pressure at -70 degrees. Eventually, all of the water in the north polar cap will transfer to the southern hemisphere. Since this has not yet happened, we know that the current configuration of the polar caps has not endured long.

As mentioned above, the rhythmic layering in the polar layered terrain suggests that the climatic conditions have been oscillating for some time. The only possible climatic controls that likewise oscillate are orbital parameters. The orbital parameter that most strongly affects climate is the obliquity, which is a measure of the angle between the rotational axis and the normal to the orbital plane. The obliquity of Mars varies over time scales of about 100,000 years, because of gravitational interactions with other planets, principally Jupiter. Other orbital parameters-such as eccentricity, the departure of the orbit from perfect circularity, or perihelion, the season at which Mars is closest to the sun-also play roles in climatic variations. For the present purpose, we will restrict the discussion to the effects of obliquity variations.

The current obliquity of Mars is 22.5 degrees, although it is thought to oscillate between 10.8 degrees and 38 degrees. When the obliquity is low, relatively more sunlight falls on the equatorial regions of the planet, and the equator-to-pole temperature gradient increases. When the obliquity is high, sunlight falls more directly onto the poles, and the latitudinal temperature gradient is more shallow. The poles warm, and the equatorial regions cool slightly. This redistribution of sunlight is important because the martian volatile inventory is distributed about the planet and continuously tries to reconfigure itself so that the volatiles can reside in the lowest temperature reservoirs available.

The martian atmosphere is the best understood of the planet’s volatile reservoirs. It holds about 150 kilograms of CO2 and about 0.01 kilograms of water per square meter, although the amount of water varies through the year.

The martian soil, or regolith, holds volatiles both as adsorbate and as ice. Both water and carbon dioxide are adsorbed onto the surfaces of martian soil materials. We cannot accurately quantify the adsorbate reservoir because we do not know how much surface area is available in the martian regolith. The uncertain parameters include not only the depth of the regolith, as already mentioned, but also the grain size of the powder. Through careful measurements of adsorption on a variety of terrestrial analogues at Mars-like conditions, however, we can estimate that the martian regolith materials hold on the average 2 x 10-7 kilograms of carbon dioxide per square meter of soil surface and about 2 x IV kilograms of water in the same area. We can guesstimate the extent of this reservoir by assuming that surface area of the soil is around 20 square meters per gram-a number that is not unreasonable based on terrestrial experience-and that the depth of the powder is about 20 meters. In that case the amount of adsorbed CO2 is around 260 kilograms per square meter of martian surface, and the amount of water is about 20 kilograms per square meter. We can only estimate the upper limit of water that may be present in the subsurface based on estimates of the thermal gradient in the martian subsurface and the porosity of the regolith. Ground ice represents an enormous potential reservoir, with up to 500,000 kilograms per square meter of ice present across the planet, mostly in the high-latitude subsurface.

The polar caps are the most obvious example of a volatile reservoir that is latitudinally restricted. Again, it is not possible to assess accurately the total capacity of the polar caps, but estimates are on the order of 5,500 kilograms per square meter. Thus either the polar caps or the permafrost may be the largest reservoir of water on Mars.

As the planet tilts back and forth as a result of obliquity oscillations, the sunlight plays across the surface, vaporizing volatiles where it warms and causing them to condense in the colder reservoirs. This exchange of mass between the volatile reservoirs may constitute the quasiperiodic climate change that is recorded in the polar layered terrains.

As the obliquity increases, the poles and high-latitude regolith warm, and the equator cools somewhat. The highlatitude regolith releases some of its adsorbed gases, which accumulate in the atmosphere, perhaps causing the atmospheric pressure to rise. At lower obliquity, the polar regions cool-enough for quasipermanent CO2 caps to form. Much of the atmosphere condenses out at the poles, and the pressure drops. As long as there are polar caps, the equilibrium between ice and vapor, which is a sensitive function of temperature, controls the atmospheric pressure. At higher obliquities, when the poles are too warm for year-round CO2 caps, equilibrium between the atmosphere and adsorption controls the pressure. The upper pressure limit at high obliquity is dependent on the total amount of carbon dioxide, which in turn cannot be known without knowing the total amount of regolith. Water ice in the regolith may also be redistributed, condensing at lower latitudes during higher obliquities and retreating poleward as the obliquity decreases.

