PART SIX
(3 billion years ago to the present day)
A WORLD OF ICE
Of all the jobs you’ve been given, this has to be one of the strangest. You’re back in your buggy, but you’re not rolling across a flat, barren, arid landscape of dust and rock. Instead, you’re looking up, up and even further up at a reclining wall of pale pink, lit by the low polar sun – a wall that you know isn’t made of layered basalts or wind-blown ash, but of solid ice.
You can see the cold. The ice gradually – one molecule at a time – turns from dense solid to tenuous vapour. It collects as streamers of white smoke and tumbles in eddies down onto the ground, where it dissipates like a summer mist. You shiver, despite the fact that your spacesuit is at the perfect temperature, neither warm nor cool, and the fans circulating the air inside your helmet are simply ticking over. The nominal temperature outside is minus seventy, and yet you feel fine.
Up on the top of the ice cap, a full three kilometres above you, is a drilling rig that’s boring out ice cores, ten metres at a time, as far down as the drill string will support it. Those cores are going to be stored in a freezer – a freezer on Mars, really? – and then analysed one by one by the resident glaciologist. She’ll be looking at the fine annual layering of dust, melting small samples in order to retrieve the atmospheric gases and looking for any discontinuities in the record. She wants to fill in the timeline of climatic conditions over the last ten or twenty thousand years, to study it for variations measured in decades, centuries, millennia.
Your task couldn’t be simpler, though. Find the oldest ice you can, use your portable rock drill to take some small cores and store them in the insulated box you’ve been given. The oldest ice is, of course, at the base of the ice cap. Usually, that would be the most difficult and inaccessible part, but here on Mars you can drive up canyons carved by the freezing polar winds as they pinwheel off the pole, and face off against bright walls of ancient ice deep in the heart of the polar cap. Even if the drilling rig can’t get down this far, here’s an easy way for the glaciologist to get the samples she needs.
It’s a strange landscape. The floor of the canyon is covered with thick dust, released by the tonne from the ice cap, that has fallen to coat the ground and then gathered up again in shapes that range from ripples to mighty marching dunes large enough to swallow your buggy whole. The permafrost lies below this mobile, inconstant surface. Your tyres plough through the fine red carpet, stirring it up: it’s heavy going, and there’s kilometre after kilometre of it.
The walls of the canyon start as nothing: mere lobes of dust-covered ice, indistinguishable from dunes. But they rise up, over the foreshortened horizon and to either side, and eventually you’re in a flat-bottomed valley, flanked by oddly carved hills made of banded ice.
You’ve gone as far as you can. The canyon ends in a blunt curve that soars above you, a high, sloping buttress. You look up and you can see patterns in it. Superimposed over the yearly signals of dark summer melt and bright winter accumulation are thousand-year wavelengths, where those same annual marks are crammed together to produce smudged stripes in the exposed ice face. They alternate with wide bands, where the water vapour froze in abundance for a cycle, and spread out the years in a broad reach.
There’s nothing to tell you just how old the ice is at the very base of this cliff; all you know is that this is where you have to drill. Perhaps there are markers in there – records of celestial events like cosmic ray bursts, solar flares, meteorite strikes – but that’s for the glaciologist to find. You park up, set out the drill and carry both it and the sample box up to the cliff.
You have to turn your suit lights on. They shine harshly in the shadowed canyon, making the ice refract and glitter. Even though it’s opaque, you can’t help but see shapes in it. They’re not real, and for a moment you wonder if your oxygen mix is wrong, but you check and it’s fine – besides, alarms would have sounded long before you noticed anything. You put your gauntleted hand against the ice – briefly – and realise that of all the things you’ve touched on Mars, this is probably the most relatable. While your ancestors were busy knapping flints and hunting aurochs across the chilly plains of Europe, chasing the edge of the retreating ice front, there was another world where the ice was advancing, accumulating on bare Martian soil winter after winter, as another random, wayward change in obliquity altered the climate for another protracted season of cold.
INTO THE AMAZONIAN
As we’ve progressed from the Noachian to the Hesperian, we’ve been able to plant markers in the timeline for major events, even if some of our placements were tentative at best and wildly wrong at worst. But for the climate-critical moments, we were faced not just with a choice between vaguely reasonable guesses, but with actively contradictory evidence that confounds any attempt at a coherent storyline.
The transition between the Hesperian and the Amazonian is even more in dispute. It happened around 3 billion years ago, give or take 500 million years. The age of the exemplar surface, the Amazonis Planitia, is difficult to calculate accurately by crater counting, not because there are too many craters, but because there are too few.
By the time the Hesperian eased into the Amazonian, impacts had become unusual events. Not unheard of, certainly, but none of the big, planet-altering asteroids were left in Mars-crossing orbits, and in order to create anything approaching accurate crater-counted ages, smaller and smaller craters have to be considered – and in order for that to happen, the photographs we work from have to be very high resolution.
For a Noachian-aged surface, craters less than 16 kilometres in diameter are barely worth counting at all. For an Amazonian-aged surface, it’s unlikely that there would be even one crater that large to count, and in order to get any idea at all of a comparative age, craters down to 100 metres and smaller have to be recorded.
