PART FIVE
(3.7–3 billion years before the present)
THE HESPERIAN CLIMATE CHANGE
The beginning of the Hesperian era is defined by a dramatic change in Mars’s climate. The late Noachian had been mild and generous – rich with water and covered by a thick blanket of atmosphere. It would have been still, obviously, uninhabitable by us, but it had potential.
Sustained volcanism changed everything. Once magma broke through to the surface in sufficient quantities, gases poured into the atmosphere and the quality of the air altered. One single volcanic eruption might have had a significant local effect on air quality – in terms of not only dust and ash, water and carbon dioxide, but also sulphur dioxide and other sulphur compounds, hydrogen fluoride, hydrogen chloride and carbon monoxide. But more eruptions, either all at once or intermittently and repeatedly over tens of millions of years, fundamentally altered the global balance.
The ocean, up until this point, was mildly alkaline, and conditions were conducive to the production and laying down of clays. While extra carbon dioxide and water vapour bolstered the thinning Martian atmosphere, the other volcanic gases readily dissolved in rain to form acid solutions. The slow drip-drip of acid then fell on the highlands and found its way into the northern ocean and the crater-lakes. It shifted the balance from clay to the laying down of sulphates – gypsum and other salts.
Mars’s water gradually became more acidic, and it stayed that way from then on. Whether or not that ended any nascent Martian life is unknown, but extremophiles are usually highly specialised organisms. Those that are resistant to heat are not usually also resistant to acid, and vice versa.
It’s thought that Hesperian-age volcanic rock eventually covered almost a third of Mars – much of it in the form of flood basalts in the northern lowlands, as well as on Tharsis and on the floors of highland craters, notably Hellas. Much of this seems to have started in the late Noachian and intensified in the early Hesperian.
On its own, volcanism would have raised the temperature of Mars a little, but there was a greater and more dramatic change happening far above sea level. The volcanic sulphur dioxide aerosols high in the atmosphere reflected sunlight back out into space, causing the air below, and therefore the ground, to cool rapidly. The gases tended to wash out quickly – within decades – and turn into acid rain, but because the volcanoes constantly belched out more sulphur, there was always enough to keep the levels up. The immediate effect was a global chilling. Frost formed where it had never formed before. High-latitude lakes iced over. The longer those conditions persisted, the deeper the cold bit. The first frosts on the equator signalled that the climate had tipped in favour of winter.
Water-saturated ground turned from being merely ice-covered into permafrost. Ice penetrated tens of metres deep. Lakes froze from top to bottom. Ice caps formed, grew larger and crept towards the lower latitudes. In the highlands, snow fell, accumulated and didn’t melt in summer. Persistent, year-round ice turned into glaciers.
Locking away water as ice interrupted the cycle of evaporation–condensation–precipitation. The northern ocean began to contract. The seabed was exposed. Lakes dried up. Most critically, the water vapour that kept the air pressure high was now trapped in crystalline ice, and the thinning atmosphere became even worse at holding heat.
The effect was cumulative and reinforcing. Throughout the Hesperian, Mars turned from warm, wet and pressurised to cool, dry and gasping. Each puff of yellow volcanic gas caused another fractional drop in temperature, which created more ice crystals, which led to a drop in air pressure, which meant the heat capacity of the atmosphere fell a little further. The colder temperatures meant less evaporation, which meant less rain to wash out the sulphur aerosols, which then persisted for longer and led to further cooling.
If Martian volcanism had been for but a season, the climate might well have recovered. But after the initial surge of eruptions had altered both sky and land, substantial periodic volcanism throughout the Hesperian and beyond locked in those changes. There were temporary, million-year-long revivals: as the planet’s obliquity turned to greater angles and the ice caps retreated, the planet warmed and water flowed again. Meteorites periodically struck the old ocean floor and melted enough of the ice and generated enough heat for lakes and even seas to form. But this was always in the context of the overall trend. It was never enough to reset the clock. The slowing of the astronomical bombardment played its part, as did the solar wind gradually stripping away the atmosphere. There was no turning back.
Mars was entering a new phase, one dominated by fire and ice.
THE BEGINNING OF THE CRYOSPHERE
That ice can be as all-pervasive as air and water is a notion that those who’ve experienced life at the poles will find easier to grasp than the rest of us, but the consequences for a previously warm planet that freezes from the surface downwards are profound and long-lasting. The cold spell that started in the Hesperian has lasted, on and off, for over three billion years and into the present day, so we need to understand the effect it has had in order to continue with Mars’s story.
As the cold penetrated water-soaked surface sediments, it froze the ground solid. Water, when frozen, occupies almost 10 per cent more space than when it’s liquid. This expansion caused Mars’s surface to buckle and heave into new shapes, disrupting the river- and lake-formed features and blurring them. Periodic warming collapsed the permafrost, causing slumping, landslides, surface sinkholes and temporary rivers. As colder conditions returned, everything slowly froze again. Landscapes changed, their previous forms erased.
Volcanic rock was also affected, although differently. When hot rock cools, it contracts, breaking along lines of stress. Joints form through the layers and water can seep down into them. Newly formed Martian ice forced those joints wider and, at the surface, shattered the rock into sharp shards. Below, where the tension had nowhere to go, the pressure inside the rock grew enormously.
The permafrost bit hard and deep. As the years turned into millennia, the ice layer penetrated further down into the crust, creating a thick – and watertight – cap over much of Mars, potentially a kilometre or more deep. This cryosphere of ice replaced much of the hydrosphere of water – although not completely, as deep down in the warm darkness, underground currents still flowed. But above, the water cycle of evaporation–condensation–precipitation became locked up in an ice cycle of sublimation and deposition. The ice did flow, but only slowly, and it needed to be renewed by fresh falls of snow. When it wasn’t, it was scoured from the top by the wind and carried away, leaving only bare frozen ground behind.
Exposed free water at the surface became ephemeral. The northern ocean armoured itself with ice. The water underneath remained protected for a while, but as the biting cold continued, liquid water froze onto the underside of the floating ice, until it had all solidified. The water currents that might have driven the ice above to break up and to flow, ceased. Deeper craters held on to their water for longer, but eventually they too succumbed.
There were limits to the march of the ice. The permafrost extended down only so far as the freezing cold won out over the rising heat of the still-warm mantle. In active volcanic areas, where the heat-flow was much higher, even the surface might have remained unfrozen, for a while at least. Hydrothermal vents stuttered on, but as they were reliant on the percolation of water down to the warm rock far below, they dwindled to hot springs and geysers, and then to nothing. The life that might have been associated with them? Unless it retreated far underground, it would have been extinguished.
Over the ocean, meteorite impacts, if large enough, punched their way through the ice and into the ground below, generating tsunamis of ice, water, steam and molten rock. On land, an impact in the permafrost unleashed an ejecta blanket of fluid, muddy debris that then froze in place. The crater stayed warm as the heat dissipated and became an oasis in a frozen desert for years, perhaps centuries – or even longer if the crater was later inundated with lava from below – but in the end this would have frozen over too.
