Mapping Mars: Science, Imagination and the Birth of a World

Looked at each other with a wild surmise –

Silent, upon a peak in Darien.

John Keats, ‘On First Looking into Chapman’s Homer’

On 13 February 1969, nine days before Mariner 6 set off for Mars and five months before Neil Armstrong was to step on to the dust of the Sea of Tranquillity, the newly inaugurated president, Richard Nixon, asked his vice-president, Spiro Agnew, to explore the options for a post-Apollo space programme. Agnew became enthused. When Apollo 11 made its historic landing that July, he talked of committing the nation to the goal of sending people to Mars. The report of Agnew’s Space Task Group, offered to the president in September 1969, discussed this possibility and many others – but more or less ignored the question of how much it was going to cost. Nixon could not allow himself that privilege.

In May 1971, the month Mariner 9 was launched, the Office of Management and Budget (OMB) informed NASA that its budget, already significantly cut back from its mid-1960s heights, would be frozen for five years. On 5 January 1972, two months after Mariner 9 reached Mars, President Nixon authorised NASA to start work on a reusable Space Transportation System – the space shuttle. There was severe doubt – at OMB and elsewhere – as to whether this was wise; NASA’s claims that it would make space travel far cheaper were highly dubious. But it was the least ambitious thing on offer that would keep people flying into space. And people in space, even if they had nowhere particular to go once they got there, was an idea that meant something to Nixon and to many of the men around him.

From 1972 onwards the space shuttle was central to NASA’s institutional survival. A national means became the agency’s end. Almost everything else was either a distraction or, if it looked expensive, a threat. The planetary missions already approved – the Pioneer 10 and 11 missions to Jupiter and Saturn, and the Viking missions to Mars – were not in too much trouble. But missions not already accepted were delayed and scaled back. The ambitious TOPS probes to the outer solar system that JPL had been planning were replaced with enhanced, enlarged versions of the now ageing Mariner spacecraft design. In the end that did little harm – launched in 1977, the Voyagers were a spectacular success. But they were the last hurrah of the ’60s horde. Between 1979 and 1991 JPL launched only two more planetary spacecraft.

(#ulink_d426cdd0-dc7c-5b23-bad2-9813324668ba)

It was in this climate of cutbacks that the Viking landers lowered themselves to the surface of Mars in 1976. For years they sampled dead soil, analysed dry winds and photographed barren landscapes at two unprepossessing sites in the planet’s northern hemisphere. In engineering terms they were a spectacular triumph. Their accompanying orbiters, meanwhile, added huge numbers of new pictures to the Mariner archive. And that was just as well, since the Viking treasury was to be the raw material for most of the next two decades of Mars research. The Viking missions were the most expensive effort in the history of planetary exploration and their single take-home message, according to most of the scientists involved, was that Mars was as lifeless seen from the surface as it had appeared to be from orbit. Expensive, dead and already the subject of overflowing data archives; to NASA budget-setters Mars looked like a pretty good place not to return to.

Which didn’t mean that scientists stopped talk about new missions to Mars. At any given time there will always be lots of ideas for missions that someone or other dearly wants to see fly. Some are little more than water-cooler chatter. Some are studied but never approved. Some are approved but then dropped. Each one that flies leaves the ashes of a dozen other dreams in its wake. The field of planetary science is full of brilliant people in their forties who have still never managed to get an instrument they defined or built on to a spacecraft, never gaining the status of a principal investigator.

In the early 1980s one of the competing dreams was a spacecraft called Mars Geoscience/Climatology Orbiter. Its proponents admitted that, yes, it did seem that Mars was a dead planet, both biologically and geologically. Although there were arguments about how to date features on the surface – arguments which will be discussed later, along with many other scientific issues some readers probably think I’m passing over too quickly at the moment – most of the interesting events in Martian history were thought to have happened billions of years ago. But dead could be interesting and besides, Mars had only been studied from a fairly narrow point of view. Most of the data were in the form of pictures. To geologists like Hal Masursky and his crew, these pictures were great. Geologists are interested in stories about which rocks are where and how they got there. While pictures taken from orbit were not terribly good guides to the nature of the rocks, their form and arrangement – the morphology of the surface – were well captured, and that provided a lot of grist to the geological mill.