In fact, because of the very low rates of crater-making events, the younger a surface is, the more difficult it becomes to date. A geological unit with a small total area – like a late lava flow or a riverbed – and with no craters at all is obviously relatively fresh, but just how young is it? A hundred million years? Ten million years? A thousand years? Counter-intuitively, it’s easier to date an older surface, simply because it will have accumulated a reasonable number of craters early on in its existence.
Time collapses in on itself. There were few spectacular events in the Amazonian. No new volcanoes. No giant craters. No great oceans. Mars entered a long, cold, sparse epoch filled with ice, dust and wind. It became a deep freeze, a vacuum flask, a desert. We can point to features that exist, but not know when or in which order they formed. We have to treat them as scattered Polaroids thrown down on the ground: we know they represent part of the whole, but not all of it.
When we started our Mars journey, we thought we already knew what the endpoint would be: our Mars. It has visible ice at the poles, significant reservoirs of permafrost under the soil and glaciers at all latitudes. The red dust that coats everything is blown about on weak winds and the rate at which it wears down exposed rock is less than a single millimetre every million years. Meteorites hit fitfully. Landslides millennia in the making tumble down slopes in clouds of dust, settle and gain the appearance of having been there forever. Marsquakes shake the ground, but do not stir it. Everything seems static and sterile compared with what has gone before.
We know, with some degree of certainty, that Hesperian Mars was an active planet, but the processes on Mars today are literally glacial, and so we ask ourselves, when did all this activity cease? Was it an abrupt cut-off or a gradual, terminal decline from warm and wet to cold and dry?
But while we’ve long been convinced that Mars is dead, that nothing has happened in the Amazonian except dust and decay, that view is now undergoing some radical revision due to a fresh – and better – understanding of the climate of Mars. We know that the obliquity of the poles changes chaotically. That the latest large swing from high obliquity to a now-moderate average of twenty-five degrees occurred only four million years ago. That low-latitude glaciers formed at obliquities of thirty-five degrees perhaps within the last two million years. The suspicion is growing that there were – and still might be – pulses that reignited the spark of Mars.
Were there seasonal lakes in the early parts of the Amazonian? Most likely, yes. Were there rivers? Again, there’s now evidence for them where there wasn’t before, thanks to higher-resolution pictures. Were these rivers simultaneous with glaciers and permafrost sculptures? Yes, and one was probably the result of the other melting. There were also eruptions of lava and ash from volcanoes and rifts that we can date not just to three billion, two billion or one billion years ago, but to one hundred million or even ten million years. Just how close to the present do we see these events occurring? Could they be happening now? Have we, in fact, simply caught Mars napping?
Potentially. The physics of an escaping atmosphere and a cooling core demands a winding-down of Martian processes, but there are good reasons to assume that, superimposed over that billions-of-years-long flattening curve, are eras of surprising vigour. We can curse ourselves that we can’t see Mars in that state, but we can still scour the surface for signs of activity and imagine. Onwards.
THE AMAZONIAN CLIMATE
Mars’s Amazonian climate is the result of a complex network of strings. Pull on one and they all move. No one factor ‘controls’ whether ice advances or retreats, whether water flows or is locked into solid form. The obliquity of the poles is the main driver determining the amount of solar heat at the surface, but we can overlay this with the eccentricity of Mars’s orbit, which determines the distance of the planet from the Sun throughout the yearly cycle, and the precession of the equinoxes, the gradual movement of the timing of the summer and winter seasons during each orbit.
While some of these variations are predictable in the short term over a few million years, in the long term they are chaotic. We can’t extrapolate back into the past from the present because we know we would be wrong, and there’s no way of telling how wrong. We can give broad upper and lower bounds to the state of Mars in the past, but we simply don’t know when any supposed temperature regime would have operated, or how long it would have lasted.
Modelling the planet’s behaviour can help here: in the same way we use climate models of our own planet to determine patterns and trends and then extrapolate those into the future, we can also make climate models of Mars and extend those into the past. We know that high-obliquity angles drive the melting of polar ice caps and the growth of low-latitude glaciers. We know that low obliquity causes those glaciers to retreat and the ice caps to grow.
Certainly, the ice doesn’t appear from nowhere – it’s not magicked into existence, nor is it dispelled with a wave of the wand; it has to come from what’s left of the Martian water cycle. But this water cycle bears no similarity to the one we see on Earth. The Amazonian sees water move from relatively warm to relatively cold regions, without ever becoming a liquid; it’s a cycle of solid-state sublimation and deposition, with the vapour transported by the wind. There are caveats: there’s significantly more accessible water-ice in the north than the south, because it appears that much of the highlands are mantled with a thin layer of dry, wind-blown dust that insulates subsurface ice. And because of the Dichotomy, winds tend not to cross from north to south.
If conditions become warmer at the poles during high-obliquity periods, they conversely become colder at lower latitudes. This might seem counter-intuitive, but whereas the poles receive half a year of continuous sunlight at a time, the lower latitudes are plunged into cold, dark night once every sol and the Sun is only overhead briefly during spring and autumn. During high-obliquity periods, the equatorial regions are cold. During low-obliquity periods, the polar regions are cold.