Volcanoes could also erupt below ice sheets, either as a new vent, or in the case of an existing volcano, beneath a glacier that occupied its summit. Heating the base of the glacier would have created huge quantities of meltwater which, once it saturated the ground underneath, had nowhere to go except out, any way it could. Those spontaneous and abrupt flooding events scoured existing channels down to the bedrock and below, and created new ones where none existed before. The water may have been fleeting, but the scarring effect was permanent.
There was a third way in which liquid water could have appeared at the surface of frozen Mars: for this we will have to go deep underground.
VALLES AND CHAOSES
There are thirty named chaos terrains on Mars, each characterised by the same jumble of flat-topped blocks, kilometres wide, tilted at every angle and separated by steep-cut valleys hundreds of metres deep. They are each set into a depression and associated with a vallis – a broad river valley with obvious flowing-water features that seem to empty away from the chaos. Some of the chaoses are huge, hundreds of kilometres across in total. They look like unhealed scars on the landscape: crazed and creased and thin.
There are almost as many theories of chaos formation as there are chaoses, but they all seem to coalesce around the sudden and repeated upwelling of great volumes of groundwater, directly out of the subsurface, that broke through the thick lid of permafrost. Such a phenomenon isn’t really found anywhere else, and certainly not on Earth, but the chaos landforms are reminiscent of places where we know sudden flooding has happened in the past, so what follows is very much a best guess. It’s dramatic and slightly complicated, but it does make a good story.
As Mars froze solid above, enough heat rose from the mantle for the crust to remain warm and thus for water inside it to stay liquid. That water may have been trapped under a solid, waterproof layer of ice and rock, but it still flowed through the cracks and pores of the fractured subsurface, rising and falling, swirling in underground currents, driven by differences in temperature and pressure. Water was potentially added to the system at the poles through the melting of ice beneath the ice caps, but all the circulation took place at depths of several kilometres. Those vast movements seemed destined to stay hidden; if they climbed up far enough to touch the permafrost layer, their flows became sluggish and icy.
Yet they were under enormous pressure: the water was trapped and squeezed by the weight of the frozen rock above. If we could have gone there and drilled into it, we would have been rewarded with a gusher of warm, salty water rising high into the thin Martian sky, freezing in the sparse, sub-zero air, falling as snow and simultaneously sublimating into vapour as it fell. Whatever the initiation event was, be it volcanic, tectonic, meteoritic or simply hydrostatic (the rock failing due to the fantastic strain the water placed on it), once some of the water had managed to punch through the ice layer, the breach inevitably widened and let out more. What this would have looked like seems almost impossible to describe, yet we have to try.
At first, there would be a rumble, a marsquake, and then a jet of water as hard as a steel rod would cut its way through the ground and roar out. Rock, eroded from the walls of the initial shaft and entrained in the jet, would widen the channel quickly and weaken the land around it. One failure would grow into multiple points of breakthrough – fractures spidering out deep underground and breaching the surface. The noise and the vibration would grow and grow until, suddenly, whole mountains would move. The seemingly solid ground would quiver; it would all sink, but some parts would sink more than others. Slabs would break off, tilt and tip, and water would be everywhere. A vast bolus of water – a sea’s worth, a spontaneous flood – would leap into existence where moments before there had been nothing but bare, dry, frozen rock. It would boil out of the ground, an explosion of liquid.
That seething, churning, mud- and rock-laden warm-water wave would gouge its way to the lowlands, digging out a channel, carving its sides, shaping its passage and only slowly losing speed and potency as it travelled. Eventually, tens or hundreds of kilometres downriver, it would subside. As the wave of sediment-laden water poured into the craters and out over the plains, the heavy material would drop out first, and the dust last, to form a single graded deposit running from coarse at the bottom to fine at the top.
Behind it, at its source, the first burst of groundwater would slacken, and as the throughput slowed, the ground would start to refreeze. The permafrosted rock layer, a huge reservoir of cold, would rob the water of its energy and the flood would turn into a river, a spring and then shut off like a tap.
The surface water, still seeking the lowest point, would pond in craters and channels, cooling and resting. Ice would form above and below until the water froze solid. The mud and the rock would freeze with it, potentially forming rock glaciers – huge, ice-cored rock flows where the ice is protected from sublimation or melting by a covering of debris. Weeks and months would pass, and although the landscape would be utterly changed, all of the surface water would have already either frozen in place or slipped back underground. The vents that brought it to the surface would be stopped with ice. The ground would quieten.
Beneath the permafrost layer, the underground water would begin to refill the depleted joints and pores, like a desiccated sponge swelling back up. With the weaknesses already created and ready to exploit in the rock above, the critical threshold of pressure to breach the surface would be lowered. The chaos may have to wait a month, a year, a decade or a century before the ground cracked open again, but it would, allowing more water to spew out along the already-formed channels, to wash them clear and widen them further. And so on, until the wound of the chaos closed over completely and the aquifer below was left so empty that it could never refill enough to break through again.
The chaos landscape that was left would be sharp and hard, as yet uneroded or smoothed over by dust. City-sized slabs of the earlier surface would lie where they’d fallen, heaved up and over by the inconstant ground. Separating them would be scoured land and flood-washed canyons. There would be new valles that marked the passage of the flood water, which had gone where it wanted, exploiting earlier lines of weakness and making new ones. Out on the plains, new layers of water-rich flood sediments would have settled and then frozen. Inside craters, the infill would have raised the crater floor, burying features like rings and central peaks, potentially leaving only the very rims exposed.
The time of chaoses eventually ceased, but they rumbled on through the Hesperian. The outwash from them coated the northern plains all the way up to the north pole. The Vastitas Borealis, the Acidalia Planitia, the Utopia Planitia and the Arcadia Planitia are home to the majority of the exposed Hesperian-age sediments, along with the Chryse Planitia, which lies at the mouth of the Valles Marineris. They are anywhere from tens to thousands of metres thick – an extraordinary amount of material transferred from the highlands to the lowlands.
Time eventually softened the chaoses. Frozen debris slumped down, and fresh frost attacked the cliffs and turned them into curtains of scree. The valles blurred with wind-blown loose sediment: dunes formed on the riverbeds and sand piled up on the leeside of obstacles.
The floodplains now crackle with frost: they heave and sag in season with the surface temperatures, and add some small measure of relief to an otherwise featureless horizon.
HERE ARE GIANTS
There are twenty-eight large volcanoes or clusters of volcanoes on Mars, and many more smaller vents and cones. Volcanoes on Mars run the gamut of sizes, from the biggest beasts in the solar system to structures that aren’t even bumps in the ground. Most of the very largest are contained within Tharsis. Purists will tell you that next-door Olympus Mons, which counts as the absolute tallest volcano and the second tallest mountain in the solar system, isn’t part of Tharsis. But whatever process built Tharsis and the region’s mighty volcanoes – Alba Mons, Arsia Mons, Ascraeus Mons, Pavonis Mons, Uranius Mons, Uranius Tholus, Ceraunius Tholus, Biblis Tholus, Ulysses Tholus, Tharsis Tholus – also sowed Olympus Mons from the same seed.