Geology, though, is not the only way to study a planet. Geophysicists are interested in understanding physical forces and processes, something they seek to do in large measure by building mathematical models. From this point of view pictures, while pretty, are no substitute for numbers. Geochemists are interested in the chemical elements from which planets are made up. Climatologists want to know whether they can understand the atmosphere’s behaviour. All these disciplines had an interest in Mars that the Viking data-set couldn’t satisfy. A modest orbiter dedicated to geophysics, geochemistry and climatology might be able to fill in the gaps in humanity’s knowledge of Mars – the mineral composition of its surface, its precise shape, the strength of any magnetic field, the structure of its atmosphere – with model-friendly numerical data.

The argument was pretty good, the prospective investigators were widely respected and the idea that the spacecraft could be a cheap modification of a design already used for satellites orbiting the earth was a plausible and appealing selling point. Indeed, the idea was intriguing enough that it started to grow. If a small geosciences spacecraft could be sent to Mars, why not send a similar one back to the moon? Or to orbit an asteroid? Buying the same design and components in bulk would keep the prices down, after all. And so the geoscientists’ Mars mission became Mars Observer, first in a new line of Observer spacecraft. Under pressure from geologists like Masursky – and with an eye to public relations – NASA added a small, comparatively cheap camera to the design; left to the geophysicists, Mars Observer would have no ability to take pictures in any usual sense of the word.

One of Mars Observer’s objectives was to get a detailed picture of the planet’s relief. The Mariner and Viking scientists had used a wide number of different techniques to try to calculate how high features on the Martian surface were. They used triangulations based on the visual images. They used the precise instants at which radio signals from orbiters were cut off as they passed behind the planet. They used subtle differences in the amount of infrared and ultraviolet light reflected from different parts of the planet through different depths of atmosphere. They used narrow beams of radio waves bounced off the surface by earth-based radio telescopes. All these different measurements were synthesised by Sherman Wu, in Flagstaff, to provide contours for the Survey’s maps. But even Wu did not think the elevations he painstakingly arrived at were accurate to more than about a kilometre.

Mars Observer was to sort all this out with an on-board radar system developed by a team at NASA’s Goddard Space Flight Center led by David Smith, a British geophysicist. Smith is a warm, affably excited man who, had he stayed in his native country, would be endlessly returning the smiles of women struck by his resemblance to the widely adored sportscaster Des Lynam. He had spent the 1970s applying the geophysical ideas attendant upon plate tectonics to studies of the shape of the earth, and he was excited about moving on to other planets shaped by other processes. Then, in late 1986, the shuttle struck again. Mars Observer had been scheduled for launch in 1990, but after the Challenger disaster the risk of the shuttle’s schedule slipping convinced NASA officials to delay the launch until the next time the planets were correctly aligned, two years later. Delaying by two years meant that the spacecraft’s costs went up, because it was not feasible simply to disband the teams already at work. Savings had to be made and so the two heaviest instruments were dropped. One was the radar.

David Smith was not going to give up. He convinced NASA to put $10 million on the table to produce a replacement instrument and, having looked at a couple of radars, decided to use a new, much less tested technology, one that bounced laser light off the surface instead of radio waves. People in Smith’s group at Goddard were already working on such an altimeter for the proposed Lunar Observer; a modified version became a relatively cheap altimeter for the Mars Observer. There were risks involved – no laser system had ever survived in space remotely as long as this one would have to – and the development was a little hairy in places. But they got the instrument finished on time and in budget. That was more than could be said for the rest of the mission. Partly due to the delays, Mars Observer’s costs rocketed – the notional later Observers were cancelled as a result. Then it was decided to launch on an expendable rocket rather than a shuttle, adding yet more to the expense.

(#ulink_43add16f-44f3-5e48-8a0a-ca4e2e8700e4) Then a hurricane hit the rocket while it was on the pad at Canaveral. Finally, on 25 September 1992, with the Mars Observer Laser Altimeter (MOLA) safely on board, Mars Observer got off the ground. And eleven months later, having been told to pressurise its fuel tanks in preparation for going into orbit around Mars, the spacecraft fell silent, never to be heard from again. It is more or less universally assumed to have exploded.

