To the millions of people glued to their televisions last Fourth of July, the Pathfinder mission to Mars was a resounding success. It gave them expansive views of a rock-strewn, undulating terrain washed billions of years ago by a flood of Noachian proportions, dust-laden pink Martian sunrises, and images—sometimes comical—of the Sojourner rover ambling like an Erector toy from rock to rock. The probe even sent Martian weather reports: there was a deep chill in the summer air and the dust devils were out.
But the most novel achievement of Pathfinder was not the images, as entrancing as they were, nor the data that went with them—it was the engineering. Pathfinder was nasa’s first big demonstration of its cheaper, better, faster credo. Charged with doing what previous projects had done but at a fraction of the cost, nasa engineers were forced to rethink their notion of how to deliver a probe to the Martian surface. We had to throw out the book and start over, says Brian Muirhead, flight system manager for the project. With Pathfinder, we needed to reinvent the entire process of doing a planetary mission.
To save on weight and rocket fuel, Pathfinder engineers dispensed with the lavish retro-rockets that eased the Viking landers featherlike onto the red dust back in the 1970s. Instead they dreamed up an elaborate landing scheme straight out of a Wallace and Gromit cartoon. To start, they used friction with the Martian atmosphere to slow the incoming probe from 17,000 to 840 miles per hour, which is slow enough in the thin atmosphere to release a parachute. Then they waited until one second before impact to have Pathfinder fire its three small retro-rockets, which brought it to a dead stop ten stories above the surface. From there the probe cut itself free of its tether and simply dropped to the ground. Encased in a cocoon of air bags, it bounced 15 times before rolling to a stop.
Defying death in such an undignified way required an entirely different engineering process from previous missions. Viking engineers, designing and building what even for the time was a less radical spacecraft, did most of their testing at the end when it was fully assembled. Pathfinder engineers, in contrast, exhaustively tested and retested almost every component of the probe from the beginning—no sense building the whole thing and then discovering that your innovative air-bag scheme, say, wouldn’t be able to prevent total destruction on impact. Tests of individual parts led to redesigns of those parts, and sometimes to modifications of the probe as a whole. Testing until it worked every time, you got to a design that doesn’t not work, says Pathfinder engineer Rob Manning of the Jet Propulsion Laboratory. Testing became our life. Manning, for one, made more trips than he can count to the world’s largest vacuum chamber, at a nasa lab in Sandusky, Ohio. That was the most Marslike place to drop air bags onto volcanic rubble, and also to test the 53 explosive devices that did everything from ejecting the parachute to freeing the air bags.
Meanwhile Pathfinder’s scientists were laboring to make sure that the probe wouldn’t go all the way to Mars only to crash on a boulder. They wanted a landing site that was not too steep and had plenty of rocks for Sojourner, the two-foot-long rover, to look at—but small ones, not yard-wide boulders that could threaten the whole lander. The images from the Viking spacecraft didn’t help much, because they show details no smaller than an average suburban lot. A team led by jpl geologist Matthew Golombek, however, filled in details from other sources. By bouncing radar signals from Earth off Mars, they pieced together a map of how steep its slopes were. Infrared images from Viking provided a clue to how rocky the surface was; rocks take longer to heat up than sand does, so after sunrise a rocky surface gives off less heat radiation.
After analyzing dozens of potential sites, Golombek and his group chose Ares Vallis, a large floodplain in Mars’ northern tropics. It appeared reasonably free of large rocks, and it was scientifically interesting because it is thought to be the site of a great flood a billion or more years ago. That flood had swept north down from the highlands, and the Pathfinder team was counting on it to have deposited a smorgasbord of rocks—solidified lava, rubble from meteorites, magmatic rocks that cooled beneath the surface and were exposed by erosion.
So when geologists saw the first panoramic images of the Martian floodplain, which showed abundant signs of the ancient flood, they cheered. Rocks nicknamed Shark, Half Dome, Moe, and Pumpkin—the nerds were getting punchy—stood in a line pointing north, the direction in which floodwaters would have flowed. The ground itself had ripples four yards high, and in the distance, on a hill called Twin Peaks, the scientists saw what looked like huge boulders left by the flood. Once Sojourner began analyzing the chemical makeup of the rocks with its handy little alpha proton X-ray spectrometer, however, befuddlement set in. The first five rocks turned out to be much alike—and all unexpectedly high in silica. I’m still struggling with it,’’ says petrologist Harry McSween of the University of Tennessee, the Pathfinder team’s lead rock expert. It surprises me that the silica content is this high. It surprises me that all the rocks look alike.
There are several theories to explain the results. The most provocative holds that Mars once had drifting tectonic plates, as Earth does now; where plates collide on Earth, volcanoes spit out lava called andesite that is rich in silica. Other theories attribute the high silica content to more mundane geologic processes, such as lava rising near the surface and going through cycles of solidifying and remelting before finally erupting. And there’s also the bad luck theory: perhaps Sojourner just happened to pick some weird rocks.
Certainly the Pathfinder mission as a whole had some bad luck after its amazingly auspicious beginning. By October the lander, and thus the rover—whose signals are relayed to Earth by the lander—had fallen silent. The lander’s batteries seemed to have died, causing the radio and computer to become too cold to operate. But researchers think there’s an outside chance of restoring contact when the spring sun warms Ares Vallis, and Pathfinder’s radio, next August.
In the meantime, the Mars Global Surveyor arrived on September 12 and started aero-braking its way down to a shallow polar orbit, from which it will make a topographic map of the Martian surface—one that may reveal evidence of plate tectonics—as well as a chemical map of its composition. The next orbiter-lander pair is set for launch in December 1998 and January 1999. That orbiter will drop some novel little probes; shaped like spears, they will plunge into the Martian surface without any braking whatsoever, penetrating several feet in the hope of finding water. The lander, though, will descend onto the south polar region using conventional retro-rockets, as Viking did. When its design was completed, well before last Fourth of July, nobody even knew if the air-bag trick would work.