We also know there are places on Mars that are cold sinks: we’ve already noted one in the region between Tharsis and the south pole, and another on the high ground between the Argyre and Hellas craters. There are other places where local conditions overwhelm global ones, too. If we run the Martian climate models to take account for geography, we find more traps for cold air: the western flanks of the Tharsis Montes and Olympus Mons, the eastern portion of Elysium Mons and the eastern basin and rim of Hellas. And these aren’t the only places where glacial features are found on Mars – there are also high-altitude equatorial glaciers on the flanks of the tall volcanoes and lower-latitude glacial activity in areas such as the Phlegra Montes, a finger of a mountain range that reaches north from Elysium and into the Vastitas Borealis.
A survey of ice-related features – signs of left-behind glacial debris that make patterns of lines and lobes and characteristic concentric slumps inside craters that show that ice once filled and then flowed there – tells a story of mid-latitude ubiquity. Between twenty-five and sixty degrees, both north and south, these markers are everywhere.
It seems that ice accumulation away from the poles is a recurring feature of Mars: every time the obliquity shifts to steeper angles, ice doesn’t so much march towards the equator as get blown there, to sit inside relatively young, fresh, steep-sided craters that can still trap a denser, water-vapour-laden atmosphere, and on the leeside valleys of high ground. If we want to see what that might look like, we can gaze down at the Korolev crater, which is in the Vastitas Borealis.
At 80 kilometres across, Korolev has a steep, high rim, a low floor and a 60-kilometre-wide dome of pale water-ice that covers the centre of the basin to a depth of almost 2 kilometres – that’s more than 2,000 cubic kilometres of ice. Because Mars is currently in a low-obliquity period, the sight of Korolev is startling and rare: the northernmost plains have few large craters, and conditions have to be just right for ice to accumulate there, even at seventy degrees north. Only a few million years ago, these pearls of ice, nestling inside their crater shells, would have been commonplace across the mid-latitudes, numbering in their hundreds. Korolev is both a survivor and a sign of another Mars.
The most important feature of the Amazonian is this: even though the astronomical variables have been chaotic throughout the entire period, they have repeated again and again over the last three billion years. The climate has veered from low-latitude glaciation to high-latitude icing and back, several times over, in whichever hundred-million-year window we look through.
This chaotic cycling between rapid change and lengthy stability has had profound effects on the landscape. The ground has swelled and shrunk. Rocks have shattered and slid. Slopes have slumped and crept. It’s impossible to tell how much of history has been smoothed and stretched by the constant churn of the climate bands across the face of Mars.
EQUATORIAL ICE
Ice on Mars is both obvious – in those pale northern and southern polar caps – and hidden. The ice caps represent huge stores of water and also carbon dioxide, which can freeze out on top of the caps. Ice, though, is almost everywhere. Permafrost on Mars – subsurface ice bound inside the fractured subsurface – is found across the whole planet. The deeper layers, 7 or 8 kilometres thick, are found adjacent to the poles, but even the equator is frozen to a depth of 2 or 3 kilometres. It’s true that this deep ice isn’t going to interact with the atmosphere – it’s not going to melt, or sublimate, or in any way affect the formation of snow or frost at the surface – but given the huge areas of permanently frozen soil, we can say that even though most of Mars’s water is bound up deep underground, there’s still plenty that’s accessible and close to the surface. Both sixty degrees north and south, the soil is usually at least half ice by volume. Taking a shovel to Mars is akin to digging into a dirty snowdrift. Ice is part of the ground, exposed in the northern plains and hiding under a thin layer of dirt in the southern highlands. We should, therefore, be able to see all kinds of landforms that are caused by, and associated with, ice.
The iciest of features are the glaciers themselves. When we think of glaciers, we probably bring to mind a pale tongue of ice draped between the jaws of sharp grey mountains, lolling out towards a valley below or down to an iceberg-flecked sea. Rivers run out from the front of the glacier as it melts, and we imagine outwash plains, ridged eskers, deep blue lakes and moraines piled high with hard, fist-sized rocks.
This was once true of Mars, back in the more-temperate Hesperian. We call these glaciers wet-based, to describe the interaction between the ice and the ground, which is lubricated with running water. A wet-based glacier can flow relatively quickly, moving as a single mass, gouging at the contact points with the rock at its base and sides as it moves. Debris is entrained within the ice from below, and this shows itself both in the output of sediment at the front of the glacier, and in the landforms it leaves behind when it retreats.
Not so the cold-based glacier. There’s only the thinnest film of water between ice and rock here, created by the weight of the ice above. Mostly, the glacier cleaves to the ground below, deforming and flowing over it like slow wax on its way downhill. Consequently, there’s no grinding at the base of the glacier, and there’s no water discharge from the front. Debris collects on the glacier’s upper surface from rocks falling from the valley or crater sides, and it gets conveyed along, accumulating thickness until the ice itself is completely hidden by a broken cloak of shifting rubble. Bare ice, in a low-obliquity period like now, will eventually turn to vapour, but the cover on rock glaciers provides a substantial (albeit imperfect) barrier to the loss of ice.