All volcanoes start the same way. Magma forces its way up through the ground from a pressurised reservoir somewhere deep underneath. When it breaks through to the surface, volcanic material – whether it’s lava or more explosive products like fine ash or cinder bombs – is expelled from the vent. When more of this collects closer to the vent than further away from it, a mound forms. Continuing or repeated eruptions build on this unpromising beginning, layer after layer, metre after metre. That’s it – year on year, century on century, millennium on millennium, the vent widens as the rock of its throat is scoured by the upward force of magma, and it rises up as the volcanic material accumulates around it. This process continues for as long as the volcano is fed fresh magma from below.
Small volcanoes can grow quickly and fall extinct just as rapidly. The more substantial ones can go through cycles of waking and sleeping, with periods of dormancy that are sometimes so long that erosion starts to wear away the summit. The magma chamber below the volcano might retreat, cool, contract and solidify; any empty space left inside the mountain will eventually be filled from above as the summit of the volcano collapses into it, erasing the volcanic crater and leaving the cliff-bounded, flat-bottomed structure we call a caldera.
Some volcanoes have just one caldera. Others have evidence of several, and crater counting on their floors shows that these volcanoes have lived and died and lived again through hundreds of millions of years, sometimes longer. Such longevity and such resurrections are only possible because of Mars’s unique set of circumstances. On Earth, a more mobile crust shifts the volcano – and the magma chamber that supplies it – away from its mantle heating source. On Mars, the volcano might remain over the same hot spot forever.
The oldest volcanoes appear to be those in the southern highlands; at least, they stopped erupting first and are left as scabs surrounded by heavily cratered lava fields. Worn down by billions of years of erosion by rain, flowing water, ice, dust-laden wind and periodic meteorite strikes, they’re barely recognisable as volcanoes at all. The layers of lava that surround them, and their rimless collapsed calderas, are all that remain.
Of the younger volcanoes, here’s the slightly disappointing truth: if we were to visit one, we’d struggle to realise that we were on the flanks of a volcano at all, let alone one with a kilometres-high peak. From above, they look hugely impressive: great, broad-shouldered structures with caldera-punched summits. Arsia Mons boasts a 100-kilometre-diameter caldera that lies a full kilometre below the highest point of the rim above it. Lava flows fan out on all sides of the lower slopes in scallop-edged formations, stacked up one on the other, raising the volcano five, ten, fifteen kilometres above the surrounding plain. Other features – side vents, lava tubes, collapse structures like landslides and chain pits – relieve the monotony of one step of solid lava after another.
But because lava on Mars seems to have been mostly very fluid, the volcanoes there are nearly flat. The average slope on Ascraeus Mons is four degrees. For every 100 metres walked, we’d climb 6.5 metres. It’s not nothing, but neither is it the steep-sided cone of popular imagination.
Standing on the lowest slopes of Ascraeus, looking up in the direction of the summit, all we’d see is the rise, not the top. And standing at the top – 18 kilometres above datum – we’d have no view at all. We could look down into the caldera, but the rest of the compass circle would be a near-level horizon a few kilometres distant. And even if we knew the direction of the next volcano in line, Pavonis, which is itself 14 kilometres tall, the curve of the planet would prevent us from seeing it at all – either the base or the summit. From our high vantage point, the horizon would be just over 350 kilometres away, and Pavonis is 800 kilometres distant. As mighty as the volcanoes of Tharsis are, they’re more or less invisible from the ground. The climbs are gentle, the summits hidden and the highest points marked by sudden kilometre-tall cliffs falling away beneath us.
Tharsis doesn’t appear to have had any volcanoes as such until the Hesperian; before then, all the gain in height and weight was due to uplift of the crust, subsurface intrusions and eruptions from rifts in the ground. Single-point volcanism was new. When it did happen, it started small, with Biblis Tholus and Ulysses Tholus in the east, Tharsis Tholus in the west and Ceraunius Tholus and Uranius Mons in the north. These are all significant volcanic structures, but they’re utterly dwarfed by the three west-of-centre Tharsis Montes volcanoes, Pavonis, Ascraeus and Arsia Mons, by Alba Mons in the north and especially by Olympus Mons in the far west, all of which grew later.
The volcanoes of Tharsis are strange by nature: complex and ancient and so very persistent. Given the Methuselah-like lives of these vast blemishes on Mars’s skin, to call them all extinct is an act of hubris. For now, they’re apparently quiet, and that’s all.
VALLES MARINERIS
Valles Marineris is another of those ludicrously oversized features that Mars seems to specialise in. It is a series of connected, steep-sided, flat-bottomed trenches that stretch for 2,000 kilometres from east to west through the almost-but-not-quite middle of the Tharsis region. These trenches – one of which is properly called a chasma, the plural being chasmata – are 200 kilometres wide, and their floors can lie 10 kilometres below their rims.
Of course this feature has to do with Tharsis – the sheer weight and size of it – but there’s more to Valles Marineris than that: there are some spectacular deep-crustal elements that combine to make this vast crack in the Martian surface simultaneously inevitable and awe-inspiring. It even created a lake system that led to some of the most significant flooding events Mars would ever see.
Before we consider how Valles Marineris formed, we ought to look at a map of it, better to contemplate its heights, depths and widths. From north to south, there are four lines of chasmata. The first consists of the shortest chasmata: Echus Chasma and Hebes Chasma, two just-unconnected valleys in the north and off to the west. Then there are three parallel scars, each with a breach in their wall into the next: Ophir Chasma, Candor Chasma and then the largest and longest system, which is really one valley but is divided up into three sections: Coprates Chasma in the east, Ius Chasma in the west and Melas Chasma in the centre.
At the far western end is the uniquely Martian landscape of the Noctis Labyrinthus, which looks like a plate that’s been dropped onto a stone floor and then clumsily glued together again. It’s a shattered region composed of cliffs with slumped and broken sides that tumble into valleys, separating high islands of the original surface. These smaller chasmata branch and split and rejoin, gradually shallowing to the far west, as they head towards the high ridge of the Claritas Rise.
At the far eastern end of Coprates Chasma there are three additional trenches, but these are aligned differently and they all meet at their eastern extremities: Ganges Chasma in the north and Capri and Eos Chasmata in the south open into the Aurorae Chaos.
The sides of an individual chasma are almost vertical – in the case of Candor, they descend 7 kilometres in one single step. The sheer cliffs have been subject to later collapse – blank walls transformed by footings of frost-shattered debris, the broken rock smoothed over with washed-down highland sediment and further blurred by wind-blown dust. But nothing can fully hide these extraordinary wounds. They are too big, too obvious and they crave an explanation.
Again, like so many phenomena on Mars, there are several competing explanations. Perhaps the chasmata are rift valleys, formed when the crust to the north and the south moved apart and stretched the ground in between – ground that might have been superficially healed by overlying Tharsis lavas, but below was still all fractured crust, kilometres-wide blocks of rock broken off from their neighbours by huge impacts in Mars’s early history.
Or maybe the chasmata were formed by the action of flood basalts scouring channels in Mars’s crust as if they were rivers – hot rock pouring from west to east, tearing at both the walls and the bed, carrying and melting the debris as it passed.