It was a terrible blow. Back when Mars missions were sent out two at a time, losing one was OK; Mariner 3 was lost, part of Mariner 7 exploded, Mariner 8 was lost, but Mariners 4, 6 and 9 did just fine. Mars Observer, though, was a singleton and the designers of its nine scientific instruments were bereft. Smith told me that while imagining ways in which the MOLA instrument itself might fail had come all too easily to him, he’d never imagined the whole spacecraft being lost. NASA’s administrator, though – a bullying, obstreperous but undeniably dynamic and often perceptive man named Dan Goldin – decided the loss was an opportunity. Goldin was sick of being responsible for the sort of space programme that launched only a couple of planetary spacecraft every decade and was determined to find ways of sending out more missions – ‘faster, better, cheaper’ missions, as he delighted in calling them. The first faster, better, cheaper programme, called Discovery, was to send spacecraft all over the solar system. Indeed, the second Discovery mission, due to take off in late 1996, was a Mars lander – Mars Pathfinder. (Mars Pathfinder was actually conceived before the Discovery programme; as its name implies, it was meant to be the first in a series of simple landers. The series of simple landers was cancelled and Pathfinder, like Mars Observer, became a one-off,

(#ulink_a2df6062-cbf5-5e91-9ead-b24e88ef730d) slotted into the Discovery programme for more or less purely political reasons.) Goldin and his advisers at NASA headquarters decided that a second line of faster-better-cheaper spacecraft should be devoted to Mars. In order to spur new thinking and greater efficiency, the size of the spacecraft and the budgets in this Mars Surveyor programme were to be tightly constrained.

The first of the missions was Mars Global Surveyor (MGS) and it has proved massively successful. Launched in November 1996, it arrived at Mars a few months after Mars Pathfinder’s landing on 4 July 1997. MGS carried copies of five of Mars Observer’s instruments, for the most part cobbled together out of spare parts. Soon after arriving it started a long series of passes through the thin upper atmosphere, a way of losing energy to make its orbit shorter and more circular. This technique, ‘aerobraking’, was new and somewhat risky. In the old days before faster-better-cheaper, changing orbits was something you did with engines, not drag. But drag is free and engines cost money.

In the end this aerobraking took a lot longer than anticipated: most of the atmospheric drag was felt by MGS’s solar panels and the arm holding one of these panels turned out to have a flaw in it. The aerobraking sequence was modified so that the spacecraft dipped into the atmosphere even less than had been planned, the force exerted on it ending up as less than three newtons – about the force it takes to lift a Big Mac. This slowed things down and it was not until early 1999 that MGS reached its final orbit, circling the planet every two hours or so, about 400 kilometres above the surface. The instruments now got down to business. The infrared spectrometer scanned the surface to see what minerals were present, and where. The camera, capable of picking out features just a couple of metres across, started adding long, thin tracks of extraordinary and frequently confusing new details to the coarser pictures of the Mariners and Vikings. And MOLA’s laser gently zapped the surface beneath the spacecraft ten times a second. The laser beam would illuminate a patch of Mars about 160 metres across and the altimeter’s clock would measure the time it took the light to get mere and bounce back (less than three thousandths of a second). The exact length of time revealed how high up the spacecraft was. Combine that altitude with tracking data showing where the spacecraft was – the tracking on MGS was exquisite – and you get a point in a global altimetry database. By the middle of April MOLA had produced almost 27 million such altitude measurements. For the most part they were precise to within less than a metre, which means that two nearby spots which seemed to have the same altitude would in reality be no more than a metre different in elevation. The overall accuracy with which the MOLA measurements determined the global shape of Mars was about eight metres.

A year after MGS reached its final orbit, in March 2000, planetary scientists from all over America and much of the rest of the world gathered in Houston for the Lunar and Planetary Science Conference, just as they have done every year since 1970, when the first such conference pored over studies of the first samples returned from the moon. For a week, the Johnson Space Center’s recreation building was turned over to them, and its basketball courts rang to the announcement of more and more news from Mars. At least a hundred papers on Mars were presented, most of them informed by MGS data in one way or another. To many of those attending, Mars seemed to be changing before their eyes. MGS measurements were discovering new features and forcing the reinterpretation of old ones. The idea that there had once been an ocean on Mars was starting to gain serious respectability. So was the idea that, far from having been geologically dead for billions of years, Mars was in fact still active. The old familiar face of the planet was taking on a youthful cast in the new light. The scientists were as reinvigorated as their planet.

