PART TWO
(More than 4.5 billion years before the present)
THE GIANT MOLECULAR CLOUD
If I’m going to explain Mars to you, I have to go back to the time-without-time, before the solar system existed. Reaching back beyond 4.5 billion years is going to be stretch, but what happened then is still apparent now. It’s all connected, a long chain of unbroken links.
Everything that Mars was and is started as a cloud of cold smoke, just like everything else in our solar system: our Sun, the planets that orbit around it, the accompanying panoply of smaller objects and the atoms that make up our own bodies. It was a cloud so vast that it stretched for 600 trillion kilometres and contained enough matter to form thousands of individual stars.
This cloud was made from the sweepings of the galaxy itself, the leftovers from its own formation and the detritus of earlier giant stars that had already burned themselves out. In the space between the matter-packed spiral arms of the galaxy, gravity was a negligible force, but it was not nothing, and so a tenuous drift of stuff coalesced. The cloud could just about keep itself together by merit of not being close to any object large enough to disrupt its fragile existence, but it still had a limited lifespan: it might fade away without consequences, it might collapse and give birth to a few massive new stars which would then burn off the rest of the cloud, or it could simply wait for the next star-rich galactic arm to pass through and spread it out again.
Despite being thin and filamentary, the cloud was still dense enough to block out starlight. Most of the molecules of gas within it were hydrogen, the default element that matter organises itself into in the universe, but other elements and molecules were present, so sparsely spread that they rarely collided, and so cold that there was no hope of a chemical reaction when they did. Inside, the cloud was almost as sterile and inert as it was outside.
Out in this wide interstellar gap, nothing existed that could stir the cloud. But for this story to be more than just a prologue, we need an initiating event, something to start the clock and begin the count. The shockwave from a supernova is one of those things.
The larger a star is, the faster it consumes its nuclear fuel, and the largest of all will attempt to burn their own ash – helium, carbon, neon, oxygen, silicon – until their cores collapse abruptly under their own weight, like a tower falling even before the upper storeys have registered that anything has happened. This triggers a supernova, the sudden brightening of a dying star. Spectacular from very far away, they are terrifying and deadly close up – close being a relative term; life on Earth would be severely impacted, if not ended, if one occurred within thirty light years of us.
For our interstellar cloud, the supernova shockwave came in two parts. The first was a blinding flash of intense energy, enough to bathe surrounding star systems in lethal radiation; the second, lagging behind, was a wave of matter, moving away from the stellar corpse at speeds of up to 10 per cent of the speed of light. A wall of particles – including exotic heavy elements that are rare in a universe dominated by hydrogen – slammed into our molecular cloud and mixed with it. Although launched from billions of kilometres away, the impact was enough to create local variations in density within the cold, inert smoke.
For the first time in millions of years, the gas began to move. Gravity’s weak grasp started to strengthen and the cloud broke up into discrete, albeit still huge, trillion-kilometre clumps. As these clumps spread out from their original positions, the subtle movements of the gas warped them into shrinking, rotating discs of material, each capable of birthing at least one star, and potentially a planetary system of its own.
Inside our unexceptional disc, matter falling from the edge towards the centre delivered energy in the form of heat as it collided with other particles. Matter interacted with matter. It stuck together. It began to pool, and over the next hundred thousand years, the initially thin, almost abstract gas formed a structure at the centre, dense and warm. That increasingly dense middle, where the vast majority of the material ended up, started to run away with itself. The more gas it attracted, the hotter and denser it became. Rather than the heat stealing away into space, energy was trapped behind the layers of falling, compressing dust. The central mass turned thick and dark: this is what would become our Sun.
Outside, the leftovers formed a thin rotating disc of matter: the protoplanetary disc. From this, everything else was fashioned: the whole halo of objects that circle our Sun, from the closest to the furthest away. The flattened disc stretched for 30 billion kilometres from the centre. Collisions of material within the disc were common, and as the disc heated up, the lighter elements – hydrogen and helium – were pushed to the outside, while the heavier elements became dominant in the midplane of the disc and towards the centre.
Chemistry happened. Where present, solid metal and silicon oxides condensed directly out of the gas, using oxygen to form stable molecules. Heat fused particles together. There was dust now, grains of matter, and these grains coalesced and melted and grew larger to become chondrules – millimetre-sized blobs of molten silicate glass. Chondrules were the basic, if tiny, building blocks of the solar system, forming and reforming during collisions over the next three million years, trapping material that might otherwise have been pushed to the edge of the disc. We find chondrules in meteorites today, frozen into the rock, packed together with dust and fragments from the protoplanetary disc.