Alternatively, they could be the result of explosive outbursts of groundwater, activated by heat rising from below, that undermined the surface and dragged all the overlying rock and dust away in a series of cataclysmic, biblical floods that spent themselves in the chaotic terrain of western Chryse.
For certain, there are volcanic and river deposits within Valles Marineris. There are small cindery cones on the floors of some of the chasmata, and there’s evidence of lava flows. Water has clearly passed down the valley system, too. There’s more than enough evidence for both of these hypotheses.
None of these stories quite satisfy, though. Why there? Why then? There are compelling reasons to believe that the chasmata of Tharsis were the result of deep-crustal movements that related to two utterly uncoincidental factors: the steadily growing size of Tharsis and the position of the Great Dichotomy.
We know that Tharsis formed on top of a stationary mantle plume, which added material above due to volcanic lavas and below due to intrusive magma. Tharsis rose and stretched and flexed throughout the Noachian and into the Hesperian, with centres of activity that slowly switched from south to north, to east to west, and back, but that always accumulated weight, always grew and spread.
The great new mass didn’t sit lightly on the ancient crust. It bore down on it, causing it to sink deeper into the mantle. But it didn’t sink evenly, because the weight of Tharsis straddled the Great Dichotomy. The crust to the south is an average of 60 kilometres thick; the crust to the north is half that. They sank at different rates in response to the same weight, creating a line of violent tension that radiated all the way up from the bottom of the crust, where it joined the mantle, to the very top, where Tharsis was unthinkingly adding more lava and jamming in more intrusive rocks.
The crust, and Tharsis above it, cracked in a series of parallel tears. Great rectangular blocks, aligned with the surface cracks, fell into the zone of maximum tension below. Lubricated by magma, they sank down an unprecedented distance and found a new, stable floor, kilometres below.
Whatever the trigger for the collapse of the crust beneath Tharsis, the Valles Marineris formed quickly. The early Hesperian seems the most likely time for the initial stages of collapse, given the ages of the lavas at the top of the formations.
But there’s more. The measured depths of the chasmata appear to be much greater than required to balance the weight of the sinking blocks, even for something as huge as Tharsis. They should have only shifted downwards a couple of kilometres at most, so why do we find them much lower than this? This time, the answer lies not below but above ground.
The newly formed chasmata obviously provided low points on the surface of Mars. We know that water was present at the time: we can still see evidence of river systems that fed into the chasmata from the high Tharsis plains, but groundwater was more likely to play the greater part in this new, colder Hesperian era. In the same way that the chaoses formed from overpressured aquifers below the permafrost layer, the kilometres-deep trenches simply exposed the water-rich layers to the open air.
The sinking troughs were rapidly inundated and became repositories for debris washed or blown off Tharsis from both north and south. Then it was just a matter of how much lower the weight of that could drive the chasmata floors: any movement down would have made room for more material to be laid on top. A stack of sediment potentially 8 kilometres thick formed, representing the whole height of the deepest parts of Valles Marineris. The resulting lakes, deep and still, most likely periodically roofed with ice, ran the whole length of Valles Marineris, connected not by surface channels but by subsurface joints and porous rock.
Careful crater counting indicates that while the chasmata are early Hesperian in age, the outflow channels at the far eastern end, where the flood-scar of the Aurorae Chaos sits, are half a billion years younger, from towards the end of the era.
The top of the lake system was kilometres above the eastern lowland, with the sediment forming a natural dam. It’s likely that water drained from the lakes by overtopping the dam, either continually, seasonally or during those longer, obliquity-inspired climatic variations. Over time an outflow channel, caused by erosion, would have formed. Every time water ran along it, the channel in the dam would have become deeper, wider and shorter, eroding backwards towards the lake.
Another day, another minor collapse. The channel opened wider and the new flow was stronger. The water tore harder at the banks and the bed of the channel, until suddenly the dam failed. All that water, penned up for so long, was suddenly free to move.
Most probably, like in the chaos-and-valles systems, flooding happened repeatedly, with the barriers failing intermittently and in stages, gradually moving the water level of the lake lower and lower. Much of the sediment that had collected in the lakes ran with the floods. There’s evidence of layered deposits in the chasmata. Kilometres-high piles of lake-type sediments are still found on the floors of many: long ridges of isolated basement islands like the Eos Mensa, the Ius Mensa and the Coprates Mensa, or banked-up deposits near the chasm walls, categorically distinct from the rock type of the walls or the bottom of the chasmata. The soft sediment was mostly carried away, though. The amount excavated was typically heroic in volume – millions of cubic kilometres of stuff redistributed onto the northern plains of the Chryse Planitia.
The Aurorae Chaos sits below the outflow from the Valles Marineris, right in the firing line of the inland tsunami. Other chaoses lie along the path north of it – chaoses that might not have been formed by the upwelling of water underneath them, but by water tearing through.
We know what the Chryse Planitia looks like, intimately. Both Viking 1 and Mars Pathfinder – with its rover Sojourner – landed within its compass. It is an undulating plain with low hills and shallow depressions, and the ground is closely dotted with loose black rocks that range from shards to boulders, all interspersed with red soil. It looks a mess. The rocks have not been smoothed over time by steadily flowing water, but are rugged and chipped from frost. They are unsorted, dropped and abandoned. This is not a gentle terrain, but scuffed and wind-seared. A mountain has been thrown down here: lava blocks on an outwash plain, a memory of floods that once crossed the land and were then stilled.
OLYMPUS MONS
Olympus Mons is the largest volcano in the solar system and the second tallest mountain. It was the bright spot that nineteenth-century astronomers named Nix Olympica (the snows of Olympus), meaning that it’s large enough to have been spotted from Earth using nineteenth-century telescopes.
Olympus Mons bears all the hallmarks of a shield volcano: this is a volcano with a roughly circular footprint, low-angled slopes running from the base to the summit and a wide, open crater – or caldera – at the top, so that it resembles a round shield dropped on the ground. Also typical of a shield volcano, Olympus Mons has lava flows and fronts down its flanks, collapsed lava tubes and, where the ground has been excavated by impact craters, exposed layers of solidified lava. The largest shield volcano on Earth is the Hawaiian island of Mauna Kea, which is 10 kilometres tall from seabed to summit. Olympus Mons is on a different scale altogether.
Olympus Mons rises 21 kilometres above datum, and because it sits on a below-datum plain to the west of the Tharsis Rise, we can add another kilometre to its actual height. East to west, it’s 640 kilometres across. North to south, it measures 840 kilometres. As we’ll discover, it used to be even larger. The slightly off-centre summit – which tends to the south-east – has six identifiable calderas that nest inside each other at the top, over an area roughly 80 by 60 kilometres. Lava overflowing the summit of Olympus Mons would have had to travel almost 400 kilometres before it reached the base of the volcano, staying liquid as it moved down the gentle gradient, pushed from behind by more erupting lava.