But if the Houston meeting was full of scientific promise, there was also a fair share of institutional foreboding. Condolences were offered to the people who would have been presenting the first data from the Mars Surveyor programme’s 1999 missions, Mars Polar Lander or Mars Climate Orbiter, had it not been for their accidents. Some of these unfortunates – the ones who had worked on an instrument designed to analyse the way the Martian atmosphere changes with altitude – had watched their instrument burn up not once but twice, first on Mars Observer, then again on Mars Climate Orbiter. The reports of a whole slew of investigative committees on the previous year’s disasters were due out in the next few weeks and everyone knew that they would make sad, infuriating reading. While Mars Global Surveyor was a wonder, the programme it had spearheaded was a disaster.

On a phenomenally wet Tuesday evening, on a set of couches in the foyer of a building on the University of Houston’s Clear Lake campus, Carl Pilcher, the man responsible for solar system science at NASA headquarters, discussed the situation with various worried and disaffected scientists. He more or less confirmed that the next Mars lander, one that shared the design of Mars Polar Lander and was due to be sent off in 2001, was being cancelled. He accepted that the constraints that had been put on the programme had proved too tough – that in the effort to force JPL to make the Mars Surveyor programme faster-better-cheaper, mistakes had been made both at the lab and at NASA headquarters. He accepted that faster-better-cheaper had meant that the scientists had worked themselves to the bone and encouraged everyone there to help NASA get it right next time round. When he’d finished it was clear that the Surveyor programme as it had been talked about just a few months ago, with plans for missions in 2003 and 2005 that would not just study Mars in situ but send samples of its surface back to the earth, was over and that as yet there was nothing to replace it. With one exception – a small orbiter that would carry the last of the Mars Observer instruments to their objective in 2001 – the future of Mars exploration was, yet again, a blank.

But Mars itself was not. Just across the aisle from Pilcher’s attempt to share the pain of his bruised community was a special presentation by the MOLA team. As befits a geophysical instrument, MOLA is in the numbers game. If you put enough numbers together, though, you can get a pretty good picture. The MOLA team had taken their data-set, arranged it on a Mercator projection and printed it out as a map. The first version of this map, published in the journal Science the summer before, had been impressive. Garishly colourful, it had shown so much detail in its crater rims and mountain tops that many looking at it had assumed it was a colourful overlay superimposed on some sort of photomosaic or airbrushed map. But every last bit of the picture came from the MOLA data-set, from simple measurements of the time it took for a pulse of laser light to reach the surface of Mars and bounce back to MGS.

By the time of the Houston conference the map had been much improved. MGS had been in its proper orbit for more than an earth year (though only just half of a Mars year, each of which lasts 687 earth days, or 669.6 Mars days). More data had been added and very large printers had been used to blow the image up far beyond the scale of a scientific paper. The version in the University of Houston foyer was about two metres long and a metre and a half high. It would have been eye-catching even if you didn’t know what it was. If you did know, it was little short of a miracle. Here were real data, as hard and scientific as you could wish, woven into the image of a planet. It was not a realistic image. The altitude data were colour coded, so that the terrain ranged from blue in the lowlands through green to yellow to red to white. Hellas, the deep basin in the south, looked out like a baleful violet eye; the rise of Tharsis, its three great volcanoes snowy white, was ringed with burning red. Faint features were enhanced by computer filtering, just as they had been in the Mariner 9 photographs, to exaggerate details. Shaded relief had been added, not by skilled artists, but by a computer program first developed for charts of the ocean floor. It did a pretty impressive job – while still suggesting, as all such shading does, that the planet knew no night and that the sun was somewhere over the north pole. No, the map was not realistic. But to the people who walked by, and stopped, and stared, it was very real.

