New Frontiers in SETI Research
Gerry Harp, who has taken over Jill Tarter’s job as research director at the SETI Institute, has high hopes for the hunt for alien life
Photo: SETI Institute
23 July 2012—In 1960, astronomer Frank Drake pointed a single radio dish at two nearby stars to hunt for signals from extraterrestrial civilizations. Since then, the search has gotten considerably more high-tech. Among those leading the charge is IEEE Member Gerry Harp, who in May took over as director of research at the SETI Institute, a position most recently held by famed signal hunter Jill Tarter. IEEE Spectrum Associate Editor Rachel Courtland talked to Harp about the track record of SETI (Search for Extraterrestrial Intelligence) so far and what can be done to improve the search.
IEEE Spectrum: The Allen Telescope Array is the only telescope designed specifically to search for signals made by an extraterrestrial intelligence. How does it work?
Gerry Harp: The Allen Telescope Array is not a single-dish telescope. It actually has 42 dishes that each make a measurement of the electric field arriving from space. We take signals from all 42 dishes and add them together with the appropriate coefficients to make an instantaneous snapshot of a very large field of view on the sky.
Because it’s an interferometer, it’s possible to take the signals from each antenna and split those signals, say, three different ways, and then phase them up at three different points on the sky so we can get three different pointings at once. That speeds up our search by a factor of three. But it also gives us a special benefit of anticoincidence RFI detection. RFI is radio frequency interference. We find lots and lots of artificial signals every day, but almost all of them are human made—so far all of them, as far as we know.
If we see a signal in one beam, and we find the signal in another beam pointed in another direction, then we know that signal must be coming from Earth. We have a lot of bright sources. We have all the geosynchronous satellites up in space, which number in the hundreds, and then all the low earth orbiting satellites that number in the thousands. If any one of those appears close to our field of view, it can generate interference. But interference is never local; it’s always a wash across the image, so we can tell by this three-beam method if something is really coming from far away.
IEEE Spectrum: What exactly are you looking for?
Gerry Harp: The typical signal [we’re searching for] is a very, very narrow band signal—we’re talking about 1 hertz wide. There’s all sorts of reasons why that seems like a good choice, not the least of which is that there’s no natural signal that is as narrow as 1 Hz. So if we find such a signal arriving from outer space, it’s either E.T. or it’s a very interesting astrophysical artifact.
It’s always been, from the very start, a guessing game. At the very beginning, Frank Drake made the guess that they would be sending us a very narrow band signal. That was a good guess, and it turns out that subsequent science supported that guess. But more recent developments in technology have called that guess into question, and now we have to ask, “Well, what other ways might they be sending us signals, and how shall we look for them?” We have to develop algorithms to find that stuff, and it’s a very challenging problem.
IEEE Spectrum: Why was it a good guess to presume that an alien civilization would send a narrowband signal?
Gerry Harp: Well, it has to do with what happens to signals as they propagate through the space between stars—we call that the interstellar medium. That interstellar medium is gas that’s residual after making the stars in the galaxies. Some of that gas is ionized, so you have free electrons in there. Light interacts with [those] electrons, and its speed changes. High frequencies are less affected, and so they travel faster and arrive sooner than the low frequencies. This is called dispersion. If you have a very, very narrow signal, there’s really only one frequency involved, so it really doesn’t matter if it gets delayed. When it arrives at the source, it’s still intact. It’s just a single tone.
But for the last several years, we’ve been developing ideas to use the telescope in special ways to expand SETI beyond what it has ordinarily been for the last 50 years. E.T. may not send a very narrow signal. The reason is [that] human communication has been going in the direction of wider and wider bandwidth.
A very narrow signal that’s only 1 Hz wide carries very small quantities of information. A 3000-Hz signal can carry info up to 1500 bits per second. Many of us think that if E.T. is going to go through all the trouble of building a beacon for us, they’re going to build one that has some information encoded on it.
IEEE Spectrum: How do you hunt for a wideband signal?
