Mention anything about the potential for life on Mars, and you’ll get people’s attention. The conversation often centres on water. Life as we know it requires liquid water, and the evidence that Mars has had liquid water in various forms throughout its history means that it could have supported life.
But Mars is a frozen, desert planet today, with a thin atmosphere and temperatures typically well below the melting point of water ice. So why do we think Mars had abundant liquid water in the past? And what caused the climate to change to one unfriendly to life as we know it?
The evidence that Mars once had liquid water came initially from orbiter images. Thanks to them, we’ve known for decades that Mars has systems of valleys that look like they formed via water runoff. We’ve seen what appear to be deposits of sediments carried by water as it flowed into the enclosed basins of impact craters. We’ve also observed that the oldest surfaces look worn down, and that few impact craters smaller than about 10–15 km wide mar the landscape, which combined tell us that significant erosion has occurred — with erosion by water runoff being the favoured explanation.
Starting with the Mars Pathfinder landing in 1997, we also began observing small-scale surface features that suggested water once flowed. We’ve identified sediment deposits that bear all the hallmarks of having been laid down in liquid water. And we’ve seen small, round deposits formed by the buildup of waterborne material, called concretions, and even specific minerals that require liquid water in order to form.
By estimating dates for these various features, planetary scientists have put together a rough timeline for water on Mars. It reveals that the strongest signs of liquid water tend to be very old. In fact, all of the geological and geochemical evidence points toward Mars having had a climate that allowed liquid water to be widespread on the surface up until about 3.5 to 3.7 billion years ago. Conditions then changed rapidly (geologically speaking), and whatever allowed water to be present disappeared over a period of only a few hundred million years, leaving behind a colder and drier planet.
There is some evidence for water in later epochs — large flood channels that appear to have been formed by the catastrophic release of water from the subsurface, very recent (within the last couple million years) gullies that may involve water — but not as an abundant, stable liquid. These later features would all involve subsurface sources and do not require a different climate than what we see today.
What would the climate on early Mars have looked like? We expect that the average temperature likely was near 0°C or higher in order to support extensive liquid water. This compares to today’s average temperature of around -60°C. The simplest explanation for the higher earlier temperatures is the presence of an abundant greenhouse gas that would trap the heat from the Sun and raise the global temperature. Carbon dioxide is the best greenhouse gas candidate, and it’s the primary component of Mars’s current, thin atmosphere. Scientists have not worked out the details of this ancient warming — whether CO2 alone could raise the temperatures or whether another greenhouse gas such as hydrogen or methane would have been required — but most expect that a thicker atmosphere early in Mars’s history will turn out to be the right explanation.
This is the picture we had prior to the Mars Atmosphere and Volatile Evolution (MAVEN) mission, which launched in 2013: that there must have once been a thicker atmosphere and that CO2 played a role. How did that atmosphere disappear? That is the mystery we sent MAVEN to solve.
My fellow planetary scientists and I had already spent many years thinking about where the CO2 could have gone. Observations had revealed CO2 -derived minerals (such as carbonates) in small amounts on the surface. In theory, these minerals could contain CO2 that had come from the atmosphere. However, scientists had spent the previous two decades looking for a carbonate reservoir on the surface or in the crust that was large enough to hold all the gas from a thick atmosphere, and they hadn’t found it.
If atmospheric CO2 had not gone into the surface or subsurface, then it might have escaped to space. We knew that at least some atmospheric gas had been lost to space: Measurements from the European Mars Express spacecraft, launched in 2003, showed ions from the planet’s ionosphere being carried away by the solar wind. Although these measurements ultimately gave us a pretty good estimate of how much gas is being lost today, we didn’t have enough information on the physical processes involved to estimate with any certainty how much atmosphere Mars had lost over time. We knew that the Sun’s ultraviolet light and the solar wind, which together drive escape, both had been more intense in the past, but without details on the role that each of the potential mechanisms contributed to gas’s removal, we couldn’t extrapolate the loss rates back through time.
Also, we had chemical measurements from the Viking mission, from Earth-based telescopes, and from Martian meteorites collected on Earth that all strongly suggested that atmospheric loss to space had occurred. This evidence comes from isotopes, the different forms of an element that are distinguished by the number of neutrons in their atomic nuclei. The ways that gas is removed from the atmosphere into space preferentially take the lighter isotope of a pair — normal hydrogen is removed more efficiently than the heavier deuterium, nitrogen-14 over nitrogen-15, and argon-36 over argon-38 — leaving the gas that is left behind enriched in the heavier isotope. Measurements of each of these isotopic pairs in the atmosphere showed just such an enrichment. This meant that loss to space probably was an important process for changing the atmosphere. However, without detailed knowledge of the composition and structure of the upper atmosphere, going from that enrichment to an estimate of the total amount of gas lost involves a lot of guesswork.
And that’s where MAVEN came in. We wanted to make measurements that would tell us how much gas is escaping to space today, what the specific processes are that are removing it, and how we could extrapolate back in time to get the total amount lost throughout Martian history. Then we could determine whether atmospheric loss to space had been the dominant process in changing the Martian climate, only one important process among several, or instead a relatively unimportant process.