I watched for an hour or so as almost every scientist with any interest in Mars passing by on the way to or from the poster presentations elsewhere in the building stopped to stare at the MOLA map. They talked to each other; they pointed out features. They got close and squinted, then stepped back to take it all in. They enthused and gestured, and then fell silent and just stared. Peter Smith, designer and operator of Mars Pathfinder’s camera and, in his youth, a photographer with serious artistic ambitions, said it was the most incredible picture he’d ever seen. Baerbel Lucchitta, a striking, stately geologist who has been at the USGS in Flagstaff since the early 1970s, traced her favourite features with a little girl’s grin. When people finally walked away, their eyes and minds full, they couldn’t help but look back over their shoulders to get just one more glimpse. Here was a map that was most definitely being treated as an icon.

And David Smith just stood by his team’s creation and beamed. Other people on the MOLA team have told me that they always expected to put together such a picture of the planet, but Smith says he had had no idea the endless stream of data points would add up to such a striking visual statement. When I’d visited him in his office me year before, when the largest printed version of the map had been about thirty-five centimetres across, we’d looked at the data laid out numerically in vast spreadsheets. Though Smith had been keen to have the biggest possible version of the map printed for me Houston meeting, he’d not actually seen the resultant poster before that Tuesday evening. He was looking at it – and showing it off – for the first time, the joy of it all over his face. Across the aisle from the MOLA map, Carl Pilcher was explaining that an era of exploration that had seemed to be just beginning was coming to an end. But Smith just kept talking and smiling and looking with pride at his map. From time to time he’d touch it, running his hand lightly across the smooth blue of the planet’s northern lowlands. As though he could feel the onset of the higher plains to the south. As though the craters might scratch his fingertips.

300 million kilometres away, an instrument he had argued for and cajoled into being and thought about every day for more than a decade was illuminating the surface of an alien planet ten times every second. And in the rain-soaked Houston suburbs David Smith was stroking the face of Mars, a picture of delight.

(#ulink_5315f089-ad1d-5e09-ad8e-773da966a2e8) At a conference in Germany in 1990, a frustrated JPL engineer named Donna Shirley told a story about a recently deceased NASA engineer asked by St Peter what he’d achieved with his life, to which the answer was ‘First viewgraph, please …’ Shirley eventually led the Mars Pathfinder team.

(#ulink_3a3450f9-c57f-54e2-b68c-c3496fe07f7f) Although a shuttle launch costs a lot of money, those costs are not typically borne by any spacecraft along for the ride, but the cost of a one-off rocket is billed to the mission that it launches.

(#ulink_c8e090c4-72da-5fa9-879c-1ad1164e4bef) This sole-survivor-of-an-imagined-series motif is a common one in the history of NASA; as individual missions grow costly, their proposed successors are cancelled. The Planetary Explorer ‘programme’ of the 1970s ended up being a single mission. So did the Mariner Mark IIs conceived in the 1980s.

Part 2 – Histories (#ulink_8d3d65e6-4d44-53ae-8a60-d207f744a78d)

When the investigator, having under consideration a fact or group of facts whose origin or cause is unknown, seeks to discover their origin, his first step is to make a guess.

Grove Karl Gilbert, ‘The Origin of Hypotheses’

Meteor Crater (#ulink_5735e81b-9ccd-5dd3-9ce4-accfa78ccbfa)

‘Craters? Why didn’t we think of craters?’

Isaac Asimov to Frederik Pohl, on first seeing the

images of Mars from Mariner 4

If you care for impressive and beguiling landscapes, Flagstaff, Arizona has a lot to recommend it. The San Francisco peaks – remnants of a shattered volcano similar in scale to Mount St Helens – loom over a town scarcely a hundred years old, wrapped in the forests that attracted its founders. To the south the beautiful canyons of Sedona, carved into the rocks of the Colorado Plateau by water draining from beneath the forests; to the east the spectacular Painted Desert; to the north the Grand Canyon itself, more than a billion years deep, more gazed at and photographed than any other hole the world has to offer. Around the San Francisco peaks sit lower cinder cones like giant black molehills, weirdly fresh. Some are intact, some thoroughly quarried: their ash grits the roads in winter. One of them, remarkably, is in the process of being turned into a vast meditation on earth and sky, light and stone, by the artist James Turrell, earth movers his chisels.