These stealthy rock glaciers are difficult to spot because they appear to be simply extensions of the landscape around them: same colour, same roughness. But there are subtle tells that give them away. The first is their shape. With good radar data that pinpoints the height of objects to within a few centimetres, a rubble-filled valley floor that shows a characteristic convex bulge from wall to wall is likely to be filled with a rock glacier. In the same way, if we measure along the length of the valley, the profile will be that of a shallow slope downhill, terminated by a blunt nose. This is exactly what we’d expect of a glacier, if it wasn’t covered by debris.
The patterns on the surface can also show what lies beneath. Because the front of the glacier tends to form a rounded lobe as it trickles like treacle downslope, the debris layer wrinkles front to back. These rounded ridges travel with the glacier, before tipping over at the nose and being subsumed. In the same way, the sides of the glacier pull out long lines of debris in the direction of flow. These curved and straight ridges, perched on the top of the glacier, are often the only visual clues to the ice below.
As a consequence of this distinctive protective mantling, we can find equatorial rock glaciers on the north-western flanks of the Tharsis Montes and Olympus Mons, and smaller ones across broad swathes of Mars, between thirty degrees north and thirty degrees south.
Ice-filled craters less than 20 kilometres across are abundant in these regions too, all hiding their ice beneath a rock-rubble casing. However, this ice has nowhere to go: there’s no downhill within a high-rimmed crater, and so all it can do is slump down inside. The walls of the crater break with age, supplying the ponded glacier with more debris, and the results are series of concentric ridges inside the crater, on top of the ice deposit. Given the utilitarian name of concentric crater fill, this was recognised as a landform well before the mechanism that caused it was understood.
While smaller craters are more able to keep their ice, as they can more completely cover the ice core with a barrier of debris, larger craters – like Korolev – will lose their ice faster. They’re likely to be shallower, with older, more eroded rims that leak both air currents and sunlight into their depths, and being larger, any debris cover they might acquire will be limited to the margins of the glacier. With the centre exposed, the ice will sublimate away to almost nothing over several million years, until only a sliver is left on the north-facing internal wall.
But what of the vast mid-latitude ice plains of the north, and the mantled highlands of the south? One of the most obvious and ubiquitous signs of ice-rich soils is the peculiar shape of pedestal craters. We’ve met these once before in the Medusae Fossae Formation; they’re craters caused by meteorite impacts in the usual way, but the whole structure appears to be raised up above the surrounding land as a rough circular plateau encompassing an area larger than the ejecta blanket, with the crater itself at the centre. The entire plateau stands tens of metres above the local level and is delineated by a steep scarp. But pedestal craters can also be explained by ice.
Icy ground forms craters like any other solid. The ejecta – rock melt, water vapour and broken, liquid-water-rich ground – is blasted out of the crater and deposited around it; the thickness and breadth of the ejecta blanket is proportionate to the size of the crater. The crater settles and then time gets to work.
If the ice is unstable – that is, if the conditions are such that water-ice turns into vapour – the soil around the crater loses volume. The ground literally deflates, sinking lower than its original level. We might assume that the crater would also simply disappear as the ice vanishes. Starting from the rim and moving downward, it would be eaten away until all that would be left is a shallow depression to mark where it once was, and even that might vanish.
However, while the ground around the crater falls away, the crater itself and its immediate surroundings seem immune to the loss of ice. This is partly for the same reason that rock glaciers persist: the ice is covered in debris. But that cannot be the whole reason, because the protected zone seems to extend beyond the reach of the ejecta, and the crater itself is also well preserved. How pedestal craters appear to armour themselves against erosion is another Martian mystery; it’s likely due to the shock of impact welding soft sediment together, but how that mechanism works in ice-saturated ground is neither clear nor understood.
What we do believe is that the height of the pedestal – the plinth on which the crater is embedded – has nothing to do with the size of the impact, but is solely a function of the erosion of the ice-rich soil layer. Some scarps are 250 metres high, but these are unusual; the much more common height of a pedestal is 10–20 metres.
Naturally, pedestal craters are not immune to the ravages of the ages. As the ice is lost from the margins, the scarp slope loses its integrity and becomes easy pickings for the wind, which nibbles away at it until the crater collapses and become unrecognisable. But so many pedestal craters are persistent from forty degrees north, or south, to the pole, they are a firm Amazonian feature – one that represents both the past and the future.
HIGH LATITUDE ICE
Closer to the poles, where the permafrost is thicker and ice component greater, we see other extraordinary landforms that we can categorise as periglacial – those that are created close to an ice sheet or glacier. The processes that make them are completely dominated by the presence and behaviour of ice, where it’s laid down and where it’s taken up.
Polygonal cracks are a maze of interlocking lines that look like the remains of a dried-out puddle, but on a much larger scale, and they’re formed by intense cold, not drought. The top of a permafrost layer grows and shrinks with the seasons. In winter, the icy ground contracts, splitting at the surface and forming cracks into which pure water-ice frost can fall; in summer, the warmed ground expands, pushing against the icy wedges and compacting itself a little more. Come winter, the cracks reopen and the ice wedges grow again. This forms the tesserae landscape common in the high northern plains.