As further shockwaves rippled through the material, these chondrules stuck to each other and collected dust on their molten surfaces. They collided. They clumped. As they grew larger, they experienced more drag against the gas they encountered: as their speed dropped, they orbited ever closer to the centre. Further out in the rotating disc, where it was colder, it began to snow. Fluffy ice crystals made of volatile, low-boiling-point compounds stuck to each other, creating mobile drifts that slowly orbited in the gas-rich extremities.
It’s at this point that things become difficult to fathom. If Mars is an unreliable narrator, the solar system speaks of itself in fragmentary, cryptic phrases. We know what we ended up with: a yellow star, a quartet of rocky inner planets, a gap filled with scant rubble, and then two giant gas planets, two giant ice planets and a corona of ice and dust. But the processes involved in getting there from our collapsing dusty disc, thick with hot glass beads and tiny, fluffy snowballs? It’s a question without a definite answer. There are signposts we can follow, though, and some of paths they lead us down are quite extraordinary.
ACCRETION
Accretion is the gradual accumulation of solid material into ever larger masses by a continual and cumulative process of low-speed collisions. It’s what we assume happened in the protoplanetary disc, but there’s a series of unresolved problems with the process that happen at key stages and ought to have prevented accretion from happening at all. And yet, we have planets. Clearly something is wrong – or right – and we simply don’t understand enough about what happened.
The first problem is this: we know that if we push ice together, especially when it’s in the form of fine, crystalline snow, it sticks. A gentle amount of pressure causes a slight melting of the ice, and the two clumps of snow become one. The same can’t be said of glass beads. Chondrules and collections of chondrules, no matter how hard you mash them together, don’t form fist-sized rocks. Quite the opposite: two pebbles thrown together are likely to break apart into several smaller fragments, not create a larger one. This is called, without irony, the bouncing barrier.
Once an object within the protoplanetary disc grew large enough, roughly a metre across, it created its own gravity gradient and so could accumulate extra material to grow larger still. But getting to this boulder stage from mere pebbles was fantastically difficult. We think this happened gradually, piece by piece, but we don’t know how for certain. Perhaps the region where these metre-sized rocks formed was hot enough for the chondrules to become sticky. Perhaps snowballs falling in from the outer disc managed to gather sufficient rocky material to form boulders before the ice evaporated. Or perhaps it didn’t happen this way at all: maybe accumulations of chondrules formed clumps and travelled around inside the disc as nothing more cohesive than a loose train of gravel caught in an eddy of gassy dust, which slowly accumulated enough mass to begin accretion through gravity.
Having swerved past one problem, we move seamlessly into the arms of another – something called the drift barrier. Once these small snowy or rocky boulders formed, they became large enough to experience aerodynamic drag from the slower-rotating gas within the disc. Thus slowed, they spiralled inwards. If they’d ever reached the central mass, they would have fallen into it and been lost forever to the surrounding disc. Indeed, in the complex computer models used to simulate the protoplanetary disc, all metre-sized boulders that form in the flattened extremities do get quickly sucked into the dense dark region at the centre. If this had happened, the disc would have become sparse and depopulated, with nothing left to build planets from, so both the bounce barrier and the drift barrier must have been overcome somehow.
As more material accumulated in a decreasing number of larger objects, the dust and the gas in the disc itself thinned out. Collisions that might have shattered a smaller mass now grew them as gravity played its part. Larger bodies, now kilometres across – planetesimals – swept up the remaining debris and each other.
Beyond the snowline – the radial distance from the centre of the disc beyond which frozen water, and other gases cold enough to, condense – planetesimals were composed of loose accumulations of chondrule-rich rocks and ices and, trapped within the mass, actual gases. These are the classic dirty snowballs we now call comets, and they dominated the outer part of the disc. Closer to the hot centre, ices were unstable and the planetesimals consisted of metal-rich silicate rocks composed entirely of chondritic material. This gave us the first major division in our solar system: rock and metal in the inner solar system; rock, ice and gas in the outer.
In a mere three million years, the disc had evolved from slowly rotating smoke to thousands of barrelling mountains of rock and ice. At the centre, where most of the matter was, the journey towards stellar nuclear ignition was continuing as more material fell inwards, heating and crushing what was already there.