If the land is wonderful, so is the sky, which seems to expand in sympathy with the majesty below. The air is dry, clean, a little thin – just the sort of place astronomers like to set up shop. Above the town, amid the ponderosa pines of Mars Hill, sits the telescope through which Percival Lowell imagined the landscapes of Mars. You can go up and have a look through it, if you like; at the right time of year you’ll be able to see Mars floating in the eyepiece just as he did, blotchy but beckoning. During my most recent visit it was the wrong time of year, with Mars best seen at about five in the morning, long after Lowell Observatory has closed itself to tourists. But even watched from a motel car park in the pre-dawn glow, Mars seemed closer in Flagstaff than it does in most places, shining clear and bright and true.

To appreciate land and sky together, drive about half an hour east of Flagstaff. A quarter of an hour beyond the line in the landscape where the Coconino forest responds to some subtle cue of altitude or precipitation and gives way to the Painted Desert, you’ll find what used to be called Coon Butte. From a distance it looks not unlike the flattened mesas that sit further off behind it, except for the fact that its heights are a little more crenellated. As you come closer, though, you begin to get the feeling that it is something quite different: smaller, lower, subtly different in form and nature. Rather than sitting on top of the desert like the low flat hills to the south, or puncturing it like the cinder cones behind you, Coon Butte seems to be a bending of the plateau itself, a twisting of the land towards the sky. And so it is.

Coon Butte is one of the places where the sciences of astronomy and geology meet. It marks the spot where, 50,000 years ago, a very small asteroid’s orbit round the sun was cut short by the surface of the earth. Most asteroids are made of stone friable enough that small ones will explode high above the earth’s surface, shattered by the shock of being slowed by the atmosphere. The fifty-metre asteroid that struck the Painted Desert that day was made of sterner stuff: iron. It pierced the atmosphere intact and ploughed on into the planet. Only after it had punched a hole through the surface of the desert did shock waves tear it apart in an underground explosion a thousand times more energetic than that of the Hiroshima bomb, throwing millions of tonnes of the plateau’s rocks back into the sky. The strata of rock surrounding the impact were bent upwards, raising the surface of the desert in a ring and forming a sharp upturned rim to the crater. Some boulders were thrown half the distance back to Flagstaff; within four kilometres the desert was covered with a thick blanket of debris. The hole left behind was about 200 metres deep and 1.2 kilometres across, excavated in seconds. The rim of raised rock stood sixty or seventy metres above the surrounding desert. After 50,000 years, erosion has smoothed it down to fifty metres.

Meteor Crater, as it is now called, is an impressive sight. By the time you reach the observation area on the north side of its rim – the only part to which the public normally has access – you have driven at least eight kilometres out of your way, you have paid for a ticket, you have walked past a gift shop and a well thought-out visitor’s centre; you know what to expect. Even so, to come across this sudden theatre of steep relief in an otherwise flat desert takes you aback. It is a big, dramatic hole, its base smoothed by the dried-out bed of a little lake, the strata of raw bedrock poking out of its sides like piers in an arena to seat a million.

At the same time, by the standards of truly dramatic valleys, canyons and volcanoes – standards the Arizona landscape requires all its tourist features to measure up to – Meteor Crater is not really so terribly large. Its sides are steep and deep, to be sure: if St Paul’s Cathedral were built at the bottom, the great golden cross on top of the dome would be well below your eye level. But the depths are enclosed in a way that almost belittles them. Craters are the most revealing of landscapes; from the rim you can quickly take in all there is to see. And this is not that large a crater. The rim is only four kilometres around. You could walk round it in a couple of hours (the walking is quite hard, for the rim is not regular); your eye runs round it automatically, limiting its scope in the process. Anything you can grasp this easily cannot give a sense of true enormity.

The most striking effect is not to look down from the rim into the crater’s depths, but rather to look straight across. To the south, the circle of the crater’s rim and that of the further horizon lie one upon the other, tangent arcs. Turn your head slowly – pan like a camera – and they become detached. The rim falls away from the true horizon; it twists into the middle distance, banking towards you as the true horizon keeps its distance, becoming a feature within the landscape rather than a limit at the edge of it. Eventually it ends up under your feet, a rampart of rubble dividing the bowl enclosed within from the great desert plain outside. And yet the rim still feels linked to the horizon itself. The great circle of the planet and the ring of the rim seem aspects of the same thing; the great void below echoes the great vault above. The effect has something in common with the old cliché of a straight road, a flat plain and a vanishing point on the horizon. But here there are no points and lines and directions: just circles turning in on themselves over 360°. This sense of a world arranged in nested circles may be something nothing else can offer as well as a deep astronomical impact with a well-preserved rim. And on the earth, there are no other impact craters with rims as well preserved as Meteor Crater’s.

