Credit: NASA
IN 1969, JOHN D. ANDERSON, an astronomer with NASA's Jet Propulsion Laboratory (JPL) in California, proposed a simple experiment.
When the Pioneer 10 and 11 spacecraft launched (1972 and 1973, respectively) flew by Jupiter (in 1973 and 1974), why not use the frequency shifts in their radio signals to map the giant planet's gravity field?
The theory was simple: as the radio signals fought their way out of Jupiter's gravity, they would lose energy, slipping slightly toward lower frequencies.
Careful measurements of that shift would translate into ultra-precise measures of the planet's gravity, which is something planetary scientists could use in order to make deductions about Jupiter's interior.
There was only one glitch: when the data was compiled, it turned out that the radio signals were slowly shifting frequency even while the two spacecraft were nowhere near Jupiter.
The most obvious explanation was that the spacecraft were changing speed. That would produce a Doppler effect in their signals, comparable to a police siren changing pitch as the car flashes past.
Such an effect, in fact, was expected as the probes slowed, on their way toward interstellar space. But they were slowing by too much. "I was having to account for a small, unmodelled force in order to fit the data," Anderson says.
As far as scientists listening from Earth knew, aliens could have lassoed the probes with some kind of force field and tugged them ever so slightly back toward the Sun. Not that anyone believes extraterrestrials were involved. The problem is that nobody's really sure exactly what is involved.
"It's one of those serendipitous things where sometimes the good science is in the surprises," says Anderson, who spearheaded the team that published their discovery in the Physical Review Letters in 1998.
The effect isn't large: equivalent to about .00000001 times the strength of gravity on Earth. That's enough to change the spacecraft's velocity by a whopping millimetre per second every fortnight, give or take a bit. Not much, but over the course of years it adds up.
Theories for the 'Pioneer anomaly', as it has come to be known, range from the mundane to the sublime.
At the mundane end: something on the spacecraft sprang a leak, producing an exhaust jet that slowed it down. At the exotic end: there's a fundamental flaw in our understanding of physics, possibly involving a previously unknown type of force.
But don't get too excited. "The thinking is that it's probably not new physics," Anderson says. "Those things don't come along very often."
Slava Turyshev of JPL's Relativistic Astrophysics research group agrees. "Before you report that you have discovered a wonderful new physics, you have to do your homework," he says.
That homework involves weeding out mundane explanations, one by one. In some cases, it's easy. A gas leak, for example, shouldn't have continued for years. "I don't think anybody's suggesting gas leaks any more," Anderson says.
Another simple prospect was that the force relates to solar radiation pressure. When sunlight hits an object, it imparts very slight momentum to it, like wind on a sail. It's a big enough effect that theoretical physicists and science fiction writers have sketched plans for 'solar sail' spacecraft that could take advantage of the effect for propulsion.
The strength of the pressure depends on the reflectivity of the surfaces being hit by sunlight. Brighter surfaces get more impulse, so if something after launch altered the spacecrafts' sheen, the radiation pressure might have changed by enough to account for the anomaly.
Radiation pressure, however, falls off with the intensity of sunlight, which drops with the square of the distance from the Sun.
The Pioneer probes were tracked long enough before NASA finally lost contact with one, in 1995, and the other, in 2003, for the scientists to determine that the anomaly doesn't follow this formula. So much for that idea.
As for the possibility that we need to revise our understanding of physics, perhaps dark matter – an unknown substance whose existence is inferred from its gravitational effect on galaxies – is somehow retarding the spacecraft as they zoom toward interstellar space.
But that would imply that there is dark matter within the Solar System, and nobody's ever seen any other evidence of that.
As alternative theories fall by the wayside, it's tempting to wonder whether it really is new physics.
Perhaps our understanding of gravity is off by a tiny amount. Or perhaps something involving gravity or velocity altered time on the two spacecraft, causing the signal drift.
Such effects indeed occur under Einstein's theory of general relativity, but all known ones have been taken into account. "You might have to modify general relativity to include the Pioneer anomaly," Anderson says.
Adding credence to this concept is the observation that the mystery force appears to be very close to the product of two fundamental constants: the speed of light and the Hubble constant, which is related to the rate at which the universe is expanding.
Specifically, the Pioneer anomaly, for both spacecraft, is an acceleration of 8.7 +/- 1.3 x 10-10 m/sec2. The product of the speed of light and the Hubble constant is about 6.7 in the same units. "Some people think that's significant," Anderson says.
But not all of the homework has been done. Another mundane explanation is that the anomaly is due to waste heat from parts of the probe facing away from the Sun.
It's another light-pressure effect, although in this case it's the probe, not the Sun, that's emitting the light. In this case, the light is infrared radiation, and while that is less energetic than visible light, it too would cause the spacecraft to recoil slightly in the opposite direction.
Some of it might also bounce off the back side of the probe's antenna, causing it to act like a tiny light sail.
One source of heat is from the probes' RTGs (radioisotope thermoelectric generators), which are electrical generators powered by heat from radioactive decay. These were designed to radiate heat equally in all directions, but if they didn't quite do so, the disparity might produce enough acceleration.
But again, there's a problem. The power from RTGs falls off in tandem with the radioisotopes' half-lives. And that, Anderson says, isn't what the scientists have been seeing.
