September Song
[Requiem for a Space Probe]

From the point of view of the vintage Galileo Orbiter, which has been tracing a complicated, cat’s-cradle trajectory among the four largest moons of Jupiter for eight years now, September 21st is bound to be an interesting day. If the camera of the spindly craft hadn’t been left shuttered after two hazardous close passes of the explosively volcanic moon Io in late 2001 and early 2002, leaving the robot essentially blind to visible light for the last two years, Galileo would witness an ominous, impressive sight. The immense banded bulk of Jupiter, by far the Solar System’s largest planet, will balloon inexorably in front of the probe as it hurtles closer than it ever has before to the Jovian atmosphere. To the east, the “fire and ice” pairing of Io and Europa, two of the most fascinating objects in the Solar System, will traverse the rapidly spinning storm systems at Jupiter’s day-night terminator in majestic sun-lit tandem. On the other side of the planet, Ganymede and Callisto, the outermost and larger of the “Galilean Moons” – objects discovered by the spacecraft’s namesake, a certain Pisan astronomer – will gradually near the Jovian horizon and then set as the Orbiter’s speed takes it closer and closer to their huge parent.

Within Galileo’s long rotating boom assembly, its five Fields and Particles science instruments will all be powered up; they’ve been conducting a live, real-time transmission to Earth since February 28th. Only two parts will be moving on the entire spacecraft: the upper, or “spun” part of Galileo, which actually constitutes the majority of the Orbiter, and which has been turning at about three revolutions per minute for almost the full duration of its fourteen years in space, and a tiny internal scanner motor in the craft’s Energetic Particle Detector, which constantly shifts the position of the detector within the instrument to allow it to survey the entire sky.

Six million miles away in a direction indistinguishable from that of the Sun, two links in the global chain of 70 meter wide deep-dish antennas that keep an Earthly ear tilted towards such endeavors will track the signals from Galileo’s science instruments and also its engineering data, including the temperatures, pressures, and voltages of the various spacecraft systems. As the probe crosses the orbit of the boulder-like asteroidal moon Amalthea, the DSS-63 antenna outside of Madrid may pulse a command, ordering Galileo to collect and read-out data from its star tracker – a device used to help orient the craft. This information will be used to search for occultations of guide-stars, potential evidence of a suspected rocky ring close to the planet. As Madrid turntables eastwards, rotating over the horizon of distant Earth, thin bars of ruby within the cryogenically cooled receiver housed in multiple tanks of liquid nitrogen at the focal point of one of NASA’s oldest tracking stations, DSS-14 at Goldstone, California, will vibrate very slightly as they pick up Galileo’s faint beacon, amplifying the steady stream of information piping from the craft’s science instruments as they convey information about the intensely radioactive inner magnetosphere of Jupiter, as well as the wispy, smoke-like “gossamer rings” weightlessly suspended inside the orbit of tiny Amalthea. Traveling at the speed of light, the spacecraft’s radio signal will take 52 minutes to span the gulf between Jupiter and Earth.

Only two human artifacts have ever penetrated this close to the planet before: Galileo’s own auxiliary atmospheric probe, which plowed into Jupiter on arrival in December of 1995, and tiny Pioneer 11, the second outer Solar System mission, which conducted an extremely close fly-by in December of 1974 on its way to Saturn.

At 12:42 p.m. PDT Galileo will whip behind Jupiter as seen from Earth, and all telemetry will abruptly cut off, eclipsed by night-shrouded clouds. Seven minutes later and while traveling at a speed of 48 kilometers per second relative to the Jovian atmosphere, Galileo’s boxy, instrument-festooned octagonal frame and the leading edges of its three extended, spinning booms will start to glow red. Within 30 seconds they’ll have snapped off, but will continue tumbling behind the now white-hot main body of the hurtling craft. By the time the Galileo Orbiter reaches the one-bar level of the atmosphere, many of the 85,000 component parts of perhaps the most complex interplanetary spacecraft ever build will have separated from each other as well, though they’ll continue burning into the atmosphere in a hail of disintegrating shrapnel. When its remains reach a depth of about 500 kilometers below Jupiter’s hydrogen cloud-tops, a region where the temperature soars to 660 Celsius, all of the Orbiter’s aluminum components will have melted, then vaporized. At a depth of 1,000 kilometers, its titanium parts will finally also melt, then turn into vapor. Rendered into a soupy haze of dispersing metal atoms, Galileo will have become an undifferentiated part of Jupiter, with no clues left as to its Earthly origins, its mission, or its original ungainly, inquisitive, insectoid shape.

Considering that Jupiter is 71,500 kilometers in radius – big enough to contain all the other planets and moons of the Solar System with plenty of room to spare – Galileo will barely have penetrated its outermost atmospheric layer.

Thus the most complex, tortured, but ultimately redeemed mission in the annals of deep space exploration ends with a commanded suicide.

* * *

Actually, it’s something of a miracle that the probe ever got to Jupiter in the first place. Galileo’s ultimate trajectory can be traced backwards, through a loopy series of linked, ever-shifting elliptical orbits around Jupiter and among its satellites, and then further back still, through the outer and then inner Solar System, in a mesmerizingly cyclical then circuitous odyssey filled with averted disasters and big triumphs. Leading the latter is a single discovery: that Jupiter's opaque, cue-ball smooth moon Europa almost certainly has an immense liquid water ocean under its frozen surface crust – a realization which in turn has led to a vociferous scientific debate about the prospects for life there.

Although Jupiter’s glassine moon certainly wasn't the only target of Galileo's cameras during its eclectic fourteen years in space, the implications of the probe's pictures of this weirdly fissured sphere – many of which show icebergs that have apparently rafted into new positions before being re-frozen into Europa’s granite-hard, but apparently quite thin, global ice-cap – produced a collective euphoria in the planetary sciences community in the late 1990's. As Richard Terrile of the Jet Propulsion Laboratory put it, “How often is an ocean discovered? The last one was the Pacific, by Balboa, and that was five hundred years ago.”