So on to a third problem. We had kilometre-sized planetesimals in chaotic orbits about the centre, scouring debris from what was left of the disc as well as devouring each other in increasingly violent collisions. Most of those collisions, large and small, reduced the orbital speed of the planetesimals and therefore their distance to the centre. As a consequence, all our models suggest that it’s the innermost regions of the disc that were the densest, and so it’s here that the largest planets ought to have formed.
This is what we see when we examine other planetary systems around different stars. When we knew of just one solar system – our own – we thought of it as typical and we assumed we’d find similar arrangements of planets elsewhere. But since our discovery of exoplanets (planets orbiting other stars, and we find them almost everywhere we look), it seems that our particular configuration of planets is far from usual. The average planetary system has planets that are all roughly the same size and evenly spaced apart. The larger the planets, the further they are from their neighbours. And as a rule, other planetary systems form far closer to their star – all of the planets lie inside the orbit of our innermost planet, Mercury.
Clearly something extraordinary must have happened early on in the formation of our own solar system that meant we have the arrangement we see today. The four rocky inner planets are unequal in size: Mercury is half the mass of Mars, which is in turn one-tenth the mass of Earth and Venus. Then there’s a gap: the asteroid belt isn’t a failed planet – the total accumulation of asteroids in this region weighs less than 1 per cent of Mars’s mass. Jupiter and Saturn are distant giants, freakishly huge gassy bodies; further out are Uranus and Neptune, smaller and icier than the other giants but still vast.
Jupiter – along with the other giant gas and ice planets – is in entirely the wrong place. It should be much closer to the Sun, and therefore the rocky inner planets shouldn’t exist at all. Mars should, by rights, be the same size as its two closest inward companions. There should, perhaps, be a fifth rocky planet of similar mass beyond it, where the asteroid belt currently is. But this is not the solar system we have.
The odds of our solar system turning out the way it did are around one in a thousand, and potentially even smaller than that. Any working theory of the formation of solar systems has to account for so much strangeness. We reach for answers and they elude us.
PLANETARY EMBRYOS
Three million years after the collapse of the giant molecular cloud, the solar system was ready for its final phase of planetary growth. The protoplanetary disc had evolved from being dominated by gas into one dominated by discrete lumps of matter – a debris disc.
At its centre was a swollen bolus of heat that accounted for almost all the gas from the earlier cloud. This was the protostar – it had all the mass of the Sun, but the nuclear fusion that would make it shine for the rest of its life had not yet begun. As the shrouding dust cleared sufficiently to let light pass through, the protostar was finally revealed in its monstrous glory: balefully huge and blindingly luminous – four times larger and sixty times brighter than the Sun is now.
As the protoplanetary disc transitioned into its debris disc phase, the infall of supersonic gas, dust and high-boiling-point compounds reduced, and the outside of the protostar cooled. From being in balance – where the urge to pull its surface inwards due to gravity was equalled exactly by the pressure of the heat pushing everything apart – the seething, roiling ball started to shrink. The matter at its very core was crushed ever closer together: lithium atoms were torn apart and hydrogen formed from helium, liberating enough energy to slow the contraction, but not stop it. The process took millions more years and would end with the full-scale nuclear fusion of hydrogen.
Orbiting this hyper-luminous protostar were the tens of thousands of kilometre-sized planetesimals. It was a chaotic time, and collisions were frequent, relative speeds low and accretion rates high. Planetesimals attracted each other even if they avoided collision during any particular pass: their gravities interacted, pulling each other into new orbits that could bring them into contact the next time around. Some experienced runaway growth, absorbing all the other planetesimals they encountered. This quickly led to fewer, larger bodies that collided only rarely, and when they did, it was with greater energy. Fragmentation, melting and mixing became important for the first time.
But where in the debris disc did this great accumulation happen? We know from our models where it ought to have been – where the greatest density of matter was, close to the primary star – and it’s reasonable to assume the same for our own solar system. Our early rock-based planetesimals all orbited in a narrow band close to the centre, a doughnut shape with well-defined borders. The inside border was close to the protostar, far inside the present-day orbit of Mercury: any closer and it would have been so hot that even metals and rock would have boiled off into gas. The outer border for rocky planetesimals was no further than where Earth currently orbits; beyond that there were insufficient bodies to accrete with.