Alases are circular depressions formed by the collapse of the surface permafrost due to ice sublimation. Material often slumps down from the margins to the middle, creating a beach-like effect at the edges. If these are low points in the landscape, then pingos are local highs: ice-cored conical hills that swell above ground level on permafrost plains. As the ice inside expands, it heaves the soil upwards to produce what look like little, steep-sided volcanoes with cracked, baked-muffin tops. Pingos grow only slowly, but a fully developed one can be hundreds of metres across and tens of metres high after centuries. They can and do deflate, leaving a circular ridge of material surrounding a sunken basin – which looks very much like an impact crater, but without the ejecta.
Intriguingly, we know from our own permafrost regions on Earth that these processes are all enhanced by the presence of liquid water. Polygonal cracks fed by meltwater form their characteristic landscape faster and more completely. Alases often hold seasonal meltwater, and pingos might use that water as their core and form at the bottom of those same dried-up lakes. This brings in the possibility that the Amazonian, like the Hesperian before it, might not have been as cold and dry as we earlier believed. We know that the climate veers wildly, and water-ice is transferred from the poles to low latitudes and back again. Could water have been stable at the surface during Amazonian times, and could it be again?
It’s true enough that there’s no current evidence of liquid water on Mars. There are no lakes, no rivers, no seasonal streams or ponding of groundwater: even the dark streaks that seemed to ooze out of crater walls and valley sides are now believed not to be salty brines, but simply small-scale flows of fresh black sand, released from the frozen surface by warmer summer temperatures. However, it’s also true that the current range of conditions on Mars is temporary.
There were almost certainly times during the Amazonian that liquid water did exist at the surface: there’s evidence that shallow rivers appeared locally in the mid-latitudes, draining into craters and, more importantly, out of them, indicating that they were at some point full to the brim. Little deltas and sinuous valleys complemented the rivers. They’re almost impossible to date, since they cover only a small area of the surface, and it’s also difficult to estimate how long they persisted for, but if we were to guess, they would have been active somewhere between the late Hesperian and the early Amazonian, turning on and off and on again, with each period lasting for several million years. Snow-melt, rather than groundwater, was probably the driver, given the locations of these features. Discrete drainage systems, based on local highs and lows, flowed as long as there was water. Then they just stopped. They ponded, they evaporated or drained away, or they froze and then turned to vapour, potentially forming alases or pingos.
The Amazonian could well have cycled between long periods where the only variation was exactly how cold it was, and brief million-year windows where the temperature – and the pressure – crept up enough to allow free water at the surface, even while the deep permafrost remained intact. But there remains the potential that in the recent past, Mars turned seasonally, if unspectacularly, wet. No wholesale melting of ice, just enough for the conditions to tip over to allow some surface flow for a few days, or a month or two, over the summer.
To show just what a knife-edge these climate calculations are on, we only need to consider the discoveries of the Phoenix lander, which touched down in 2008 on the Vastitas Borealis, at sixty-eight degrees north. Aiming to operate during one northern hemisphere summer, Phoenix was deliberately designed to look for water and conduct water-based experiments. Previous surveys had suggested that there was ice within a metre of the surface of the polygonally cracked ground; in fact, as the rockets that dropped the lander safely down also blew the surface layer of dirt away, they revealed bright ice within a mere 5 centimetres of the top of the soil.
It was light at the landing site nearly all the time, as expected for that high latitude. The Sun rose and dipped during each sol, but never dropped below the horizon. Phoenix relied on solar panels for power, so it was able to function continuously, until the Martian winter came and the Sun slipped below the horizon. Phoenix succumbed to the cold and the dark after a remarkable 157 sols.
During its months of operation, Phoenix reported on everything its cameras could see and its sensors could touch. Temperatures rose to –20°C during the day and fell to –80°C at night. Thin fog formed close to the ground when the Sun was at its lowest, and frost covered the rocks, before sublimating away as the Sun rose higher. Phoenix could see clouds, too, at various heights: high in summer, but as winter approached they formed lower and it snowed – lightly, but still snow. At no point was liquid water ever seen, and the ice the lander exposed, either with its rockets or its robotic arm, was unstable at the surface and began to pit and vanish within days of exposure. But it was close, so close. Given a slight shift in conditions – if Mars’s closest approach to the Sun had coincided with the northern summer, for example – the temperature would have scraped above zero. With a higher obliquity, there would have been a hundred days of that climate.
The conclusion – perhaps a startling one – is that we have seen hints of a completely different Mars that is actually still possible. This means that the ice caps, both northern and southern, aren’t permanent. They are recent constructions, temporary and cyclical, changing not just with yearly variations, or growing and shrinking as the seasons turn, but being partly or wholly obliterated by larger changes. We need to look to the poles for evidence of whether any of this is more than just wishful thinking.
THE POLAR REGIONS
The Amazonian ice caps are different to the ones that were present in the Hesperian, and not just because they’re in subtly different places. The Dorsa Argentea esker deposits we visited earlier are most likely the site of an old ice sheet, but we don’t know which route the poles took, nor the date by which true polar wander was complete. We don’t even know if its rate increased at all, or whether the direction of travel reversed at any point.
Neither do we know whether the Amazonian ice caps were once wet-based, like their earlier Hesperian incarnations. Looking at both the north and south poles now, and the complete absence of contemporary outwash plains and other water-related signs, everything seems to indicate that the current ice caps are cold-based. That might also have been the case in the past, even during periods of high obliquity, but we need to determine how old the current ice caps are if we’re going to make any attempt at answering our climate questions.