In the end, Galileo’s record reads like a litany of firsts. It conducted more fly-bys of more planets and moons than any other probe in the history of the genre (forty, including repeated encounters). It was the first to swing close by an asteroid; the first to discover a tiny moon orbiting an asteroid, on its second foray into the asteroid belt; the first to examine the third stone from the Sun, a.k.a. Earth, on a classic fly-by trajectory; the first to discover life on Earth (in an intriguing test experiment devised by Carl Sagan); the first to orbit one of the immense outer planets; the first directly to observe fire fountains erupting from the surface of Jupiter’s moon Io; and the first actually to fly through a plume from Io, a lurid yellow-orange firecracker with an estimated 200-300 volcanoes erupting across its pitted face at any one time (a finding itself attributable to Galileo). In July of 1994, while still outbound, Galileo provided long-distance direct observations as fragments of Comet Shoemaker-Levy 9 slammed into Jupiter at radical speeds, producing multiple fireballs that rose over the planet’s limb – a series of frightening detonations more powerful than the largest H-bombs and entirely invisible from Earth. Another first, one of the first.

During the mission’s last years, when it was run by a skeleton crew on a low budget and had taken more than four times as much of the fierce Jovian radiation as it had been designed to withstand, Galileo’s systems faltered frequently, but it continued to come up with discoveries. In November of last year, for example, its star trackers registered the presence of up to nine tiny moons orbiting close to Jupiter. In 2000, its on-board magnetometer came up with the strongest evidence yet that a liquid water ocean exists – right now; contemporaneously – under Europa’s ice. Galileo even made a discovery outside the Solar System. In June of 2000, its star tracker suddenly failed to recognize the bright star cluster Delta Velorium, which flares brightly (and, as it appears, unsteadily) in the southern-hemisphere constellation Vela. Subsequent observations from Earth confirmed that this grouping of stars – which have a greater magnitude than the North Star, and which has been catalogued and observed since ancient times – in fact contain a dual-sun system, with one of its component parts periodically eclipsing the other, resulting in the variable light output that puzzled the spacecraft’s instrument.

Galileo thus became the first interplanetary probe ever to make an interstellar discovery.

* * *

Given this litany of accomplishments, which outstrips that of any living human astronomer, it may come as a surprise that this was also the single most troubled mission in the history of robotic space exploration. Its problems started well before launch, and they didn’t end there. I’ve been keeping an eye on Galileo for years, mostly out of sheer fascination with its pictures of Europa and Io, many of which exceed the most improbable visions of 20th century science fiction, and during that time whenever I got a chance to meet some of Galileo’s Earth-bound handlers, I took it. These people have what amounts to the most fascinating desk job of all time: they get to explore strange new worlds, and boldly go where no one has gone before, all in symbiosis with a two-ton spacecraft that’s six million miles away.

In the spring of 2000 I met Bill O’Neil, a square-jawed, silver-haired engineer who served as Galileo mission director for the bulk of its main mission to and around Jupiter. O’Neil described the turbulent ups and downs of the probe’s ongoing journey with the ongoing muted euphoria – a low-key exultation really – of someone who’d actually done it: he’d brought it off, beating the odds that had so mysteriously stacked up against the mission.

Although O’Neil has played a major role in almost every American robotic space mission since the first Surveyor touched down flawlessly on the surface of the Moon in 1966, Galileo dominated his career. He worked on the project for eighteen years—originally as the manager of its science and mission design, then as project manager for the entirety of its main mission to and around Jupiter. O’Neil described Galileo’s sorry state as it endured a series of launch delays around the time the Challenger shuttle exploded in 1986. Some were due to that tragedy, some not. Galileo’s entire flight plan had to be redesigned an unprecedented five times as the rocket power available to take it out of Earth orbit diminished (for various reasons) and the configurations of the planets shifted (for reasons easily explainable with reference to Newton). The probe was repeatedly trucked back and forth between California and Florida; it was disassembled, cleaned, stored, and then re-assembled; hair turned gray as the mission went from its late 70’s design to its actual 1989 launch to its extremely attenuated Jupiter trajectory and then finally on to a series of revelatory encounters with the planet’s eye-opening satellites.

One of his stories in particular stuck in my mind. It involved an obscure JPL trajectory specialist named Roger Diehl, who had single-handedly saved the mission, which had been in serious trouble even before it left the ground. Galileo’s most critical pre-launch problem was that it was saddled with a woefully underpowered solid-fuel upper stage booster that could barely get it out of Earth orbit. This circumstance had come about because after the Challenger disaster, a newly (and as it now appears, temporarily) safety-conscious NASA decided that Galileo’s more powerful Centaur upper stage – which was liquid-fuelled and therefore more dangerous than its weaker solid-fuelled alternate – couldn’t be lofted along with the Shuttle’s human cargo. And Galileo had been designed for Shuttle deployment.

The result was that, after nearly a decade of development, testing and assembly, the spacecraft was on the verge of a one-way trip to the Smithsonian. JPL responded by deploying its best talent and kicking into high gear. Soon after the Challenger explosion, mission design manager Bob Mitchell assembled a cutting-edge team of trajectory specialists, including Roger Diehl, Lou D’Amario, and Denis Burns. Their mandate was to figure out how get what amounted to a Mercedes Benz of a space-probe to Jupiter with a lawn mower engine under the hood.

It was Diehl who ultimately came up with the exquisitely unlikely trajectory actually capable of getting the mission to its destination. Late in July, with the doomed robot riding its last orbit in towards Jupiter, I called him at JPL. “Yeah, Galileo,” he recalled ruefully. “You know, all these deep space missions have their problems, but Galileo really seemed to have more than its fair share. We used to have a party every year, a ‘five years to launch’ party, and then a ‘three years to launch’ party, and what was funny is that each year the numbers of years would jump around – you know, forwards, and then backwards. Finally at one of the later parties we had a plot up that showed exactly where we thought we had been at each party over the years.” A disembodied laugh tumbled through static.

I asked him what had led to his mission-saving epiphany, and Diehl explained that the existing upper stage only had enough juice to get Galileo to Mars or Venus – intriguing Solar System objects to be sure, but not ones it had been designed to study. So his first efforts were directed at getting the spacecraft to Mars and then using that planet’s gravity to sling it onwards to Jupiter. It turned out, however, that as Galileo’s launch kept on getting postponed, Mars had moved inexorably onwards from its ideal placement in between Earth and Jupiter. Getting a boost from the Red Planet was less and less of a usable proposition.