Still, Turyshev says, preliminary assessments by Gary Kinsella at JPL found that there is enough heat from the RTGs to account for at least part of the anomaly. "What part, we don't know yet," Turyshev says. "Basically, there is no number yet."
More importantly, during Anderson's initial studies, there were only 11.5 years of data from Pioneer 10, and 3.25 years of data from Pioneer 11.
Thanks to work by Viktor Toth, a Hungarian-Canadian software engineer (who, like a number of people studying the anomaly, works on the problem in his spare time), old records have been located and digitised, allowing the probes' flight paths to be examined in much greater detail.
"Now we have roughly 20 years of Pioneer 11 and maybe 18 years of Pioneer 10," Turyshev says. And in addition to the Doppler data, his team recovered data from 114 on-board sensors. That means more than simply modelling heat loss from the RTGs; it permits modelling of infrared emissions from other equipment in exquisite, day-to-day detail.
"That information can be used to build a high-fidelity thermal model of the spacecraft," Turyshev says. With that information in hand, the scientists can then create a computer model that will 'fly' the two spacecraft "year by year and manoeuvre by manoeuvre," he says.
It's likely that there will still be an anomaly, but at least this will further refine our estimates and possibly reveal part of the cause, Turyshev says. "That's a guesstimate. We should have results in about 10 months."
An obvious question is why subsequent probes to the outer Solar System haven't shown the same effect. Sadly, these craft weren't designed to look for it.
Some, like Galileo, simply didn't go far enough out. Others, like the Voyager or Cassini missions, used thrusters for attitude control, rather than being spin-stabilised (or spun about an axis to maintain orbital control, similar to a gyroscope), as the Pioneers were.
Uncertainties in the power of thruster bursts are enough to overwhelm any Pioneer-style anomalies in the trajectories of the other probes.
The best hope for the future lies with the New Horizons probe, which is presently headed for Pluto. "It wasn't designed for precision tracking," Anderson says. "[But] it may yet return some good data because it's on a good trajectory, and it's a spinner."
Interestingly, however, several closer spacecraft have shown anomalous behaviours. The first of these was Galileo, which whipped by Earth on 8 December 1990, en route to a 1990 rendezvous with Jupiter.
When the spacecraft emerged from its speed-boosting fly-by of Earth, it was quickly clear that it had gained more energy from the manoeuvre than expected. Since then, similar anomalies have been seen in some other Earth fly-bys.
"We get a little boost that we didn't expect," says Peter Antreasian, a spacecraft navigator at JPL. "I was working on the NEAR project in 1998 [Near Earth Asteroid Rendezvous probe], where we saw a sizable increase in energy, equivalent to 13.5 millimetres per second."
In an article published in March 2008, in Physical Review Letters, Anderson and four colleagues examined all such fly-bys for similar effects.
Sometimes it was there; sometimes not. Galileo, for example, made a second pass in 1992, but no unexpected deviation occurred. But in that case, the satellite's trajectory took it much closer to Earth.
"We were watching for [the effect]," Antreasian says, "but we were at a lot lower altitude, and you start to get drag effects from the atmosphere. The uncertainty in that masks out this anomalous delta-v."
Similarly, the Cassini data was useless because the spacecraft was firing its thrusters at the time of the fly-by. Of the others – NEAR, Rosetta and Messenger – only Messenger failed to show a fly-by anomaly.
What's happening is anybody's guess. Despite some sensational media reporting, the only thing the two have in common, Turyshev believes, is that both involve spacecraft, and both are called anomalies. "Nobody has established that the two are connected," he says.
Anderson's not so sure. He's wants to know if the new, extended trajectory data on the Pioneer missions will show that the Pioneer anomaly is somehow "turned on" by the probe's encounters with Jupiter and (in the case of Pioneer 11) Saturn. If it is, the two might be linked.
"We can see how it's affected by going by a planet," he says. Turyshev, on the other hand, finds this unlikely; any evidence of a 'turn on' effect is an artefact of the way in which the data was initially recorded, he says.
Anderson also found that the amount of the fly-by anomaly depended on the angle at which the spacecraft slid by Earth.
Messenger whipped closely around the Equator and had no measurable anomaly. NEAR and, to a lesser extent Rosetta, came in with more of a north/south angle.
"So it appears to us that the Earth fly-by anomaly is related to the spin of the Earth somehow," Anderson says. "How? We don't have an answer for that. It's another mystery."
Turyshev and Antreasian, however, doubt that it's a gravitational effect at all. Most likely, they say, the anomaly is simply an error in the computer code that is used to shift between Earth-bound and space-based coordinate systems.
"That's done quickly in the fly-by, and if you have mismodelling, you will get a velocity-dependent error and latitude-dependent error," Turyshev says.
Meanwhile, it's not likely that either anomaly will cause spacecraft engineers to crash a probe. "We navigate spacecraft really well," says Antreasian. "This kind of discrepancy doesn't affect our accuracy much."
Still, he says, "we don't know everything". He's speaking specifically of the fly-by anomaly, but the message could apply equally well to either discrepancy. "That's the thing. There's always something there to explore and try to understand in more detail."
RICHARD A. LOVETT is a science writer in Portland, Oregon, USA.