Making these measurements was no easy task. The many ways that the upper atmosphere and the solar wind could interact, not to mention the ways these ways could interact, make the work similar to unraveling the proverbial Gordian knot. Mars is a complex planet. Goings-on in its deep interior, surface, atmosphere, upper atmosphere, and solar wind all link together, sometimes strongly, and we have to understand the interactions between each of the components in order to understand the system as a whole.
First, we needed to measure the solar properties that drive escape of gas from the Martian atmosphere and the specific ways the upper atmosphere’s composition and structure respond. With its nine science instruments, MAVEN measures the amount of solar ultraviolet light hitting the planet; the solar wind speed, density, and magnetic field; and the solar energetic particles that are emitted from the Sun by solar storms. On the receiving end of the physical system, MAVEN also measures the basic state of the upper atmosphere’s temperature, neutral-gas composition, and ion composition, as well as the electron properties in the ionosphere. Recently, we’ve been able to add measurements of the neutral and ionic winds in the upper atmosphere, too.
We’re also determining how much gas Mars is losing today, following the clues in the upper atmosphere most likely to be important:
The ions in the atmosphere that are being picked up and stripped away by the solar wind;
The ions that are swept up and then flung back into the atmosphere, knocking other atoms into space in a process called sputtering;
The properties of the ionosphere that tell us how much gas is being removed by photochemical processes; and
The hydrogen distribution in an extended ‘corona’ surrounding Mars that tells us how much hydrogen is escaping to space. Together, these measurements enable us to follow the chain of evidence to determine the importance of each of the likely loss processes.
MAVEN entered orbit in September 2014. With it, we’ve collected measurements for longer than a full Martian year, and we’ve seen Mars at all seasons. During this time, we’ve seen several tens of solar storms hit Mars, including a couple of big ones, and we’ve also seen the intensity of the solar ultraviolet light change significantly as the Sun has gone through part of its 11-year cycle. We’ve made observations at essentially all solar zenith angles (the angle between the Sun and the spacecraft as measured from the centre of Mars), at all local solar times and at most latitudes. We’ve observed at a wide range of locations on the planet, including over the regions of strong crustal magnetic fields and over regions with no magnetic field, and including all geological provinces.
The major atoms that we’ve observed being lost to space today are hydrogen and oxygen. (Other elements are being lost as well, but they’re harder to observe than hydrogen and oxygen are.) These come from H2Oand CO2 , broken apart by the Sun’s ultraviolet light. Hydrogen is leaving by thermal escape, which means that the gas is hot enough that some of the hydrogen atoms naturally move fast enough to escape Martian gravity. The hydrogen isn’t being lost at a constant rate, however — we see a factor-often variation in the escape rate throughout the Martian year, with the greatest loss rates occurring during the seasons when the atmosphere is the dustiest. We think that the dust increases the atmospheric temperature, allowing the water molecules that are the source of hydrogen to rise to higher altitudes, where they can be broken apart more easily and where escape is possible. Recent work with data from the Mars Reconnaissance Orbiter supports this picture.
The major way Mars loses oxygen today is by what’s called photochemical loss. In this process, an ionised oxygen molecule (O2 ) collides with an electron and recombines to form a neutral molecule. This recombination releases enough energy to break the molecule into two separate oxygen atoms and give them enough energy that, if one is moving upwards, it will escape into space (if it doesn’t hit anything else first).
Adding these losses up, we’ve discovered that Mars is losing gas to space at a rate of about 2 to 3 kg/second. That’s 10 million billionths (10 ) of the total atmosphere lost each second. That may not sound like much, but it’s enough of a trickle that, over the more than 4 billion years of Martian history, it could have removed enough oxygen and hydrogen to create a global layer of water a few metres thick. Equivalently, the oxygen that has been lost would have made enough CO2 to produce an atmosphere ten times thicker than the present one of 6 millibars atmospheric pressure.
We don’t think that the loss rate has been constant in time, however. The Sun’s ultraviolet radiation and its wind of energetic particles were both more intense early in the Solar System’s history and would have driven greater gas loss. Now that we know the specific processes involved, though, we can calculate how much loss would have occurred earlier in Martian history if these same processes were at work.
Using these extrapolations, it appears that more than a half bar of CO2 could have been lost to space. This is roughly 100 times the amount of CO2 that is in today’s atmosphere, and it is enough to have produced significant greenhouse warming. Given that the young Sun might have unleashed solar storms more often than it does today, even more CO2 could have been stripped away.
This result might sound a bit hand-wavy, but we can determine the total loss more directly, too, thanks to argon. Argon’s heavier isotope naturally settles lower in the Martian air than the lighter one, creating a predictable ratio of argon-36 to argon-38 in the upper atmosphere. This ratio will be different at the surface, where things are well mixed. The argon ratio at the ground has been measured most accurately by the SAM instrument on the Mars Curiosity rover. Combining Curiosity and MAVEN measurements with the argon ratio seen on other Solar System bodies tells us that Mars has lost to space about two-thirds of the argon that had ever been in its atmosphere.