“So then the next thing that I recall was saying ‘Well, let’s launch to Venus,’” Diehl said. I pointed out that, apart from that planet being in an entirely counter-intuitive direction – Venus is in the inner Solar System; Jupiter’s a very great distance in the opposite direction – Bill O’Neil had told me that Galileo had never been designed to go inwards and closer to the Sun before going outwards to the frigid space around Jupiter. Hadn’t that presented a thermal problem? Yeah, Diehl responded, but the mission was on the verge of cancellation anyway, and his boss Bob Mitchell had said consider anything. Even though the spacecraft’s designers might’ve been appalled at the idea of heading towards the Sun rather than away, given a workable trajectory, and perhaps with some kind of heat shielding, that could sort itself out later.

“I remember I read that when a person gets pumped up to really work a problem, the adrenaline flows,” he said. “Its almost like you’re working at a level that you haven’t worked at before. And I sort of experienced something like that.” At this point even the tone of Diehl’s voice had subtly changed – gradually shifting to a higher pitch as his words came faster. “I would go to bed at night, and my wife said she could even hear me talking about trajectories in my sleep. And then I’d wake up in the morning and I would have something that I would immediately want to try.

“And so the night before I found the first trajectory which identified the concept, I remember thinking ‘I’m going to totally ignore Mars.’ In my previous trajectory work I had done a lot of the initial tour designs for how you fly by the different satellites of Jupiter. So I said to myself ‘I’m going to think of the problem as doing a tour of the planets of the Solar System with the goal of getting to Jupiter.’ And I didn’t care how many years it would take to do it.”

There was a pause at the other end of the line. “You know, in the back of my mind I felt that there was something out there, and why haven’t I found it, I know it’s there,” Deihkl finally resumed. “And throwing Mars out of the equation was like opening the flood-gates to being able to find it. And the next morning I went in, and within fifteen minutes I found it.”

His trajectory required that Galileo return from Venus and swing past the Earth again not once but twice on its way to Jupiter. After putting up some initial resistance, the spacecraft’s designers soon came up with light-weight thermal shielding to protect the spacecraft from the harsh Venusian sun, and meanwhile Diehl’s colleague Lou D’Amario went on to take the initial trajectory concept, quickly dubbed VEEGA (“Venus Earth Earth Gravity Assist”), and spent months improving it substantially, building in more favorable Earth departure and Jupiter arrival times and incorporating two asteroid fly-bys for good measure. Flight time from Earth to Jupiter had gone from three to six years, but it seemed a small price to pay – particularly given a resuscitated mission and some worthy science objectives along the way.

I asked Diehl if he thought he had received the recognition he deserved. “Well, the fact that Galileo ultimately flew a trajectory that I came up with, that was the ultimate high for me,” he answered slowly. “I mean, nothing could replace that. But yeah, I was well known and at the same time within the team I think it caused a little bit of… well, a little bit of hard feelings. You know, everyone was working equally hard, and I was getting a lot of recognition. And at JPL you have a lot of people who are very high achievers. So you know, in one way I felt very good but at the same time there were these hard feelings. So it was a little bit tempered by that.”

There was another ruminative pause. “But to this day I’m still very proud, and my car license plate says ‘VEEGA,’” Diehl concluded. “So every morning I go out and I see the word.” A pleased chuckle floated across the line from the air-conditioned blaze of noontime Pasadena.

* * *

Diehl’s brainstorm allowed Galileo to be taken out of storage and reassembled in the womb-like clean rooms of JPL, but the game wasn’t over, not by a long shot – and we’re talking here about one of the longest shots ever conceived and executed by the species. By the time Galileo was actually deployed from the reactivated Space Shuttle in 1989, seven years after its original launch date, its on-board processors were six generations behind what anyone with a couple thousand bucks could buy at the local neighborhood PC store. (They were in fact radiation-hardened, rebuilt versions of the RCA 1802 chip – the same processor which ran the most primitive early video arcade game, “Pong.”) And by the time Galileo’s underpowered upper stage fired, sending the craft puttering towards Venus to pick up momentum, one of its most important component parts, which was originally manufactured in Florida, had already vibrated its way across the full span of the continental United States in the back of a truck four times.

Throughout the entire first part of the trip, the spacecraft’s umbrella-shaped high gain antenna – intended to be its main communications link to the Earth from way out at Jupiter – had been snugly folded, exactly as an umbrella generally is in the absence of rain, i.e., along a central spine projecting from its center. The importance of this particular device can’t be overestimated: Jupiter’s so far from Earth that if you tried to drive your Ford Galaxie there, it would take about 79 million years, floored. The plan was to deploy Galileo’s vital main antenna only in late 1991, after the probe had receded far enough from the Sun, nineteen months after launch.

But first came “Earth-1” – the first Earth fly-by, which happened to coincide with the first Gulf War, to the point where JPL had to inform NORAD (the North American Air Defense Command) that the hurtling blip that would appear on their radar screens on December 8th – an incredibly fast-moving object that might well appear to originate from the Middle East, not Venus, and be on a ballistic trajectory towards the continental United States, not the outer Solar System – was actually a NASA-origin space probe, and not one of those famous disappearing Iraqi “weapons of mass destruction.”

“I’ll tell you, I can’t really explain this,” O’Neil had said to me in Paris, “but Earth-1 was the most euphoric professional event of my life. Now why wouldn’t it have been arriving at Jupiter, when everything worked perfectly?” Yeah, I wondered, why not? We sat there together, baffled, until finally I ventured a longish explanation: they were sitting there at JPL on Earth, had finally managed to get their machine out into the wine-dark etcetera, and there it was already coming back to them, but only temporarily, and showing them themselves, in fact all of us here on this marvelous blue-white sphere; of course you were euphoric…

O’Neil cut me off in full improv. No, he said, the still-furled high gain antenna meant there weren’t any “real-time, gee-whiz images.” Instead Galileo had been forced to rely on its small, stumpy, omnidirectional low gain antenna for the entire first part of the cruise, a device which had been put onboard only for near-Earth and emergency communications purposes. Such a low-wattage antenna didn’t allow for the rapid delivery of bandwidth-hogging pictures. No, Bill O’Neil’s “Earth-1” euphoria had most likely been because “everything seemed perfect. We had been through all this, and in particular, we had been to Venus, we had demonstrated that we had overcome the challenge of taking a spacecraft that was never intended to go there, there and back successfully – and the answer is, probably, that it looked like all the challenges were handled. That we had survived.”