On Mars, by way of contrast, there may be a quarter of a million impact craters the size of Meteor Crater. And there are craters of all other sizes, too. There are great impact basins large enough to put the European Union into; there are craters small enough to use for tennis courts. There are craters that overlap like the circles of an Olympic flag. There are craters on the rims of bigger craters. There are craters within craters within craters. Some are as young as Meteor Crater itself, some even younger. Some are more than 80,000 times older, landscapes more ancient than anything on the earth’s shifting surface except a few tiny zircon crystals preserved by chance.

And those are just the ones you can see in the airbrush maps and the Viking pictures; the ones with clear, well defined rims. One of the discoveries made with the data from Mars Global Surveyor’s MOLA altimeter was that there are hidden craters, too, craters yet more ancient than the visible ones, if only by a little. The MOLA team has developed all sorts of ways of using brightness and colour cues to bring out different aspects of their vast data-set. One of their best tricks is a way of looking at the planet slice by slice on a computer screen. The spectrum of colours that allows the eye to understand what it is seeing is concentrated into a thin range of altitudes – just a few hundred metres, perhaps – with all lower places darkly blue, all higher grimly purple; the highlighted range can be moved up or down at will. Look at a mountain this way and you will see a circular band of rainbow with a dark centre. Toggle the highlighted range upwards and the noose of light will tighten to a solid disc at the summit; lower it and the ring of colour will expand slowly until it smears itself out across the plains at the mountain’s base.

Run this magical palette over the surface of Mars and crater rims will stand proud as thin, hollow crowns. But rimless craters can be found too: solid circles of equal altitude. These are old, eroded craters, craters the unaided eye would never pick up. These shadow craters can be quite big: one of the first to be discovered this way was about 450 kilometres across, giving it an area about the same as Michigan’s or England’s, and definitely putting it in the first division of Martian craters. And they are quite numerous. In the summer of 1999, seventy flat circles of various sizes that looked like ancient impact scars were discovered by one high school student doing an internship with the MOLA geology team.

Discoveries on such a scale mark a peculiarly auspicious beginning to a scientific career. But if the intern was spectacularly successful in how she did her job, she was not particularly distinctive in the job she was doing. Almost every geologist who looks to the skies as well as the earth starts off counting craters; most will still be doing so, now and then, decades later. They are the way that astrogeologists measure time. On the earth, geological time is measured in layers; layering is history and depth is age, as a drive to any edge of the Colorado Plateau will demonstrate. Stratification, though, like embonpoint, is best seen in profile; on planets looked on only from above the study of strata is geometrically challenging, to put it mildly. But craters, too, are the testaments of time; like sediment on a sea floor, they accumulate over the years. Most planets with rocky surfaces are amply supplied with craters: the earth, endlessly reinventing its surface through erosion and plate tectonics, is the great exception. Reading the record of craters has made sense of the geology of the moon, has revealed global cataclysms responsible for remaking the surface of Venus and has provided, at least in outline, the history of the Martian surface from the most recent sharp-edged scar to the most ancient rimless basin.

It was through Meteor Crater mat people first learned how to read such records. It gave them what geologists most need: an analogue through which to understand processes not yet understood in any other way. Analogy sits at the heart of geology; it has long linked the past to the present, and now serves to tie the earthly to the alien. Meteor Crater allowed geologists to understand impacts, and its nested horizons became the door to other worlds and other times. Its role in understanding was not just theoretical. In the 1960s Meteor Crater was one of the sites chosen to train the only men from earth ever to walk anywhere else; strain your eyes and by the lake bed at the bottom you can see the statue of an Apollo astronaut that commemorates them. From his point of view there is no double horizon; beyond his little bowl of a world there is just the great urgent vault of the sky.