As lander, cruise stage and probes drift away from each other, perspectives alter. Mars stops being a vast wall in front of the spacecraft and becomes a strange new land below them; the ice-white limb of the planet barring the sky becomes a curved horizon. The outer reaches of the atmosphere begin to stroke the lander’s protective aeroshell, too thin at first to have much effect, but getting thicker by the second. Soon on-board accelerometers decide the breaking force is getting strong enough to be worth bothering about and tiny thrusters start firing to keep the aeroshell’s blunt nose cone pointed the right way. The atmosphere’s grip tightens further. Within a minute or so, the deceleration is up to 12g – the sort of force you’d feel if a cruising airliner came to a full halt in a couple of seconds. The nose cone is at 1650°C and the air around it is incandescent. The Polar Lander is a minute-long meteor in the Martian sky.
Three minutes after atmospheric entry begins, the worst is over, though the lander is still moving at 1500 kilometres an hour. A gun at the back of the aeroshell fires out a parachute and the thin air rips it open seven kilometres above the surface. Ten seconds later the charred front of the aeroshell is jettisoned and a camera pointing downwards starts to take pictures of the landscape below as it rushes upwards. If they make it back to earth, these descent images will make quite a movie.
While all this is happening I’m picking at a tuna sandwich in the JPL cafeteria. I chat to some of the scientists from other projects who are gathering round the television monitors that show what’s happening in mission control, then wander back across the plaza, past the model in the sandbox, to the press room. There’s no hurry – the probe is silent during the landing sequence and is only due to pipe up twenty-three minutes after touchdown. Even then there will be a fourteen-minute delay as the radio waves creep across the solar system at the speed of light. Plenty of time.
A quarter of a billion kilometres away, Mars Polar Lander’s legs snap out from their stowed position, ready for the ground below.
Four months later a board of enquiry decided that this was the crucial moment. When the legs snapped into position, they apparently did so with a touch more vigour than was necessary, flexing a little against the restraints meant to hold them in position. Little magnetic sensors in the spacecraft’s body seem almost certain to have interpreted this flexing as meaning that the legs had encountered resistance and were bending under the weight of the spacecraft – just as they would at the moment of touchdown. The state of these sensors was being monitored a hundred times a second by the part of the spacecraft’s software that was in charge of turning off the engines straight after landing and, since the legs took more than a hundredth of a second to reach their proper position, the sensors reported that the spacecraft seemed to have touched down on two successive checks. If it had heard this report only once, the software in charge of turning off the engines would have ignored the reading as a transient glitch. Hearing it twice, the software in charge of turning off the engines after touchdown concluded that the spacecraft had indeed touched down. Unfortunately, it was still almost four kilometres up in the air.
A bit more than a minute later, when the spacecraft’s radar said that it was only forty metres above the surface, the misinformed software had its virtual hand put on the virtual switch that controlled the engines. It turned them off straight away, unable to know or care that the spacecraft was still moving at almost fifty kilometres an hour. After falling that last forty metres, Mars Polar Lander hit the surface at something like eighty kilometres an hour, a speed it could never survive.
Back in December, no one knows any of this. About an hour after lunch on Friday, we know that the first transmission from the surface hasn’t happened, but though that’s a little disappointing, no one is really worried. Spacecraft are programmed to be flighty things and at the slightest sign of something out of the ordinary they are apt to go into ‘safe modes’, which means shutting down all non-vital systems for a set amount of time. The lander’s ability to go safe had been turned off during the descent sequence – when wilful inactivity would have been fatal – but once it got down to the ground this override would turn itself off and the spacecraft would be free to go into a silent funk if some subsystem or other had exceeded its safety levels during the landing.
Over the next few days the silence gets worse. Scott and Amundsen, the ground-penetrating microprobes, are never heard from at all. To this day no one knows what happened to them. The team running the polar lander itself methodically lists the things that could be stopping the probe from communicating and tries to work its way around them, using various different types of radio command. Is the main antenna facing the wrong way? Then send the lander instructions to scan its beam across the sky. Did it not hear those instructions? Send them over another frequency. Did it go into a different sort of safe mode, or go safe twice? Listen at the later times when it was meant to transmit. Each possibility is a branch on what the engineers call a fault tree, and every branch has to be checked out.
While all this is going on up at JPL, down at the Pasadena convention centre the Planetfest rolls on. The fact that there are no neat new pictures of the surface to be seen puts a damper on it, to be sure – but not too terrible a one. People still come to hear the assembled luminaries talk about the great future of Mars exploration. They hear from astronauts and scientists and engineers and Star Trek actors and Bill Nye me Science Guy, proselytiser by appointment to PBS. And they hear from the science fiction writers. From Larry Niven, who has just written a fantasy in which all humanity’s dreams about Mars come true at the same time; from Greg Bear, whose Moving Mars imagined the planet’s future as a backwater from which settlers watch the ever more high-tech earth redefine what is human; from Greg Benford, whose The Martian Race, published this very weekend, sets a new standard of technical accuracy for first-mission-to-Mars stories. And from Kim Stanley Robinson, whose books Red Mars, Green Mars, and Blue Mars provide the fullest picture yet attempted of life on that planet. Unlike every previous generation of science fiction writers, these men have had data from Mars orbit and the Martian surface on which to base their visions, and they are scrupulous in their use. In their hands, the physical facts of planetary science and the romance of travel to other worlds are brought as close as they yet can be.
Meanwhile, up at JPL, what seemed so close is slipping away. After each new attempt to make contact an ever more despondent flight team comes out to face an ever smaller press corps and tell us that nothing was heard. They were so excited on Friday morning – by the early hours of Sunday, some are almost in tears. On Monday morning most have had a chance to rest, but though the faces are fresher and the eyes clearer, a certain resignation has settled in. By Monday night, all the one-fault branches on the fault tree have been evaluated; it’s clear that at least two separate systems must have failed. The team will keep climbing ever more unlikely limbs of the fault tree for a week or so yet, but for the rest of us that’s it. The lander is lost. The last tents in the media caravan are folded up just after midnight; we don’t even have the ingenuity, or stamina, to find a bar.
‘I think it’s part of the nature of man to start with romance and build to a reality.’
Ray Bradbury, in Mars and the Mind of Man
Mars Polar Lander was JPL’s thirteenth mission to Mars and its fifth failure. Mariner 3 died with its solar panels pinned to its side by the wrapping in which it had been launched in 1964; Mariner 8 fell into the Atlantic in 1971; Mars Observer exploded as it was trying to go into orbit round Mars in 1993; Mars Climate Orbiter burned up in the atmosphere in 1999; Mars Polar Lander made its mistake just forty metres up a few months later. An optimist might point out that each got closer to the target than the previous failure. A pessimist might point out that the frequency of failure seems to be on the increase.
It’s hardly surprising that, with so few missions, everything that has not been a failure has been counted a terrific success. Mars exploration is still too new for there to have been any hey-ho, business-as-usual missions. But among all these successes one stands out: Mariner 9. Mariner 9 was the first American spacecraft to go into orbit round