This simple climatological cycle is rife with feedback mechanisms. For example, higher atmospheric pressures may allow dust to be raised into the atmosphere and deposited around the planet more easily Dust deposition can darken the polar caps, raising their temperature and causing more volatiles to sublimate into the atmosphere. The removal of ice from a mixture of dust and ice may lead to a mantling of the polar caps with dark dust, eventually choking off the sublimation of additional ice. Indeed, such a mechanism has been posited as an explanation for the origin of the polar layered terrain. Without a better understanding of the composition and structure of the layered terrain, however, it is impossible to confirm any hypothesis for their origin.

Implications for Prebiotic Chemistry Every model for the earliest climatic history of Mars invokes liquid water either at the surface or in a shallow groundwater table. As Chris McKay and Carol Stoker at NASA Ames have pointed out, the only unambiguous requirement for the abiotic origin of life that can clearly be identified from the study of the earliest life on Earth and the prebiotic chemistry that brought it about is that liquid water must be present. Mars readily satisfies our limited criterion as a possible site for the origin of life. We do not believe that timing was a problem, in spite of the limited duration of the period in martian history that was, if not warm and wet, at least damp and chilly. Life had already appeared on Earth by 3.8 billions years ago, when Mars likewise had liquid water.

What about the persistence of life on Mars? There is absolutely no definitive evidence that there is, or ever has been, life on Mars. Nonetheless, the possibility that some organisms survive in some sheltered oasis cannot be discounted. In the past 20 years, a continuing series of discoveries have demonstrated that life on Earth has adapted to an astounding variety of environments, some of which have parallels on Mars.

Terrestrial organisms have been found living in cracks in deep rocks under the Columbia River valley, inside rocks in both hot and cold deserts, in Antarctic lakes under permanent ice covers many meters in thickness, and at mid-ocean ridges, miles below the depth to which sunlight penetrates. It is not too fanciful to entertain the notion that once life gets a grip on a planet, it somehow manages to hang on, exploiting every possible nook and cranny, no matter how marginal.

The only conceivable niches that might provide a habitat for martian biota are in the subsurface, which would provide shielding from the ravages of solar ultraviolet radiation and perhaps maintain enough humidity to sustain a specialized ecosystem. Although even subsurface liquid would be unstable if the molecules were free to migrate to the atmosphere, localized subsurface water could persist at depth if physically trapped there by impermeable layers. Further, liquid water may be associated with subsurface hydrothermal systems in the vicinity of volcanic activity The presence in the martian atmosphere of neon and helium, both of which rapidly escape to space and therefore must be resupplied, as well as the presence of very young volcanoes, demonstrate that Mars is still degassing and very well may not be volcanically, or biologically, dead.

The Future of Mars Exploration

The martian climate has evolved continuously throughout the history of the planet and is continuing to evolve in ways we do not fully understand. A number of key observations have been made, however, and there is every expectation that many of the uncertainties described in this article will soon be resolved.

A suite of spacecraft from the United States, Russia and perhaps the European Space Agency will contribute mountains of data toward resolving many of the controversies about martian climate evolution. The structure and composition of the polar layered terrain will be studied by cameras and infrared spectrometers to be flown by both the U.S. and Russia. We will look closely with cameras that can resolve details of the valley networks as small as one meter, providing clues to their origins. Infrared spectrometers will search for the massive carbonate deposits that must necessarily have been the final fate of the putative massive CO2 atmosphere, perhaps responsible for an early greenhouse atmosphere. The many Hesperian-age features that may or may not be glacial in origin will be studied closely for clues to their genesis. And we hope that by early in the next millennium, one or more samples of the martian surface from a well-characterized site will be returned to Earth to allow us to calibrate the crater frequency curves, date the martian surface and resolve the enormous uncertainties so evident in Figure 2. It is again time for the textbook writers to be twitchy.