* * *

His high lasted only another five months. With Galileo now heading out towards the cooler climes of the Asteroid Belt, and in fact towards the very first encounter between a robot and an asteroid, the time had come for it to open its high gain antenna and presumably begin pulsing the anticipated luxurious flood of real-time, 134 kilobytes per second data towards Earth. Galileo’s data rate was designed to give it enough bandwidth to fire one picture home from Jupiter per minute, while also feeding information from all its other science instruments simultaneously. Its antenna, in fact, was the largest ever to have been sent out of Earth orbit – so large that it had to be folded even in the cavernous confines of the Shuttle bay.

Perhaps not surprisingly, though certainly tragically, when JPL finally ordered Galileo to open this key device – it stuck. And subsequently refused to budge, no matter what they tried to do. Later analysis determined that a design flaw combined with the vibrations during all the trucking across the country were largely responsible. It was Galileo’s second potentially mission-terminating catastrophe, and it was also a devastating blow to everyone’s morale.

JPL responded by deploying its best talent and kicking into high gear. This was not the myopic Hubble Space Telescope, which could be affixed with what amounted to a monocle by visiting astronauts; Galileo was already nearing the orbit of Mars. Within a week of the antenna failure, two engineering teams were formed. One was dedicated to figuring out how to get the thing unstuck (essentially through alternately heating, then freezing, than heating the antenna by rolling the spacecraft’s aft-end towards and away from the Sun, then ordering the small antenna deployment motor to pulse; then trying multiple variations of that over a period of months while cursing through clenched teeth).

The other team was dedicated to figuring out how to make the mission work without the use of the high gain antenna. It was comprised primarily of telecommunications specialists from JPL’s Deep Space Network, and was assembled under the leadership of the head of research and development for the network, Leslie Deutsch.

The Deep Space Network has been a central node of the Information Age since well before it was called that – since the short-lived Space Age, in fact. People who’ve developed systems and procedures for the DSN have gone on to define (and sometimes profit mightily from) the protocols and standards and frequency modulations and complex decoders and other gizmos that govern that alternative universe of routers, chips, transmitters and receivers that are the backbone of the Age. And the Deep Space Network has also frequently served as what one of Deutsch’s predecessors called “a million mile screwdriver” – the only way for JPL to fix distant problematic robots, of which Galileo was not the first nor the last, though certainly the most challenging.

One day after calling Roger Diehl, and with distant Galileo already almost three million miles closer to its fatal rendezvous with Jupiter, I picked up the phone again and dialed Leslie Deutsch. I asked him what his first order of business was when the sickening realization dawned that JPL’s billion dollar flagship, now finally on its way after multiple delays, was — metaphorically at least — dead in the water. “There was a crises,” he acknowledged. “I got together with a few people, and we did some brain-storming. And we said suppose the high gain antenna never opens. Suppose it never gets any better than this? What do you think we can do? And we were doing that within a week of this event. We worked pretty fast on it. And we did some thinking.

“And first we said, suppose we don’t change anything, what’s the data rate going to be when we get to Jupiter? If we just have to continue using this low gain antenna? And if we had, we would have been a factor of ten thousand lower in data rate than we had planned for the mission. Instead of being a hundred thousand bits per second we would have been at ten bits per second.” From one picture a minute they had gone to one picture per month.

As Deutsch sketched out the disastrous scenario, I realized that in effect, a kind of transmigration of souls was necessary. The physical part of the spacecraft was out of reach – or to the extent that it was in reach, the other team, the long-distance antenna repairmen, were mandated to explore that side of the problem. With the physical spacecraft largely unchangeable, the extended hand of the DSN rescue squad could only have its effect on Galileo’s information-processing side. It could control, in other words, the bits that effect the spacecraft’s atoms; the software. This was the ghost in a machine named after a ghost, and it was changeable, not through séances or crystal balls, but through electromagnetic impulses pulsed through huge antennas.

But first the on-board computer had to be up to snuff, and ready to receive the spirit that moves the body. “The bad news was that these computer processors were ancient,” Deutsch said. Because computer chips have to be entirely re-engineered and re-built to withstand the vibrations of launch and the harsh radiation levels of deep space, most deep space missions are launched with computing power three generations older than that available in the current generation of PCs. And in the case of Galileo, Deutsch said, that lag had been compounded by all the delays since its conception.

But the good news was that shortly before launch, the spacecraft had been outfitted with twice as many memory chips as originally intended. This would be the boon allowing the Deutsch team to do something that had never been attempted before: to change the spacecraft’s software – in effect, its entire operating system – from the ground, painstakingly, in mid-flight, using the sluggish low-gain antenna. And changing the software would enable them to introduce advanced data-compression techniques into the spacecraft, which in turn would help to make it possible for Galileo to send useful pictures and other science information over the low gain antenna from as far away as Jupiter. Not nearly as fast as originally planned, but still at a rate many times faster than the dismal original estimate of 10 bits per second. The spacecraft would now be able to send several hundred images to Earth per month. Although Galileo’s instruments would now have to be pointed with the greatest of care, and although some of its science objectives would now have to be thrown out altogther, the mission, it was beginning to seem, wasn’t a complete write-off.

And there were other important elements to Deutsch’s strategy. Millions of dollars would have to be invested in adding to, and electronically ganging together, the network’s globe-girdling chain of dish antennas – work which would immediately benefit all other space missions. And the antennas themselves would also be improved, with the cryogenically cooled ruby receiver prongs at their hearts cooled down still further – in practice by placing a freezer tank within a freezer tank within a freezer tank, matrioshka doll style – to help differentiate between the spurious noise of their surrounding electronics and the distant zeros-and-ones whisper emanating from Galileo’s impossibly (or rather, it was beginning to seem, just possibly) weak and distant low gain antenna.