Mars’s ice caps – their presence or absence, their size and their persistence – are related to the complex interplay of astronomy, climate and resources. Obviously, there’d be no ice caps without water to make them, and even then, early Mars probably didn’t have permanent ice because the atmosphere was too thick and held its heat too well, despite the planet following the same chaotic obliquity patterns. Ice would only have formed as the atmosphere changed and the pressure dropped: seasonal at first, then all year round. The records for those earlier ice shields have been lost, and we don’t know when that record began to be preserved. Ice, by its very nature, is volatile. It can be carved into the most spectacular shapes, but when it’s gone, it’s gone. It leaves few permanent traces: much of its ephemera can be overwritten by the next winter’s fresh snow.
However, we still need to look. The current Martian polar deposits both have a threefold structure: the elements of this are different at each pole, since the higher altitude of the south pole, and the trapped weather system around it – created by the combination of the Argyre and Hellas craters – gives it a persistently colder quality.
They both have a temporary, seasonal layer of frozen carbon dioxide, which forms only during the coldest part of the winter and vanishes quickly when temperatures rise. It forms as a transparent glaze a few metres thick, but it still represents a significant portion of the atmosphere: a third of the Martian atmosphere freezes out every winter, switching between the north and south poles.
Below that is the first proper layer: in the north, it’s a thick residual ice cap of white water-ice that grows and shrinks from winter to summer, but persists all year round. In the south, it’s mainly carbon dioxide, with some water-ice. Both ice caps are currently growing. The smaller southern one because carbon dioxide freezes with the water-ice and traps it by virtue of its extreme coldness. The northern ice cap because it’s the main repository for sublimated mid-latitude ice. But overall, water is being seasonally transferred from the northern hemisphere to the south pole, slowly but inexorably. At some point in the future, the amount of water-ice there will exceed the carbon dioxide ice, trapping the supercold layer below the surface: some kind of balance will be reached and the transfer will stop.
Beneath the residual ice cap layer is a series of thinly stacked deposits that make up the greater part of both polar ice caps. These are great domes of interleaved water-ice and dust, cut by ravines and chasms that spiral out of the main mass in the direction of the wind – anticlockwise in the north, clockwise in the south.
The northern ice cap is roughly circular, 1,000 kilometres across and 3 kilometres high. A great curved scar cuts kilometres down through it, almost to the bedrock in one part, called the Chasma Boreale, and on the southern side of that the ice cap continues as the Gemina Lingula. It looks for all the world like a great Catherine wheel of ice.
The centre of the southern ice cap is actually offset from the south pole by 150 kilometres and, as a structure, it’s less well defined. It sits uncomfortably half in, half out of a large crater, Prometheus, and its ice deposits are dirtier. It’s also irregular: 3–4 kilometres thick at its maximum, it extends northward in places as a plateau of dust-and-rock-covered ice. Like the northern cap, it has large chasmata – not one, but several, Chasma Australe being the largest – and this gives rise to three distinct lobes protruding from the main cap: Australe Lingula, Ultima Lingula and Promethei Lingula.
These chasmata, both north and south, have nothing to do with ground movement or the passage of water. They are created by wind alone. They do, though, allow us to see the interior of the ice caps without resorting to drilling. Inside, we can see a kilometres-high ladder of narrow, alternating light and dark bands of ice, with the colour of each band depending on the thickness of dust entrained within it. The ice is solid and the chasmata are steep-sided, with broken debris congesting their basements.
There’s another layer at the bottom of the polar piles: the northern one is simply called the north polar basal unit, and the southern one we have encountered before – the Dorsa Argentea Formation. The north polar basal unit is a much more sand-rich deposit, and thickly layered, as if it’s been made from the sublimation of the layered deposits above it. Which it probably has: take away the ice and we’re left with a pile of wind-blown dust where the ice cap would have been. Sand that has blown out of the basal unit forms fields of moving dunes around the polar cap – the adjacent Olympia Planum has the same construction as the basal unit, except it’s almost completely covered by dunes. It’s not a stretch to suggest that reducing the ice cap to almost nothing would allow these dunes to advance onto the pole itself.
The southern ice cap appears to be more stable – and certainly more persistent – than its northern twin, but there are good reasons to believe that over the length and breadth of the Amazonian, both ice caps have died and risen and died again. The northern ice cap may only be a few million years old at most, and potentially only a few thousand years – crater counting really doesn’t work on a surface that renews itself yearly. The southern structure is older, anywhere between a hundred million and ten million years old, but these ages have to be acknowledged as just best guesses.
The number of mid-latitude ice ages within the last few million years can be easily counted in the tens. Throughout the three billion years of the Amazonian, that number rises dramatically, so the ice caps become flickering, ghost-like features, with the ice switching from the low- and mid-latitudes to the poles and back again countless times.
The amount of water-ice held in the polar regions is about a 20-metre global equivalent layer. If the ice caps disappear, the atmospheric pressure doubles. If the pressure increases, so does the temperature. As the Martian climate yaws from low-obliquity polar ice caps to high-obliquity low-latitude glaciers, here’s the tantalising hope: when the ice caps are dispensed with, we might just see free water again on the surface of Mars.
