small logo Editorial

Observations Must Be Paired
by Dr. H. Paul Shuch, Executive Director

A few months ago we reported the first anomalous signal detected by our Project Argus search. That hit turned out to be interference from a military satellite in Low Earth Orbit (LEO), a fact which was readily established by software analysis of the signal's Doppler shift. But classifying candidate signals is not always that straightforward. As more and more amateur SETI stations come on line, we'll be getting more interesting signals, and finding a need to formalize our signal verification and follow-up procedures.

A case in point is an interesting signal received last month by SETI League member Daniel Fox. As he tells the story:

"While doing testing on version 1.05 [of his SETIFOX program -- ed.] I had one detected carrier that showed up that might be worth checking into. Using a very long integration time and compensating for the Doppler shift of the Earth's rotation I only had one hit on the system and this was shortly after I moved the dish and looked at a portion of the sky I don't normally look at.

Information below:

When the R7100 was tuned to 1420.001500 a carrier was detected in bin 226 which works out to a 904Hz tone. The actual frequency of detection would be 1420.001500 plus 904 minus an offset for USB mode (about 400Hz).

The integration routine is set up so that the power spectrum of the sample is added to the old power spectrum times a decay factor (in this case .95). This was repeated indefinitely. An estimate of its signal strength may be difficult since it could be a strong signal on for a short time or a weak signal on for a long time.

By the way, the reason I picked that particular declination was that it was near the same declination as the Crab Nebula. I had also been recording the relative power and plotting it versus right ascension for various declinations to see what was detectable with my system. I notice that the noise level varies quite a bit with outdoor temperature and have been considering some sort of temperature regulation on the amplifiers outdoors. "

signal spectrum Dan's received power spectrum, shown here, is not so readily identified as was our first candidate signal. In fact, with but a single sample, we cannot rule out terrestrial interference, satellite interference, or equipment malfunction. Neither can we confirm them. So we're left with something of a mystery, probably the first of many we will encounter in our search.

Nearly all previous SETI studies have uncovered similar unexplained phenomena. The best known of these, the Ohio State University "Wow!" signal of 1977, was as brief, intermittent and elusive as our present example. The several dozen signals detected by Project SERENDIP's Arecibo deployment were consdered consistent with random noise. None of the interesting signals detected by Paul Horowitz's Project META ever repeated, although their distribution along the Galactic plane certainly appears non-random. And none of Project Phoenix's many hits at Parkes was seen by the follow-up detection device at Mopra.

The rule in SETI has been, and must remain, "unverified signals prove nothing." Tantalizing as these signals have been, they do not constitute the existence proof of other civilizations, which SETI seeks. And neither does Dan Fox's signal.

Are we ever to be tantalized by unverified mystery signals? Not if The SETI League has its way. Once we have achieved critical mass (and that will take far more than our present two dozen search participants), we should be in a position to verify our detections. One strategy endorsed by SETI pioneer and SETI League advisor Dr. Frank Drake is to pair our observing stations. To be effective follow-up detection devices for one another, two Project Argus stations will need to measure frequency, and especially Doppler shift, most accurately. As Drake tells us:

"The important point is that with the two dish system, a true extrasolar signal does not give the same Doppler pattern at each antenna (unless the source is on a certain great circle, as the Phoenix crew learned to their dismay). Because the rotation of the Earth creates a Doppler offset in the received frequency at each antenna, the received frequency and its change with time will be different at each antenna (except for sources on that great circle). So there will be a different Doppler offset at each antenna, and the change with time of the offset will be different for each antenna. These differences, which are precisely calculable for any source position, create the really powerful discriminant against RFI. They are the heart of the method."
Right now, we're in the shakedown phase. We're still trying to help as many members as possible to build reasonably sensitive radio telescopes, and learning how to use them for SETI. But within just a few years, it should be possible to assign each participant a partner. Linked through the Internet, calibrated via GPS, paired stations will be in a position to instantly verify their candidate signals. Most will be ruled out, as either not simultaneously received by two distant stations, or not fitting the differential Doppler profile which Drake outlines above. But with luck, someday we'll receive a signal which passes all the tests. The challenge for today is to figure out what those tests must be. Certainly, independent verification is one of them. Drake again:
"I think it is good to push hard for the amateurs to use the simultaneous two-dish approach, with careful recording of frequency and frequency drift, so that the Doppler signature can be used. Otherwise the observers will always have inconclusive results, unless, perchance, a clear extraterrestrial TV broadcast pops up on the screen!"

Click to email the Webmaster
| Home | General | Memb Svcs | Publications | Press | Technical | Internet | Index |
entire website copyright © The SETI League, Inc.
this page last updated 4 January 2003
Click for top of page
Top of Page