Perhaps inevitably, a type of low intensity warfare developed almost immediately between the telecommunications specialists that were endeavoring to figure out ways to continue the mission without the use of the high gain antenna and the engineers who were trying to open the damn thing from the ground. “The very fact that we were sanctioned bothered them, because it was like people saying they were going to fail,” said Deutsch. “And yet on our team we were always saying ‘You know, we’re doing all this great theoretical work but we’d really love never to have to put it into practice!’” He laughed. “The conflict arose when we got to the point where we had to say ‘Look, in the next six months we’ve got to make a decision on this or there’s no time to do it the other way.’ And so eventually we had to make that decision. And they were basically told, ‘Ok, now you are the second class citizens, and you have to fight for time on the spacecraft to try to do what you want to do.’”

From there on out, occasional attempts were still made to open the antenna, but they all failed. As for the intramural conflict, it mostly took the form of glares between rival encampments across the JPL cafeteria, but never escalated to the food-fight stage.

* * *

I had heard that some of the cerebrations built into the software solutions to Galileo’s high gain antenna problem had made their way into daily use in the wider world, and was having a hard time pinning down just which went where. So I asked Deutsch. It turns out that two innovations in particular could be traced directly to Galileo, with the first linked to the crises itself and the second pre-dating the antenna problem.

When Galileo was launched, Deutsch said, it had a coding technique built into its circuitry designed to reduce errors in the spacecraft’s signals. “It’s called a Reed-Solomon code, and was originally developed as a mathematical curiosity,” he said. “Both Reed and Solomon were consultants here. That particular code is what enabled the CD industry, compact discs. It’s two levels of that code that enable you to resist dust and scratches on a CD. And that’s a multi-billion dollar a year industry.”

The CD industry is already in decline even before Galileo’s terminal dive – but its decline has everything to do with the second great innovation which saved the mission. CD sales happen to be nose-diving because of the relative ease of trading music for free on-line – and the other concept devised by Deutsch’s team – the one necessitated directly by the antenna malfunction – was the development of a so-called “packet-based” information downlink. It was soon to be the technique enabling efficient information transfers on the Internet.

Previous to Galileo, and in fact initially on Galileo, a spacecraft’s instruments fed an ongoing live stream of telemetry to the ground – essentially an EKG reading reporting on spacecraft health and accompanied by the science results of various instruments. But there was a lot of wasted space in that feed, the Deutsch group realized – “empty” bits where a particular instrument was switched off, or simply not vital at the moment, leaving holes or useless data in a signal that was nevertheless transmitted to the ground. Galileo could no longer afford that luxury, and so by the time it got to Jupiter, the spacecraft’s instruments were putting their information in packets, storing it, compressing it, and sending it to Earth whenever time was available. Although each packet now had to have a header specifying when it had been recorded and from which instrument, the added address bits were far fewer than the subtracted empty or irrelevant ones.

The decades-long trajectory of this mission, in other words, arches over the CD revolution, then the net-based MP-3 counter-revolution, playing an integral part in both! At this stage I found myself looking down at the tiny wheels spinning in the tape recorder that was absorbing the distant words of Leslie Deutsch and considering how impossibly retro, how positively archaic, a 20th century Walkman, a device entirely without silver discs or solid-state storage capacities, actually is in 2003 – when suddenly I remembered the other key element that saved Galileo: its on-board tape recorder.

* * *

Yes, Galileo has a tape recorder, much bigger than a Walkman, a defiantly pre-digital box of mechanized wheels-and-springs workmanship manufactured by the Odetex Corporation of California, and practically indistinguishable from other, less radiation-shielded reel-to-reel tape recorders attached to the higher-end stereo systems of the sixties and seventies. If the jammed communications dish was Galileo’s Achilles heel, its boxy archaic tape recorder became its greatest single redeeming feature, much to the surprise of everyone concerned. None of the wizardry of the Deutsch Telecom team, revolutionary though it was, would have amounted to a pile of Greek war helmets without that vintage recorder – and yet the bulky device was capable of making a “Pong”-era microchip seem about as advanced as a solar-powered cell-phone.

The tape recorder had been incorporated into Galileo’s design for one reason, and one only: to back up data from the craft’s auxiliary atmospheric probe, which was scheduled to tunnel into Jupiter’s clouds upon 1995 arrival, release its heat shield, deploy a parachute, and go about the business of uplinking information about the Jovian atmosphere as it sank into atomized oblivion. The whole procedure was supposed to unfold over the course of an hour, and was originally to happen in real time, with a live link between the atmospheric probe’s seven instruments, Galileo, and Earth. But if there were to be a technical problem – if, for example, Galileo’s high gain antenna had failed to deploy, or if the receiving station on Earth had just been struck by thirteen bolts of lightening – then that invaluable feed (another Galileo first: the first direct contact between an outer planet and Earth-origin instruments) would have been lost.

The solution was to back the probe data up on that tape recorder, and the only reason why two such recorders hadn’t been sent within the spacecraft’s bulky carapace – standard operating procedure for a mission-critical component – was that it was considered redundant to the Orbiter’s primary task. But Galileo’s tape recorder became mission-critical at the moment its JPL handlers, in evaluating their situation sans main antenna, realized that it could be used – in fact now had to be used – to store all the incoming images and other scientific data gathered by the spacecraft’s instruments during its multiple fly-bys of Jupiter’s moons.

The complex kineticism of the Jovian archipelago meant that on arrival, Galileo would necessarily have to go into a series of elongated, months-long elliptical orbits between such encounters. These delays, once thought of as necessary evils, had now become much-needed windows during which the stored fly-by pictures and other data could be fed from the recorder to Galileo’s computers (for compression by its new software) before being slowly trickled to Earth (over, you guessed it, the low gain antenna).