So how did Olympus Mons form? At some point in the early Hesperian, a hot spot in the mantle melted the crust beneath the Amazonis Planitia, and lava first came to the surface. The area where this happened is close to Tharsis, but close is a relative term – the three volcanoes that make up the Tharsis Montes are over a thousand kilometres to the east. But something definitely happened deep below this part of Mars that built this vast edifice rapidly. Most of Olympus Mons was complete within a few hundred million years, still within the Hesperian.
The reasons for believing this speed of construction are a little tentative, but they are compelling. The Hesperian was the high point of Martian volcanism, and some surfaces, hidden away on the extremities of Olympus Mons’s long flanks, most likely date to the Hesperian. While the rest of the volcano is now veneered with later lavas, the core of it still sits inside.
The two most obvious features of Olympus Mons, after its sheer height, width and breadth have been considered, are these: firstly, the cliff that circles its base and secondly, the ring of debris that extends hundreds of kilometres out from there in obvious, discrete flows. Such features are not usual for Martian shield volcanoes, and they need explaining.
Shield volcanoes are formed by the flow of very fluid lavas. This defines them – it gives them their shape, their internal structure and determines how they interact with the encircling landscape. Lava is supposed to run down the flanks of the volcano from the summit and from vents in its side as rivers, forming lava tubes and levees that initially confine the flow and then break out later into broad fronts that spread onto the encircling plain. This is how a shield volcano slowly increases both its height and its diameter without substantially altering its overall shape. Shield volcanoes are supposed to have low-angled slopes of one or two degrees at the point where they meet the surrounding land. There definitely shouldn’t be a cliff, let alone one that is, in places, 5 kilometres high.
Then there’s the debris ring, usually referred to as the aureole, which stretches out in all directions from Olympus Mons’s base; north and west across the Amazonis Planitia, south towards the Dichotomy and east in the direction of Tharsis. There are several distinct flows of material that form long fields of rubbly terrain, resting behind vast arced fronts of piled-up, thrown-down rock.
The obvious conclusion is that parts of Olympus Mons have broken off. The falling material formed giant landslides, which swept across the plains and ended up lying where they spilled, either stretched out on the lowlands or banked against local ridges. That conclusion encompasses the idea that the edge of the volcano was originally some 100–200 kilometres further out than its current cliff-faced position.
But there are immediate problems with such a scenario: kilometre-thick layered lava flows are really very solid and not prone to spontaneous collapse, especially when the deposits are at angles barely above one degree from the horizontal. Also, there’s no known mechanism for a landslide to travel between 200 and 700 kilometres from its starting point – even factoring in Mars’s reduced gravity, lower air pressure and every other conceivable environmental component. It simply couldn’t happen.
And yet, the obvious conclusion is somehow correct. If we gather up all the landslides, trace them back to their beginnings and pack them up against the cliffs, we can make a perfectly shaped shield volcano out of them. Our assumptions must be wrong, somewhere.
They are. Detailed pictures of the cliff faces don’t show serried banks of black lava flows edge on, but hard ridges of lavas interleaved with soft, grey, undifferentiated rock that avalanches dust. The angle of the cliffs isn’t quite as sharp as expected either: rather than being near vertical, the average slope is somewhere between twenty and thirty degrees. This is steep, but certainly not as steep as could be made from solid blocks of basalt. The indication is, then, that the outer reaches of Olympus Mons aren’t constructed primarily from rugged rock, but from loose ash, which if stacked up high enough, might well become mechanically unstable.
Even then, a collapsing mass of light ash and pumice can hardly be expected to travel hundreds of kilometres in a single coherent flow, before slumping neatly down on the ground with such a well-defined border – let alone repeatedly, towards all points of the compass. It might stretch to tens of kilometres, but the friction between the components of the slide would be high, and the energy of the fall would soon dissipate.
However, going back to the earlier contention that Olympus Mons was mostly formed in the early Hesperian, we might have a way out. What if these were not dry, dusty landslides composed of rattling rock that might have arisen out of the failure of a 5-kilometre-high slope? What if the base of the volcano had actually been submerged in a Hesperian-age ocean, and the collapse, transport and deposition of the aureole had all taken place underwater?
This proposition is startling given that everything we thought we knew about both the longevity and the depth of the northern ocean suggests that a deep sea shouldn’t have been there at that time. But if – if – Olympus Mons was surrounded on all sides by water, then the erosion of its ashy flanks by waves would explain the presence of cliffs, and it would also explain the size and shape of the aureole.
Underwater landslides are of an entirely different quality to those in air. The water turns the debris into a dense slurry that minimises contact between the solid materials within the flow, while the difference in density between the flow and the surrounding water also prolongs the distance it can travel by lubricating the gap between the solid material and the seabed. We know such things happen around Hawaii, where submarine slumps and debris flows are common to all the volcanic islands in the chain.
In context, then, Olympus Mons probably started as a submarine volcano on the sloping sea-floor west of the Tharsis Rise, sometime during the late Noachian or early Hesperian. It grew rapidly over the next ten to a hundred million years, breaching the surface of the sea on its way to its 21-kilometre height. While lavas flowed from the central crater, and probably also from vents on the flanks, the outer reaches of the volcano were largely composed of loose ash and other aerial volcanic debris. Saturated with water, the edges of the volcano repeatedly failed, forming huge submarine landslides that travelled underwater for hundreds of kilometres, before running out of energy or banking up against submarine ridges. Each collapse was associated with a tremendous tsunami that echoed up to the pole and back. The scarred cliffs carved by the collapses formed scarp slopes that were twenty degrees steeper than either the aureole deposits or the volcano itself.
This scenario does require the northern ocean to have persisted for longer than we had thought. It does require the crater-counted ages of the aureole to be wrong – the rubbly plains around Olympus return later dates. But other explanations are difficult to swallow. Transporting that amount of material that far from its source without the aid of some kind of fluid is highly unlikely: wind will not do, and neither will purely volcanic processes. The deposits in the aureole didn’t form in place: their shape and position show they were moved there from the flanks of Olympus. And surely we can’t suggest a deep ocean that lasted even longer? To push it from the Noachian into the Hesperian is difficult enough.
We are left with a paradox: Hesperian-aged deep-water deposits in a sea that should have been both retreating and freezing by that time, from a shield volcano that seems to have an outer portion composed of ash instead of lava that dates – using our most reliable technique – to an even later time. Such is Mars. We acknowledge its mysteries and move on.
ELYSIUM
Tharsis isn’t the only volcanic province on Mars. It would be simpler if it was – then we would have one mantle plume that created the Great Dichotomy, and then wandered and created Tharsis. Yes, volcanism also occurred in and around large craters, where the thinned crust rose and the drop in pressure at the boundary with the mantle caused spontaneous melting, but much of Mars is underlain by flood basalts that appear to have no connection with either the craters or the plumes, and we simply have to accept that other mechanisms – ones that relied on very different conditions – were always in play.
The other major volcanic province is Elysium, where three volcanoes – one mons and two tholi – rise out of the Elysium plain in stark contrast to the low relief that surrounds them. Situated to the north of the Dichotomy, the largest volcano in the group, Elysium Mons, rises 12 kilometres above datum from a floor that is 2 kilometres below it. To the north is Hecates Tholus, a mere 8 kilometres tall, and to the south is Albor Tholus, only 6 kilometres high.