Icy planetesimals stretched that border outwards somewhat, but as they collided and grew, they were subject to the same forces as their rocky counterparts and they moved closer in. The whole of our solar system’s planetary formation process became squeezed into a comparatively busy and narrow region. Somewhere in there was Mars – or rather what would become Mars, a proto-Mars that existed as a thousand-kilometre-wide planetary embryo at this point, along with the other few hundred planetary embryos formed by runaway accretion. They began to dominate their own particular orbits. They collected other stray planetesimals and increased their size without being deflected from their paths.
We didn’t end up with a few hundred planets. While these embryos formed the apex in the hierarchy of the rapidly clearing debris disc, there was a little way to go yet. Proto-Mars, along with proto-Earth, proto-Venus and proto-Mercury, was already present, as were dozens of other embryos that never managed to gestate to term. Critically, just on the other side of the snowline from proto-Mars was the monster that killed most of them: Jupiter.
Together, Jupiter and Saturn represent 90 per cent of the planetary mass of the solar system. Jupiter alone accounts for 70 per cent. It’s about as big as a planet can get – if it had more mass, gravity would make it denser and therefore smaller in size. Forming where it did, near the snowline, it started as just one of the planetary embryos in the debris disc. Roll the dice a thousand times, and in nine hundred and ninety-nine cases Jupiter would have become a reasonable-sized rock-cored planet, covered in layers of ice. But we live in that exceptional timeline where Jupiter voraciously consumed mass – as well as attracting ice and rock, it was still ploughing through the remnants of the dust and gas that were more abundant in the outer edges of the disc. It grew into a body that was over a hundred Mars-equivalents.
As it grew, Jupiter slowed and spiralled inwards. As it crossed the current orbit of the asteroid belt, it gobbled up everything in its path: planetesimals and planetary embryos alike, material that should have coalesced with proto-Mars but now never could. Earth and Venus were still relatively well supplied, while Mercury – half the mass of Mars – appears not to have grown much beyond the planetary embryo stage. Objects that Jupiter failed to absorb, it flung around the disc. Saturn was briefly left behind at this point, but in the absence of the ever-inward-moving Jupiter, it too started to accumulate gas. It grew rapidly and followed Jupiter towards the inner region of the disc.
There was potentially nothing to stop Jupiter’s inward march. We see planets of similar size around other stars, parked in very close orbits, where the heat of the star strips away their vast reserves of gas. These ‘hot Jupiters’ have a limited lifespan. They eventually become naked cores, huge baked cinders of planets perpetually roasted in the intense starlight. But Jupiter did stop. More than that: it moved away again, and it dragged the rocky planets back out with it.
Orbital resonances are odd things. At first glance, two planets orbiting a star seem so distant from each other as to be unable to affect one another’s motion at all. And yet, planets and moons can and do fall naturally into cycles, the subtle drag of gravity pulling bodies towards each other as they pass by. For much of their formation period, Jupiter and Saturn had a 3:2 resonance: for every three orbits of Jupiter, Saturn made two. When Jupiter moved inwards, that resonance was broken.
But when Saturn fell sunwards too, it reattached itself to the 3:2 resonance frequency. Jupiter, then located around the current orbit of Mars, started to move out again, caught up in its resonance with its smaller outer companion. Jupiter retreated to beyond the asteroid belt once again, scattering debris across the solar system for a second time. Saturn’s presence pushed Uranus and Neptune further out, into the icy debris at the very edge of the disc, and possibly caused a hail of distant snowy planetesimals to arc inwards towards the more-populated centre.
Giant planets swinging backwards and forwards through the early solar system like sailing ships, knocking rocks aside as they went, sounds preposterous. And yet the fine mill of computer simulations shows that this scenario is not just possible but probable. We have to get here from there, and it seems there are very few ways to do it. Can we explain the low mass of Mars? Why do we have an asteroid belt? Where are the missing terrestrial planets? Why is everything so far away from the Sun? How did Jupiter become so vast?
This theory – the Grand Tack theory – is our current best guess. If our calculations are correct, it was a startlingly swift process: Jupiter’s innermost incursion, reaching the current orbit of Mars, took place a mere 100,000 years after it formed, and it moved back out to its current orbit in just another 400,000 years. There are variations on the Grand Tack theory, and there are other theories that end up in the same place (of course they do, otherwise they’d be rejected), but in the Grand Tack’s favour, it continues to give gifts long after it’s finished.
For now, we have marshalled everything into position. Our solar system is complete and all the planets are in their orbits. Proto-Mars is at the right distance from the protostar; it has the right mass and the right chemical composition. Mars the planet is ready to be born.