So the filament of magnetic-tape spooled in Galileo’s tape recorder became one of the thin threads on which the mission’s destiny hung, and took its place within the blinking gizmo-chain that was assembled to save the mission, and by the time of the probe’s first encounter with a Jovian moon, its entire operating system had indeed been replaced. It was an unprecedentedly risky move, “a complete brain transplant over a 400 million mile radio link,” as one team paper put it, and any error could have meant losing the spacecraft altogether. But it was utterly necessary. Given that it extended out across half of the Solar System, and consisted of a dialogue between human controllers and an intricate, sophisticated, flawed machine, the whole assemblage was a kind of fantastical Rube Goldberg contraption, one held together by super-cooled ruby receiver prongs the length of your pinkie, 70 meter wide dish antennas, pre-PC microchips, and a reel-to-reel tape recorder that wouldn’t have looked out of place in an “Eagles”-era recording studio. It was also a remarkable illustration of the new reach the human race has, via elongated digitized feeds and pinwheeling hardware, across a Solar System it could once only peer at with telescopes. And finally, it was a cobbled-together, seat-of-the-pants series of fixes, workarounds, and software patches in the great Jet Propulsion Laboratory tradition of deploying the virtuosos, kicking into high gear, and figuring out how to eke out a hardwired, long-distance living again.

Maybe literally so. As Deutsch pointed out, losing Galileo could have meant losing interplanetary exploration altogether. The ultimate footer of the bills, after all, is the public, as represented by its sometimes zealously budget-cutting representatives in Congress. And so the Galileo crises extended well past the fate of a single mission, and could have become a cancer in the whole tenuous enterprise of deep space exploration. The stakes, in other words, were very high, but the million mile screwdriver, it worked.

* * *

We rewind, in a blurred digitized shriek, past the stored data of Galileo’s years of satellite touring, past the successful deployment of its Jupiter atmospheric probe, and to the mission’s second foray into the asteroid belt, ten years ago last month. Enter Paul Geissler, planetary scientist at the University of Arizona’s seriously edgy Lunar and Planetary Sciences Lab.

I first met Geissler at the annual conclave of the American Association of Astronomers’ Planetary Sciences Division, which in October of ‘99 was held in the thermal baths resort town of Abano Terme, just outside the walls of ancient Padua – the city from which Galileo first observed moons in orbit around Jupiter. He was part of a controversial but well-respected team led by an original member of the Galileo imaging team, Rick Greenberg. It included Randy Tufts and Gregg Hoppa, the lead author of a paper to be delivered at the conference. Their findings, based largely on some innovative conceptualizing of fresh Galileo images, were the first great widely accepted keystone in an effort to establish the existence of a liquid water ocean on Europa. I had come to meet Greenberg’s group because of their Europa revelations, but now I was talking to Geissler about Galileo’s second asteroidal encounter, which had brought the first positive surprise from the mission in a long time.

“Way early on, back in ’92, I was given a project to work on, and given lead authorship of that project,” he said, stirring a perfect macchiato at the conference’s outdoor café nexus, “and in 1995 we flew past the second asteroid to ever have been looked at, and the first one to have been looked at in high resolution, Ida.”Both of the spacecraft’s asteroidal encounters had been complicated by the keyhole strictures of its trickle-down data rate, which were not yet ameliorated by the solutions of Leslie Deutsch’s team. “It was wonderful, we were locked into a room and sworn to silence,” Geissler said. “Because we didn’t have a high gain antenna, the data came in as what we call ‘jail-bars.’ Galileo would send down a line, and then skip twenty lines, then send down another line, and then skip twenty lines and send down another line, and the issue was, is the asteroid in the frame at all, and should we use our precious bits to send down this frame or should we save it for the next frame?

“In one of these jail bars you could see Ida, and then it dropped off back into space again, and then there was another little “blip.” And that’s all we had, Ok? And we were sitting in this room, we were locked up together, you know, and were being threatened with ten kinds of death if we made a peep about this until we had better verification! These particular jail-bars had three lines and then skipped a bunch, and this blip was in all three of the lines, and we were dead certain that it wasn’t a cosmic ray hit or anything like that. We knew there was something there. But we waited until another instrument on Galileo that happened to be looking in the same direction at the same time had a confirmation of it, and that’s when we announced it. But there were a wonderful few weeks when we were confident of it, and we would sort of see each other in the hall and whip out this picture, you know?”

That three-pixels-wide blip eventually materialized into a punctuation mark of a moon; effectively a tiny object circling a small object. Although asteroids had been suspected of having moonlets before, this was the sheerest hypotheses-terminating confirmation possible, and it was also a reassuring illustration of what could still be achieved by Galileo even in its compromised state. The moon was soon named “Dactyl” – Dactyls being the pixyish magicians that, according to Greek legend, live on Mount Ida in Crete. No more than a kilometer and a half across, it resembled nothing so much as the small sphere occupied by the similarly diminutive prince on the cover of the most popular edition of Antoine de Saint-Exupéry’s famous children’s tale. And its parent asteroid was revealed to have an interestingly stratified, irregular topography.

Geissler, one of the leading image processors in the planetary sciences community, has frequently been among the first to get the results of Galileo’s photography. I asked him how it felt to be the first person ever to “see” something in deep space. “There was one thing that I was the first human being to see,” he responded, “and that I think was probably one of the most thrilling episodes in my career. We had gotten two pictures of Ida up close, from different perspectives. So as the spacecraft flew past the asteroid it would snap a picture, at high resolution, and then it flew a little bit farther and then snapped another picture of the same region, again at high resolution.”

He soon realized that this separation allowed for the creation of a stereo image of the kind which, when done properly, can make an object leap into vivid, three-dimensional life. “So I processed those pictures, and shot negatives of them, and brought them home, that was late on a Friday. I had a darkroom at home, and still on Friday night I made eight-by-tens of these two, and I had pinched a stereoscope from work. And late that night I popped in these two wonderful eight by tens and saw a stereo image of an asteroid for the very first time at high resolution!” He peered at me from under raised eyebrows to make sure I understood how fundamentally cool this was. “And that entire weekend anyone who came close to my door was dragged over: ‘Look at this!’ You know, the mailman, the babysitter… That was really a thrill.”