Elysium Mons is closer to what we think a volcano should look like. Its upper slopes are comparatively steep, but despite that, the lower slopes are so shallow that there’s disagreement as to where the volcano ends and the surrounding plains begin. Still, at some 400 kilometres across, Elysium Mons is a beast of a shield volcano, topped by the usual nested caldera that indicate several collapses – widely spaced apart in time – of the magma chamber below the peak.
Looking at the map, there’s a sinuous line waiting to be drawn from Hecates, through Elysium and Albor, continuing southwards through the Cerberus region where flood basalts poured from rifts and barely raised small shield volcanoes, to Apollinaris Mons, sitting directly on the Dichotomy, and then further into the highlands to connect with Apollinaris Tholus and Zephyria Tholus, a low-relief shield volcano sitting on ancient Noachian rock. Making a connection between these volcanic centres is tempting and leads us to see a pattern of northward drift by some rogue plume fragment.
From all we can ascertain, though, there was no grand cause. Elysium began erupting in the very late Noachian or shortly after the transition to the Hesperian. Hecates started some half a billion years later and Albor a billion years after that. The eruptions on Cerberus look much, much younger. Just as Tharsis has no easy south-to-north ageing pattern, neither does Elysium, and we can only acknowledge that the order of eruptions was, if not arbitrary, then at least controlled by processes we do not yet understand.
Once again, we have to consider the extraordinary longevity of volcanism on a world where plate tectonics did not operate to create, move or destroy crust. Elysium Mons was mostly completed, built from the ground up, within a few hundred million years of the first eruption, and it still continued to erupt sporadically for the next 1.5 billion years. For whatever reason, Elysium’s molten rock stayed under the Elysium region for four billion years.
However, the power of the Elysium generator was an order of magnitude less than that of Tharsis. The Elysium region rose up from the surrounding plain, but it doesn’t have the weight that Tharsis has. Elysium Mons is a significant volcano in its own right, but it’s just under half as high as Olympus Mons. Tharsis dominates.
One property that volcanoes in the Elysium region do seem to possess is that at least some of the eruptions there weren’t comprised of fluid lavas pouring from shallow cones and fissures to flood the surrounding terrain, but were composed of explosive, volatile-rich magmas that created significant quantities of fine ash, and with that, scalding-hot clouds of volcanic debris called pyroclastic flows.
Lavas that are more viscous, or sticky, keep their gases entrained for longer: as they rise up towards the surface, the sudden drop in external pressure causes the gas bubbles in the lava to abruptly expand, forcing it to foam and spray. This lava spray turns into the microscopic fragments of volcanic rock we know as ash; larger globules turn into frothy pumice. The eruption itself takes the form of a heavy, fiery blanket of gas, ash and rock that rolls down the flanks of the volcano. Often following existing valleys or folds in the landscape, these types of pyroclastic flow can drive on for tens of kilometres, and when they collapse they can add metres’ worth of material to the ground in a single pass.
Other scenarios are even more violent. If the viscous lava forms a plug in the vent, pressure can build up behind it, and when mechanical failure occurs – either of the plug or of the containing vent – then the resulting surge of debris explodes outwards and can lay down huge quantities of rock, much of it gouged out from the existing volcano.
The steepness of the upper slopes of the Elysium volcanoes appears to indicate that the lava here was stickier than at Tharsis, at least some of the time, and that this would have been associated with both pyroclastic flows and voluminous ash production. In fact, volcanic ash plays a major role in one of the Hesperian’s most significant geological formations – one that directly affects how we see Mars today.
THE MEDUSAE FOSSAE FORMATION
The one thing that we know about Mars – that we’ve always known – is that it’s red. Since the first people looked up at the sky, since we constructed the first names in the first languages, Mars has always been the planet of fire, war, destruction and death, and it has been associated with gods that have those aspects. In Sanskrit, it was called Angaraka; in Egyptian, Her Deshur; in Hebrew, Ma’adim – all these names mean ‘red’.
But Mars is not inherently red. Much of the rock that covers the surface is black basalt. The sediments that form from the erosion of this rock are mostly grey. The ice caps are white. Yet the colour of Mars is the colour of rust, from a sandy ochre through deep browns and reds to a dust-laden, baby pink sky. Where did this redness, this rustiness, come from?
The obvious answer – that it is rust – is both right and wrong. We can identify the chemical compound that’s responsible for the redness, but also know it couldn’t have been formed by the usual method.
Rust is the reaction of iron in the presence of water and oxygen to form iron hydroxides, which then desiccate to become iron oxides. Iron oxides are brittle compared with iron, and because they expand away from the surface of their parent metal, moisture and air can creep in through the outer layers of rust and corrode beneath them. While there’s no native iron in Mars’s crust or mantle – the thorough melting of Mars shortly after its formation sequestered all the metal iron in the core – iron-bearing minerals, silicates and oxides containing iron, form part of the usual suite of crystals found in both the crust and the mantle.
These iron minerals can break down in wet, oxygen-rich environments, gain more oxygen and turn into haematite. This transition is important, because while most iron minerals are black, haematite is red.
However, although Mars was wet at times, it was never – not once – oxygen rich. In dry climates where there’s plenty of oxygen, only a thin veneer of rust can form, as it’s water that opens up the chemical pathways and allows the easy transport of reagents. But in wet climates where there’s no oxygen, no reaction can occur because the basic building-block of oxidation is not present. Without atmospheric oxygen, there’s no oxygen dissolved in the water. The usual mechanism for converting black iron minerals into red haematite is simply not available.
Yet there’s demonstrably red haematite on Mars, invariably as a component of the ubiquitous dust that blows around the entire planet. If we can talk about a water cycle that circulates water from ocean to clouds to ground and back, then we can also talk about a Martian dust cycle that circulates dust. Not all of the dust is haematite – a substantial fraction of it is magnetite and other volcanic silicates – but almost all of the haematite is dust. It’s this dust that coats highlands and lowlands, that layers within and on top of ice. This is what makes Mars red.
Even the sky is red. On planets with an atmosphere, the sky looks blue because the molecules that make up the air scatter the blue part of sunlight more effectively than the red part, diffusing the blue light across the whole of the sky. As the Sun sets, the low angle of lighting means the blue is scattered out into space, and we tend to see more red. At the opposite, extreme, scenario, where a planet has no atmosphere, the sky is always black, even at midday, as there’s no air to scatter the light.
Mars has an atmosphere, albeit a thin one now. As a result, we’d expect the sky to be blue-black during the day. Instead, it’s pink from dust so fine that it can be carried high by the Martian winds. Part of the effect of this is to add a red hue to everything else, whatever its original colour.
Knowing why Mars is red still leaves us with two problems: firstly, how does the dust become red when we’ve no obvious way of making haematite? And secondly, how is there so much dust? The two questions are linked – of course – but we need to deal with them in reverse order: dust first.