Gerry Harp: Recently we’ve started to use the telescope more as an interferometer. The imaging system gives us the ability to form, instead of three beams, 1250 beams in the field of view all at once. So you can imagine this is much faster than [the] three beams approach. We can cover a lot more sky at the same sensitivity; however, the spectrometers are not very high resolution. The smallest resolution is about 3000 Hz instead of 1 Hz. By nature that means we’re looking for a different kind of signal. We’re looking for alternative kinds of signals that may only be on for a short period of time.
IEEE Spectrum: When you hunt for such wideband signals, don’t you have to take this frequency-dependent delay—this dispersion—into account?
Gerry Harp: That’s normally what you would have to do. There’s a fellow named Andrew Siemion at UC Berkeley who works on the SETI@home project. He looks for dispersed pulses and uses other people’s computers to do the processing.
I’ve been working on another technique called autocorrelation spectroscopy. Autocorrelation spectroscopy costs computationally about twice as much as the conventional spectroscopy search, so it’s something you can do without much extra effort. And it happens to have this feature that it doesn’t matter whether the signals are dispersed or not. If the signals are repeating—maybe it’s a string of symbols sent over and over again—they will still show up in the spectroscopy independent of what the dispersion is. That’s the approach that we’re excited about. But you can see that there’s competition in this. Young people are getting into it and trying new ideas.
IEEE Spectrum: Do you find it at all surprising that we haven’t seen any promising signals since that “Wow!” signal found by Ohio State University in 1977?
Gerry Harp: The Wow! signal has never been seen again. It turns out that we see signals like [that] all the time—spurious signals that are detected only once and then they disappear. We’ve pretty much figured out these have to be human signals. I think the reason we haven’t heard a lot of excitement in the last 35 years is because we made the correct decision to get the humans out of the loop and let the computers do their work with algorithms that are scientifically valid.
IEEE Spectrum: Are you at all looking forward to the arrival of the Square Kilometre Array [SKA], the large collection of radio dishes that’s set to be constructed in Australia and South Africa?
Gerry Harp: The Allen Telescope Array [ATA], even if we’ve built up to 350 dishes, will still only be 1 percent the size of the SKA. It’s going to be a really, really big and sensitive telescope, so we’ll be able to look at stars that are much farther away from us and be able to detect very weak signals.
With a radio telescope, it’s possible to amplify a signal and split it many ways. We do that at the ATA. We have two correlators and three beam formers. We’re always splitting the signal five different ways to different instruments. There’s really no limit on how many different ways you can split a radio signal. We fully anticipate that on the SKA they’re going to reproduce the output of their array ad infinitum and provide workstations for anyone who can afford to pay a small premium for a spigot of information out of the telescope. The only thing that you can’t share is the pointing of the telescope—it can only point in one direction at once.
IEEE Spectrum: Are there any big challenges facing SETI research?
Gerry Harp: For the last more than a decade, there hasn’t been any government funding. The National Science Foundation and NASA—just two places that you think might support SETI—have not supported it at all. I think that if the public were more aware of it they might wonder why. It really has to do with shortsightedness. There’s a real unwillingness to support high-risk, high-reward science.
IEEE Spectrum: But it must be hard to demonstrate progress in this field to potential funders, since you’ve searched for decades without finding anything.
Gerry Harp: The search we’re doing now is practically incomparable to the search that was done 35 years ago. We’re sifting through so many more signals. The search keeps expanding at an exponential rate. It doubles like Moore’s Law. We’re looking at millions and millions of frequencies at a time, and we have these multibeam techniques we’re starting to work on.
And we are learning something, although what we’ve done so far is just a drop in the bucket. When SETI becomes more complete—when we have more complete information about what sorts of signals are arriving from stars in the galaxy—we are going to start making statements of the probability of there being intelligent life in the galaxy. Even a null result is a positive result, because we can put upper limits on the frequency that some life like us comes along. So far, we really haven’t done the search to make much of an exclusionary statement, but we can say it is certainly true that not every star system has something as interesting as us on it. That tells you something about humans and our place in the galaxy and in the universe. That’s an important question we can answer even if we never discover another species like us.
IEEE Spectrum: I’ve heard you have a bottle of champagne permanently on ice, just in case.
Gerry Harp: That is definitely true. Every couple of years we take the bottle out and open it and drink it because champagne doesn’t last that long. Then we replace it with another bottle. We have a celebration of what we’ve accomplished so far and then carry on.