THE DUST CYCLE
Dust is not the same as sand. What we call sand is simply the fraction that is driven before the wind. Sand saltates; it gets picked up and travels in a ballistic arc, striking the ground further on. Dust, on the other hand, becomes part of the wind, and only drops when the wind can no longer hold it up. In practical terms, since Mars’s atmosphere is so thin, this refers to particles no larger than two-millionths of a millimetre across.
Despite this size barrier, dust is ubiquitous. There’s dust in the layered deposits at the north pole, and it covers the ice cap at the south pole. There’s dust on the wide open northern plains. There’s dust mantling the southern highlands. There’s dust on the flanks of volcanoes and in the centres of craters. It forms ripples and dunes. It’s carried from one part of the globe to the other, and eventually everywhere.
The Martian weather system, like all weather systems, is chaotic, but with a wide enough lens, trends appear and we can make predictions. It turns out that Mars has a dust season. It starts suddenly, with the southern hemisphere spring. The surface layer of winter ice turns to vapour, and the dust is left sitting there, waiting, poised for the first gusty winds to pick it up and carry it aloft. The lack of airborne dust early in spring allows the Sun to heat the ground directly – dust devils, spirals of warmer air that rise from the rock, are a constant factor in loading dust into the atmosphere. But as the Sun climbs higher and spring turns to summer, the winds that turn around the southern mid-latitudes get stronger, and they pick up dust from wherever it can be found.
Then there are dust storms. Most years there are only minor storms, but every so often there’s one that shrouds the entire planet. They start in places like the Thaumasia Planum on southern Tharsis or around the Hellas crater, with a great plume of dust rising up into the sky. Mostly they fizzle out within a few days – Mars is quick to anger and equally quick to quieten again – but at other times, in other years, the storm gains in intensity. The atmosphere turns opaquely ochre and sunlight heats the cloud, adding more and more energy to the storm. The dust cloud spreads out across the southern hemisphere. Because of Mars’s low gravity, the Martian atmosphere is tall and the dust can be carried a long way up, but Mars’s atmosphere is also thin, so as soon as it stops churning, the dust falls out. Constant movement is a necessity.
Seeing a dust cloud rise up around Hellas and expanding so that it wraps around the entire southern hemisphere isn’t an uncommon experience. The big storms, though – the ones where the dust crosses the equator and encroaches on the north – happen only once every decade or longer. The surface vanishes from view – even Olympus Mons, the Tharsis Montes and Elysium. Mars becomes a smooth, rust-red ball with only the pale poles visible. It can stay this way for weeks, and then it clears.
If we know something about how the storms start, grow and persist, we know very little about how they end. They should, reasonably, last longer, driven by the energy of the Sun striking the dust-laden air and heating it to new heights of opacity. But they don’t. The winds don’t even seem to lessen. The dust just seems to get tired of being up and comes down instead.
To be in a Martian dust storm is a confusing thing. Even though the sky grows dark and the Sun turns into a shadow, the fineness of the dust makes it look more like fog than objects in motion. The wind-speed isn’t so troubling – barely topping out at 100 kilometres an hour, storm force but nothing greater – and the thinness of the air is such that, if we were standing in the middle of the storm, we would experience no buffeting or risk losing our balance. We would see bands of dust advancing on us like a mountain of cloud, cutting our vision to a few metres, and we would hear a low susurrus of sound, a thousand whispers just on the other side of our spacesuit helmets.
Despite the size, longevity and thickness of a big Martian dust storm, there’s very little lightning. Detailed and concentrated study has shown that it’s simply not a regular feature on Mars. This disappointing and somewhat counter-intuitive finding can be explained by understanding what good conditions for lightning look like.
Lightning needs, firstly, a very large difference in electrical charge between the ground and the cloud, and secondly, a gap between the two that doesn’t habitually conduct electricity. We can see how Mars lacks both of those requirements: the cloud is already at ground level, and every level above it, and the dust itself behaves like an electrically conducting fluid – each tiny particle of dust carries a minute part of the electric charge, and all of them are in motion, giving no opportunity for any great difference in charge to build up.
This last property gives the dust its most dangerous feature, though. Because it’s so easily charged, and so small, it sticks to everything, coats everything, gets into everything and is incredibly difficult to dislodge. The static cling of Martian dust means it can find its way into gears and seals, wearing them down and causing them to fail. It reduces the efficiency of solar panels. It contaminates samples and, because it’s so very fine, if humans ever do reach Mars, it’ll get into our artificial habitats and everywhere else, inevitably ending up as a feature of every Martian visitor’s diet.
Even though dust is created slowly, the length of time that it has been accumulating – and being recycled – allows for deep deposits of wind-dropped dust. More than that, the dust can form rock: the electrostatic charge that individual particles carry can cause them to clump together so that they’re no longer easily picked up. These duststones then form a stable surface and are eroded by the wind, not as loose cover but as soft rock, forming their own distinctive, wind-carved landforms. Duststone is the predominant rock type of the Amazonian, and the dust itself is a product of conditions that have persisted for three billion years. That’s three billion years of material being stripped off the vast deposits of volcanic ash in the Medusae Fossae Formation, of sand grains grinding against each other and creating their strange mechanical–chemical interactions, of desiccated clays being lifted from crater floors and the bed of the old northern ocean, of a steady, thin patter of meteoric dust from space.