This brings us to the Medusae Fossae Formation, which is a deposit of rock hundreds of metres thick that sits on the lowland plains, just north of the Dichotomy, between Elysium in the west and the edge of the Tharsis Rise in the east. Precisely what the Medusae Fossae Formation consists of has been the subject of fierce debate for decades: is it lake sediments, or a stranded raft of lighter-than-water pumice, or old glacial deposits or a carbonate shelf that formed at the edge of a shallow sea? For a while, it was even thought to be a vast deposit of dust, caught in the lee of the Dichotomy slope – a dust trap to end all dust traps.
Careful analysis brought about a surprising conclusion: the Medusae Fossae Formation isn’t where dust goes to die; it’s where it’s born. The chemical signature of the global dust is very similar to that of the exposed rock of the Medusae Fossae, with just the right proportions of sulphur and chlorine. Material is being eroded from there and then dispersed by the Martian winds to every part of Mars.
If that’s the case, then the Medusae Fossae Formation has to have been significantly larger than it is now to account for the sheer amount of dust on Mars, and it should also show wind-eroded features, both ancient and modern. That is what we find. The landscape is rich with signs of wind erosion: not just dunes and ripples, but more sculptural formations such as yardangs – carved, steep-sided ridges, tens of metres high and kilometres long, which align with the predominant wind direction – and pedestal craters, where an impact has locally fused softer sediment together, creating a hard-to-erode circle around it.
Because the Medusae Fossae Formation appears to be composed of loosely cemented volcanic ash and pyroclastic material, its surface is being continually reworked. It exists as both a primary deposit, where it has been laid down and is now being excavated by the equatorial wind, and also as a secondary formation, a wind-blown layer of sediment that overlies much younger lavas. It’s a fascinating, mobile mess of tearing down and building up, and it makes calculating a crater-counted age enormously difficult.
It’s almost impossible to pinpoint the formation’s origin, but there are volcanoes nearby. If it is made from ash, the fact that it’s flanked by the two major volcanic regions of Mars isn’t going to be a coincidence, and Apollinaris Mons sits centrally on the southern edge of the deposit. Yes, crater counting can give a falsely young age – the whole technique is based on crater preservation, not crater erasure – but the surfaces the formation sits on are late Noachian or early Hesperian, and Hesperian and later lavas lie both on top and underneath in interleaved layers. The early to middle Hesperian is a guess for when it was formed, but a reasonable one.
Why red, though? Ash is grey, magnetite is black. How did this dust change colour in the absence of oxygen and, increasingly, in the absence of water? It turns out that it’s the wind itself that creates the haematite through a process of mechanical milling.
Sand and other fine grains, when blown by the wind, move through a process known as saltation: essentially, a grain flicks up from the loose surface, is carried a short distance and then lands again. During its flight, the grain might hit another, and when it lands it certainly will. Each tiny impact abrades both it and whatever it strikes. As the particles get smaller and smaller, and create more and more dust, the surface area of each particle compared with its size increases dramatically, and this gives greater and closer contact during collisions. It’s during these collisions that chemistry can occur that would normally only happen at higher temperatures and in the presence of water.
Mill a mixture of nine parts quartz sand and one part magnetite in a carbon dioxide atmosphere for long enough and you’ll end up with haematite dust. The exact mechanism for how this occurs is poorly understood, but it does work. Pick up enough of the Medusa Fossae Formation and carry it in the wind for a few billion years, and Mars will gradually, inevitably, turn red.
TRUE POLAR WANDER
We know that Mars’s polar axis wobbles – that the angle of obliquity (the angle which determines where the poles are pointing) changes periodically and chaotically. However, if we took a globe of Mars, put a finger on each pole and spun it, whether we held it upright or at a steeper angle, our fingers would still be on the poles. The poles remain in the same place on the globe, and the globe always rotates about the same axis.
But true polar wander is where the axis of rotation itself moves – the poles physically shift. This isn’t unheard of in general, and it isn’t unheard of in the history of Mars either. It’s already thought to have happened once, when the Great Dichotomy formed. That event was so long ago that no real evidence of it has managed to survive, but that the Dichotomy line ended up parallel to the equator – implying the poles moved to accommodate the change in the distribution of Mars’s mass – is well established. How, then, is the Dichotomy now twenty degrees askew from that?
That this is the case is undeniable. The evidence is there in the differing depth of the crust, measured from the gravity map, and also in the line we can draw on the surface between the southern highlands and the northern plains. And then there’s the belt of late Noachian and early Hesperian river systems that were formed in a band twenty-five degrees south of the equator – the old equator, that is – fed by rain coming in off the northern ocean. Map out all those lines now, and they make sinuous, parallel paths around Mars, crossing the new equator and back. In order to straighten these inconsistencies out, the poles need to have shifted.
Where were the old poles? There are more than a few degrees of uncertainty here, but the north pole was probably located in the part of the Vastitas Borealis called the Scandia Colles, while the south pole was somewhere in the Malea Planum, to the south of Hellas crater. These are the places we need to look to for evidence that – for perhaps 1.5 billion years or so – these areas were covered, at least some of the time, by ice sheets.
So what do we find? The Scandia Colles is characterised by odd, hummocky terrain that’s suggestive of both subsurface ice melting that collapsed the ground above, and also liquid water run-off from a retreating ice cap. The water would have gone deep underground, but there’s still a significant amount of water-ice in the area. It’s promising evidence, but not conclusive.
In the south, the Malea Planum is characterised by long ridges, valley systems and other features like muddy ejecta splashes from cratering events that suggest unusually saturated ground in an area that should have been dry, cold desert, given where it sits in the highlands. Again, not conclusive evidence, but what we find doesn’t contradict our supposition that a pole used to lie here.
So if we decide that the poles did move, that Mars’s axis of rotation did slide twenty degrees from where it was to where it is today, at some point between the formation of the rain-belt valley systems of wet Mars and present-day frozen Mars, how could it have happened? The answer is obviously Tharsis, but that’s not the whole story.
The centre of mass of Tharsis, its point of balance, is below a point on the western flank of Pavonis Mons, the middle of the three huge Tharsis Montes volcanoes that dominate the west of Tharsis. This point now lies directly on the Martian equator, yet if the preceding information is true, then this same point was once twenty degrees north of the equator, before the poles moved.
Tharsis on its own has the same mass as the dwarf planet Ceres, and it covers over a quarter of the surface area of Mars. Something that huge could, and eventually would, unbalance the axis of rotation of Mars itself. At a certain point it grew so large that the planet’s spin – its moment of inertia – became unsteady. There needed to be a correction, and that was best achieved by shifting the axis such that Tharsis’s centre of mass was sitting exactly on the line of maximum rotational speed: the equator.
So far, so extraordinary. That it happened is agreed upon, and we’ve located a cause. All we need to do now is identify a mechanism to allow for it. But turning the weight of an entire planet against its existing moment of inertia isn’t something that’s easily accomplished.