Argon also reveals how it was lost. Because argon is a noble gas, it can’t be removed by chemical processes such as interacting with the surface and forming mineral deposits there, as carbon dioxide forms carbonates. Only the physical sputtering mechanism will work. If Mars lost a lot of argon via sputtering, it probably lost similar amounts of carbon dioxide that way, too. Other processes, such as photochemical loss, would have removed even more of the CO2 , giving us confidence that our extrapolations from current oxygen and hydrogen loss aren’t totally off the mark.
Following the trail of evidence from MAVEN and what we know about the Sun, the majority of this loss would have occurred early in the planet’s history, when the Sun’s ultraviolet light and the solar wind were most intense. In the very earliest epochs, Mars’s global magnetic field likely protected the atmosphere from stripping — the field would have kept the solar wind standing off at a greater distance, as happens on Earth today. Based on magnetic records in the planet’s surface, we think the global magnetic field turned off roughly 4 billion years ago. That switch would have allowed the solar wind to hit the upper atmosphere directly, triggering significant atmospheric loss to space. The bulk of Mars’ atmosphere then would have been lost within a few hundred million years. As the solar radiation and wind calmed down later in history, the loss would have slowed to a gradual leak, but the damage would have been done. The loss we see occurring today would be the tail end of that slow leak.
This timing of the Sun’s stripping of the atmosphere will sound familiar: It matches the timing of the climate change inferred from the geologic features left by water. That’s not by our design; it’s the result of totally different lines of evidence converging on the same picture. Scientists suspected that the bulk of the Martian atmosphere has been lost to space, but MAVEN has confirmed it. This means that escape to space was the major process responsible for changing the Martian climate from its early, potentially warm and wet environment to the cold, dry planet that we see today.
Where do we go from here? Our results are based on one Mars year so far. But not every year is the same on Mars. The lower atmosphere’s water cycle and the behaviour of dust there can vary significantly from one year to another and can affect the supply of gas to the upper atmosphere. MAVEN hasn’t observed the effects of a global dust storm yet. The planet’s dust storm season begins this summer, though, so perhaps we’ll get lucky.
In addition, the Sun goes through an 11-year cycle of behaviour, which changes the properties of the solar wind and our star’s ultraviolet output. MAVEN arrived just after solar maximum in the current cycle and was able to see how a moderately active Sun affects Mars’s atmosphere, but it was a very weak solar maximum, with only a few strong solar storms hitting the planet. We aren’t entirely sure yet what the ‘most common’ behaviour of the modern atmosphere is. This means that the extrapolation of present-day loss rates into the past carries some uncertainty.
The spacecraft and all of its science instruments continue to operate nominally (a wonderful word to hear for anybody involved in the spacecraft world!). We plan to continue observations in coordination with the European Mars Express mission, which has been in orbit for more than a decade. And we’ll coordinate with future observations from other spacecraft: The European/Russian Trace Gas Orbiter has finally settled into its circular, near-polar orbit at Mars and is beginning science observations, and the United Arab Emirates Hope orbiter will launch in 2020 to study how the lower and upper atmosphere connect.
MAVEN’s current extended mission runs through September 2018, but it has enough fuel that we think it can survive until 2030. We plan to continue our science observations as long as possible, along with serving as a communications relay between Earth and rovers and landers on the Martian surface. It’s our hope that, in addition to what we’ve already learned from MAVEN, teaming it up with current and future spacecraft will teach us even more about Mars’s atmosphere and the history of habitability on this small, frozen world.
• Planetary scientist and MAVEN Principal Investigator BRUCE JAKOSKY is Associate Director of the Laboratory for Atmospheric and Space Physics at the University of Colorado, Boulder. He has been exploring Mars since the Viking missions in the 1970s. HALOS MAVEN has detected coronae of atomic oxygen and carbon around Mars that are en route to escaping. More dramatically, an extended cloud of hydrogen atoms (blue image) extends at least 10 Mars radii beyond the surface. The red circle marks Mars’s location. SEASONAL CHANGES Mars loses the most atomic hydrogen to space during the northern hemisphere’s late autumn and winter.
MARS TERRAIN: ESA / DLR / FU BERLIN, CC BY-SA 3.0 IGO; ILLUSTRATIONS: NASA GSFC. LAYERED ROCKS: NASA / JPL / CORNELL; CONCENTRIC RINGS: NASA / JPL / UNIV. OF ARIZONA; HALOS: MAVENTEAM / UNIVERSITY OF COLORADO / NASA; SEASONS GRAPH: GREGG DINDERMAN / S&T, SOURCE: AUTHOR. IONS: NASA’S SCIENTIFIC VISUALIZATION STUDIO / MAVEN SCIENCE TEAM. ADUST STORM: NASA / J. BELL (CORNELL UNIV.) / M. WOLFF (SSI) / HUBBLE HERITAGE TEAM (STSCI / AURA); AURORAS: NASA GSFC / UNIVERSITY OF COLORADO; TIMELINE: TERRI DUBÉ / S&T ■