Just how much dust is there? The thickness in any particular region obviously depends on whether it’s a net producer or accumulator of dust. For example, Arabia Terra, the only highland region that projects north of the equator, finds itself mantled in at least 20 metres of dust, and in places it can be as much as 60. Other regions clearly have nakedly exposed volcanic rock at the surface.
If a planet can have a water cycle, why not a dust cycle? Dust endlessly transported in a loop between low latitudes and high latitudes and back. The periodic cold, dry climate allows the equatorial regions to create the dust, while the higher-obliquity times release the reservoirs of trapped material in the polar caps. Some of it becomes stuck, some of it becomes free and there’s always the wind to scour more from the volcanic source regions. If there’s one thing Mars does well, it’s making dust.
AMAZONIAN VOLCANISM
The start of the Amazonian was three billion years ago. No matter how much we try, we can’t comprehend how distant that time is from us now. The earlier stages of the Amazonian – call it the first 500 million years or so – are further from us now than all but the very oldest terrestrial rocks, but that point marks the almost-end of our Martian journey. From the very beginning of Mars, when the hot mantle failed to set in motion a viable system of plate tectonics, we could have predicted that, after a frenetic start, all activity would tail off to a disappointing end.
So it’s with a degree of discomfort that we learn that there were volcanoes still erupting in the Amazonian. Why that happened is one of those greater mysteries. Mars has had an extraordinarily persistent and active life, far beyond anything we could have expected. Because of its size and its history, by this period it should have been, if not cold, at least cooling like a casserole in a turned-off oven – one with a very solid lid.
So we appear to have a record of volcanic eruptions throughout the Amazonian, of the same very fluid lavas that we had in the Noachian when volcanoes first rose up on the rim of Hellas. Tharsis is coated with such lavas, as is Elysium, and there is little variation from the beginning until now.
Alba Mons is the northernmost of the Tharsis volcanoes, and at first sight it’s barely a volcano at all. It’s fantastically flat for most of its construction, which stretches east to west for a full sixty degrees of longitude, almost 900 kilometres side to side. It’s pretty much the size of France.
But Alba Mons appears to be two volcanoes, one on top of the other. The broad, near-flat apron of lava is surmounted by a more conventional shield volcano that is itself 360 kilometres across and 2 kilometres high. All this is balanced on the very edge of the Tharsis Rise, where it glides down into the northern plains. Surrounding it, and cutting through much of the earlier apron, are fields of linear tension faults – the Alba Fossae to the west and the Tantalus Fossae to the east both wrap around the dome-like summit, while to the south, the Ceraunius Fossae run steadily northward until they encounter the mass of the volcano, and falter against it.
The apron lavas are solidly Hesperian in age, but the summit is not, and neither are the two lobes of lava that extend east and west of it. The early Amazonian lavas were just as fluid as those that erupted in the Hesperian, but not quite as abundant. And even when Alba Mons appeared to be shutting down a few tens of millions of years later, there was still enough magma below for two small shield volcanoes to have appeared within the calderas of the collapsed summit.
Towards the middle of Tharsis, the three Tharsis Montes were also busy erupting. They continued to do so throughout the early and middle Amazonian, topping out their monumental structures at over 14 kilometres above the surrounding plain. And while their main building phase finished within a few hundred million years – itself an extraordinary length of time – they simply kept going. The youngest lavas on the flanks of Ascraeus Mons are barely twenty million years old.
Meanwhile, over on Elysium Planitia, between Elysium Mons and the on-the-Dichotomy Apollinaris Mons, potentially the youngest flood basalts on Mars occupy parts of the Athabasca Valles, with an estimated age of just two million years. In the planet’s long history, this counts as yesterday.
If we take the suggested ages in these studies at face value, then there is one obvious, startling proposition we can make: volcanism is not yet finished on Mars. Mars appears to be still active in ways that we didn’t think possible even a few years ago.
This is the enigma at the heart of our study. What are the chances of us sending spaceships to Mars at the precise moment that active volcanism – which has lasted for the best part of 4.5 billion years – ceases? We know that many, if not all, of the processes that have shaped the planet throughout its history, and certainly throughout the three billion years of the Amazonian, are continuing. The water cycle, powered by ice sublimation and deposition rather than water evaporation and condensation, is not just a yearly seasonal event, but one that extends over longer climatic periods and can melt the ice caps and allow free-flowing water on the surface. The dust cycle is intimately entwined with that movement and is a major source of new rock. And now there’s the prospect that volcanism is not finished, and that we have simply had the misfortune to arrive in a quiet period.
Our view of Mars as cold, airless and dead is in need of serious revision. Its heart may be slowing, its lungs barely inflating, its blood sluggish, but there are planetary-scale systems still in motion. It might not be exactly what we want to see. For our own selfish reasons, we crave the idea of a second Earth – and to have one so close and yet so far is tantalising – but we know that Mars was never that. It was never ours. It is its own and we have to accept it for what it is.