One of the things we do know is that solid crust can be carried along, at a rate of millimetres a year, on a slowly moving mantle current. It happens on Earth; it might happen on Venus; it might have happened on early Mars. But this mechanism applies to rafts of crust broken into continent-sized slabs, with new crust being formed in one place and being dragged down and destroyed in another. By the Hesperian, we’re certain that Mars’s crust – brittle, fractured to a depth of tens of kilometres, measuring 60 kilometres thick in the south and 30 in the north – acted as a single ‘stagnant lid’: a solid, continuous shell of light, cold rock enclosing the totality of the mantle and pierced only in very few places by volcanic eruptions.
Could the crust have moved as a single entity, hauling itself into its new position, not pulled by movements of the mantle, but driving itself using Tharsis as the counterweight and dragging the mantle along with it? It’s a journey of over 1,100 kilometres between the current pole and the old pole, one-twentieth of the circumference of Mars. Over a billion millimetres. At a rate of two millimetres a year, it would have taken 500 million years to complete the process. No one said this was going to be fast – but it demonstrably had to happen.
Is there any other evidence we can find for this slow, inexorable rearrangement of the geography of Mars? Again, we’re confronted by the problem that we’re looking back billions of years into the past. Such a gradual turn would have minimised the stresses on the rigid crust, but it wouldn’t have eliminated them completely. The Martian surface is scored with both extension cracks (graben valleys) and compression features (wrinkle ridges); not all of them are associated with Tharsis and its extra weight, nor can the remainder all be linked with the other volcanic provinces. Might they be evidence of the shifting crust? If we map them out and try to make sense of the patterns, it turns out that the results are patchy, confusing and contradictory.
There is, though, one feature on Mars I have already mentioned that might lend this story an extra layer of credibility. Take a look at the Tharsis Montes, the three principal volcanoes of eastern Tharsis. From south-west to north-east, these three huge shield volcanoes – Arsia Mons, which stands 12 kilometres above the surrounding plains, Pavonis Mons, 14 kilometres, and Ascraeus Mons, 15 kilometres – lie almost exactly in a line.
Three huge volcanoes in a line speaks of a common cause: a single plume of mantle material underneath that part of Tharsis. Either the plume moved, breaking through the thickened Tharsis crust each time, or Tharsis itself moved and the plume remained stationary. We have precedents for both: the single mantle plume that tracked towards the Dichotomy in the early Noachian; and the linear Hawaiian islands on Earth, shield volcanoes all, built successively over a stationary plume as the Pacific plate travels at a relatively brisk 9 centimetres a year across it.
At this distance, we cannot tell which was the cause of the Tharsis Montes; it is most likely a bit of both. But it does look extraordinary. The clearest indication would be a firm date for the first eruption from each volcano, but that data is denied to us, buried under kilometres of subsequent lavas. All we have to go on are tentative last dates, given to us by crater counting the collapsed summits of each volcano in turn, and this tells us only that volcanism persisted long after the Hesperian ended.
As with much of Mars’s history, we are left with answers that may be wildly, extravagantly wrong. At least, though, we know what the questions are. Mars’s poles of rotation have moved and we have a culprit, even if we cannot quite work out how they committed the act.
THE ICE CAPS
The move into the Hesperian brought physical and climatic changes at even the highest of latitudes. True polar wander was dragging the poles away from their previous positions, incrementally, into new territory. Because the planet was cooling, there were extensive periods of time when the temperature never rose above the freezing point of water, even at the equator. Mars’s atmosphere was also thinning: water vapour was dropping out as snow and staying locked up as ice. Even the carbon dioxide was starting to solidify as temperatures at the poles plummeted. The paradox was that the lower the pressure became, the more likely carbon dioxide was to form a gas at any given temperature. Dry ice sublimated away as the pressure fell and formed a positive feedback loop that prevented both temperature and pressure from reaching absolute zero.
Ice caps formed – potentially irregularly and inconstantly – at Mars’s poles. Periods of high obliquity exposed the poles to long summers and long winters, and seasonal snow and ice came and went across large swathes of Mars. In periods of low obliquity, it stayed uniformly cold at the poles all year round: snow accumulated there, pressed down on the layers beneath, and fluffy snow became hard ice. When a warmer climate returned, as it did chaotically for tens or hundreds of thousands of years at a time, the ice caps, both north and south, retreated. The resumption of the cold climate reinflated the ice caps and they crept back across previously ceded ground.
But as the Hesperian progressed, the lack of new snowfall, the sublimation of carbon dioxide and true polar wander meant that the existing ice caps began to retreat in earnest. They did so unevenly, especially in the south, where there was more geography to interfere with the weather patterns. Cold sinks – areas which were unusually frigid for their latitude – sat between the pole and Tharsis, and also on the high ground between the Argyre and Hellas craters; the depths of the craters themselves, with their higher air pressure and lower elevation, were warm by comparison. The southern ice cap was always colder than its northern counterpart, due to its being some three to four kilometres above datum. It was simply colder at altitude when the atmosphere was relatively thick.
When an ice cap melts, it leaves behind footprints – telltale markers of interaction between ice, water and rock – and when we look at the Martian polar regions, we see these signs in abundance. In the north, there are underground ice lenses and surface collapse structures, especially in the Scandia Colles region, where the old pole is believed to have been. In the south, the Dorsa Argentea Formation is crossed and crossed again by features called eskers.
When water melts on the surface of an ice cap, it runs through crevasses and joints within the ice, widening them into tight tunnels which work their way down from the top of the ice cap, all the way to the bottom. Melting also occurs at the base of an ice cap, where it sits on the bare rock below; pressure and rising heat – ice is an excellent insulator – can cause hidden lakes to form, entirely covered by the ice above. These under-ice rivers and lakes meet, and the flow of water picks up sediment – ground-down rock flour, entrained dust, larger broken fragments lifted from the basement by the dragging of the ice as it flows under its own weight to the edges – and it makes its own debris too, using what it’s already carrying to wear at the subglacial riverbeds.
Confined to their icy tunnels, these fast-flowing and voluminous rivers rinse through the ice caps until they emerge from the leading edge in great, frigid flows. There they spread out and drop their fine loads across sandy outwash plains. New river systems form outside the ice cap, allowing the water to continue on into basins near and far, until it settles and freezes solid.
If the ice cap retreats, the riverbeds are slowly exposed. Each ice tunnel melts down to just its banks on either side, which then disappear completely. The sediment that was confined within the tunnel settles on the bare rock, leaving a sinuous ridge on the surface, a bank of debris that marks the route of the river along the underside of the ice. These lines of sediment – these eskers – can be significant features, both in height and width, and also in how they print the landscape with an obvious climatic marker. If there are eskers present, winding their way across a plain, then there was once a melting ice sheet on top of it. And while eskers can be erased again by a resurgent ice sheet, they are remarkably persistent once they form – on Mars, erosion rates are low in the cold environment of the far north and south.
If the Dorsa Argentea Formation was once covered by ice, then the Argentea Planum was the temporary holding lake for the water on its northern edge. River channels from there cut across the Noachian highlands and snaked towards the locally deep low of the Argyre crater, where Hesperian-age sands and muds now cover the basin floor.
The ice caps continued to grow and fade in time with the astronomical calendar, but without replenishment their maximum extent would never be as great as it was at the end of the Hesperian. Mars’s drying and cooling was surely taking a fixed course now, not to be diverted.