* * *

If one had to choose a single piece of elegant inductive reasoning to serve as the most compelling example of the science findings resulting from Galileo’s Jupiter mission, it would have to be Randy Tufts’s and Gregg Hoppa’s untangling of the semantics of Europa’s lines. Europa is the ice-clad moon that was discovered to most likely have a sub-surface liquid water ocean, but that discovery, to lift a line from The Sun Also Rises, came in two ways: gradually and then suddenly. In a paper delivered in Abano Terme, Hoppa’s hour-by-hour analysis of the powerful shifting gravitational fields that play across Europa during each of the moons’ Jupiter orbits, teamed with geologist Tuft’s insights into tectonics and faulting, yielded one of the most downright aesthetic findings ever to come from space research. After a good deal of excitement over Galileo’s photographs of Europa’s rotated and then apparently re-frozen ice-bergs – provocative images, clearly, but still deemed inconclusive – Tufts and Hoppa were the “suddenly.”

Even before Galileo’s predecessor probes, the twin Voyagers, zipped through the Jupiter system at approximately the speed of a rifle bullet in 1979, scientists have known that three of the four Galilean moons have high concentrations of water ice. But only the hardiest optimists among them dared to speculate that liquid water could exist all the way out at Jupiter, more than half a billion clicks from the Sun. Europa’s average surface temperature is estimated at 100 degrees Kelvin, or about –260 degrees Fahrenheit. The North Pole in February is a steam bath by comparison.

But despite the evidence of the thermometer, two stubbornly contrarian clues surfaced in the Voyager photographic record. The most obvious was Io, Jupiter’s innermost large moon. Squeezed by the huge hand of its parent planet’s gravity, yanked the other way by the shifting gravitational fields of its three large Galilean sisters, Io produces seemingly endless chains of active volcanoes. At 3,240 degrees Fahrenheit, they are far hotter at their source than any on Earth. Io is the most volcanic object in the Solar System; the mere proximity of such an excitable object to Europa suddenly rendered sub-surface liquid water more imaginable under its ice. If such active volcanism was present on Io, why couldn’t there be some erupting from Europa’s sea bed?

The other Voyager-era clue was very subtle and mysterious – a faint whisper of potential meaning, albeit one discernable from 124 million miles out, which is the closest either probe got to Europa. (By contrast Galileo has veered to within 124 miles of the moon.) These were the long, looping chains of scalloped cracks, each joined to the next in a kind of cusp, that snake across large spans of Europa’s surface. Looking oddly like telegraph wires slung in descending, then ascending arcs – only in this case, arcs dwindling in length with each span, as though the installers had progressively run out of steam between poles – these “arcuate cycloids” extend for hundreds of miles across the crystalline topography encircling the moon’s poles. The largest of these seemingly inscrutable features were already clearly visible in the Voyager images, and Galileo had sent back many more examples at a far higher resolution. They appeared to be unique in the Solar System.

But scrutability’s in the eye of the beholder. Someone once compared the situation of a poet to that of a person standing in an open field, waiting to get struck by lightening. If he’s lucky, he’ll get hit more than once in a lifetime. Tufts’s role in cracking Europa’s arcuate cycloids code was unambiguous: he was the guy who got zapped.

A tall, free-ranging individual, intellectually and also physically, Tufts had a kind of angular bony grace as he walked – loped really – along the stone-cut walkways of Abano Terme. It transpired that he had been struck by lightening three times in his life. Before his formal training as a planetary geologist, he had been an amateur spelunker. In 1974, he and a friend stumbled on the recessed entrance to a large, unexplored cave system buried beneath the Arizona desert. They had managed to keep it a secret for an astonishing fourteen years, until it could be protected from damage. Arizona eventually invested 28 million dollars in the cave, installing heavy weatherproof doors and a misting system to keep the dry desert air out, and a month after I met Tufts in Italy, they finally opened it with much fanfare as the Kartchner Caverns State Park. “The whole idea is to develop it so that it’s environmentally preserved,” Tufts said. “I don’t know, it’s a paradox, but…”

Tufts’ second lightening strike had been his 1998 discovery of an immense fault line in the southern hemisphere of Europa. The crack, which was subsequently named Astypalaea Linea, and was revealed by Galileo photographs to be longer than the San Andreas Fault, was important because it gave clear evidence of something separating the Europan crust from the rocky core of the moon – a clear indication of a possible liquid water “decoupling” layer. Still, it wasn’t yet the clincher.

It was the third lightening bolt – his arcuate lines intuition – that we soon fell to talking about. Tufts had been fascinated by those weird ridges even before Galileo reached Jupiter in December of 1995. He recalled printing out multiple copies of the less distinct Voyager pictures of them and handing them out to his non-scientist friends, the idea being to see if they might miraculously intuit the cause. He even took the pictures to a glass-blowing factory in downtown Tucson, and asked if they’d ever seen anything like it. What did they say? I asked. “No!” Tufts laughed, scratching the back of his balding head. “I was just casting about for any kind of analogue, anything that might do it.”

With the Astypalaea fault as his subject, Tufts was working on his doctoral dissertation one night in the summer of 1998 when it occurred to him that one explanation for a slight curvature in his fault-line could be the regular shift, in both direction and amplitude, of Jupiter’s gravity during each of Europa’s revolutions around the planet. With its parent planet weighing in at over 300 times the mass of Earth, immense gravitational stresses inevitably play across Europa’s flexing ice shell. Tufts remembered going into the lab in July to “play with” Gregg Hoppa’s detailed maps of those stress fields, which calculated their evolving orientation and changing force levels. Squinting down at the print-outs while sketching lines in a small notebook with a stub pencil, he felt a growing excitement: when he followed Jupiter’s shifting influence on the Europan surface, one hemisphere of which is always facing the planet, he ended up with looping cracks that propagate in curving, stop-and-go chains – exactly what they really do.

I asked him why the cycloids do that – stop and go. Tufts explained that Europa’s slightly elliptical orbit meant that Jupiter’s gravity increases and decreases with metronomic regularity; as a result, cracks start propagating, but then as Europa recedes from Jupiter, they stop again. By the time the stresses pick up again an orbit later, they’re oriented in a different direction – one closer to the starting direction of the previous link in the chain, in fact. The procedure results in those bizarre linking cusps where the cycloids suddenly make an about-face. Finally and most intriguingly, the whole process couldn’t happen without the existence of a large body of sub-surface water to exert tidal pressure from below – something which Gregg Hoppa had been the first to realize.

The whole idea ultimately has an almost sculptural simplicity, and later I couldn’t help but thinking of Roger Diehl and his VEEGA trajectory. It too, had somehow been waiting for discovery, lost in plain sight among a tangle of alternate trajectories; it too ultimately looked simple, the way a triple pirouette by an ice dancer might look simple, though it had presented itself as a solution only after much obsessive work; and like Tufts’s cycloids, it too curved gracefully through space and time, its arcs and reversals subject to gravity’s uncompromising but explicable cable work. And one revelation hinged irrevocably on the other; Tufts’s Europa insight would’ve been impossible without Galileo data, and Galileo wouldn’t have gotten near the moon without Diehl’s VEEGA trajectory.

Tufts had tossed a two-decade career as a community organizer to focus on Europa research, and it turns out he had one overriding motivation beyond sheer scientific curiosity. “Because I was interested in politics, I thought a lot about what kinds of things would best promote world peace,” he told me. “One of those, it always seemed to me, would be to find life somewhere else. It would give us a vastly new perspective on existence.” He laughed. “I mean, on the one hand, it might take us down a peg, which always could be useful. And the other thing it might teach us is that life is what the Universe does. What is the Universe? It might be a great mechanism for creating consciousness.” And then he excused himself: he had a date with a working group that was proposing instruments for what was to have been Galileo’s successor, a Europa Orbiter. It was supposed to have been launched this year, but the mission was cancelled in 2002 for budgetary reasons.

In late April of that year an obituary appeared in The New York Times. It gave an accurate account of two of the three lightening strikes that had graced Randy Tufts’s life in the 53 years before a rare bone marrow disorder suddenly felled him: the Arizona cave discovery and that of Europa’s 600-mile long Astypalaea Linea fault. It didn’t, however, say anything about the feat of deduction that had unraveled Europa’s cycloidal ridges conundrum and become the first great confirmation of a sub-surface liquid water ocean there. Maybe this was because Gregg Hoppa was the lead author of that paper, or maybe it was just because space constraints precluded trying to explain the thing in an obituary. How, after all, to put this story of pure logic – the logic of natural forces in cyclical motion, but also the force of the natural workings of the human brain, that mysterious instrument capable of using robot visions to deduce the origins of encrypted inscriptions on the face of a moon that’s six million miles away – into a newsprint death notice? It may have been the most wonderful revelation to have happened to this science-minded poet of the field, and it may yet prove to be a key finding on the way towards the discovery of extraterrestrial life. But three strikes and you’re out.

* * *

The obit, however, ended on a prescient note. Just as Tufts protected his cave, his wife Ericha Scott is paraphrased as saying, so he wanted to establish safeguards to protect whatever life might exist on Europa from damage by spacecraft. And that’s the leading rationale behind Galileo’s death dive. Unlike NASA’s Mars landers, which are sterilized before launch, Galileo may still harbor some of the microorganisms which inevitably hitch a ride on our space robots. If left in its orbit around Jupiter after running out of propellant, there's a chance that it would eventually crash into Europa, potentially seeding the moon with aliens from Earth. (The spacecraft also carries potentially dangerous plutonium pellets in its two power generators.)

In late July I called Arthur C. Clarke at his home in Sri Lanka and asked him to comment on Galileo’s death sentence. I was interested in his view of the planetary protection reasons behind it. Clarke has had his periods of Europa fascination; in fact he put a mysterious form of intelligence in the Europan ocean in his sequels to 2001: A Space Odyssey. But instead of steering me towards 2010: Odyssey Two he mentioned an old short story of his, Before Eden, which was published in 1956. “It’s all about the danger that we might contaminate new worlds,” Clarke said. Later I found the story, which describes a scouting expedition to Galileo’s first fly-by destination, Venus. The expedition left behind that archetypical human artifact, a bag of waste. The waste ended up contaminating a strange Venusian life form they’d discovered there, ending its evolution. I concluded that Clarke probably endorsed NASA’s plan to destroy Galileo.

In January of 1997, shortly after Galileo had first eked Earthwards its pictures of piecemeal ice floes in the Europan ocean, Galileo project director Bill O’Neil and mission science director Torrence Johnson had an audience at the Vatican with the Pope. It was a scene the Roman Catholic Church’s most famous heretic no doubt would’ve appreciated: the scientists running the mission named after him meeting with the very pontiff who had finally acknowledged that the shrewd astronomer might’ve had a point after all. Presented with robot Galileo’s photographs of Jupiter’s excellent strange satellites, multilingual John Paul III studied the branching, forking, curving cracks that fissure Europa’s silvery surface and pondered for a minute. Then he looked up. “Wow,” he said.

Three years after that Vatican audience, and three hundred and sixty after the first Galileo was dragged from Florence to a nearby Inquisition courtroom, a mouthful of a committee – the National Academy of Sciences' Space Studies Board's Committee on Planetary and Lunar Exploration, or COMPLEX – delivered itself of a verdict on his successor. Asked by NASA to study the various options for ending the Galileo mission, it recommended disposal of the craft either through a controlled trajectory into Jupiter or into its presumably sterile volcanic moon Io. They ruled out another option, that of slinging the probe out of Jupiter orbit, because of “the very small, but nonzero, chance of eventual impact with Earth." Galileo would not be given the slightest chance to come home.

When I asked Leslie Deutsch what his reaction had been when he’d heard of the decision, he said he was initially angry, though he understood the rationale behind it.Over the years, he admitted, he had become emotionally attached to the distant robot emissary, adding that it was only the second time that NASA had deliberately killed a functioning spacecraft. When I asked Bill O’Neil the same question, Galileo’s long-serving project manager – and one of the key architects of the effort to save the mission – mulled it over for a few days, then sent me an e-mail. Galileo’s end end would bring a personal sense of satisfaction at what had been achieved, he wrote. Still, he found it ironic “Galileo Galilei only got house arrest by his sponsor the Roman Catholic Church for discovering things they didn't want to be true, whereas our Project Galileo gets a death sentence from NASA for its greatest discovery – the prospect of life on Europa."