With Tarter in the lead, the privately funded SETI Institute (Congress cut all NASA funding for SETI in 1993) conducted the most comprehensive search to date, called Project Phoenix, from 1995 to 2004. Phoenix targeted about 1,000 sun-like stars, spending a few minutes listening for radio signals from each. “There wasn’t a transmitter pointed our way when we looked,” says Tarter. “Does that mean there’s no technological civilization? I don’t know. That’s a much harder conclusion to draw.”
Phoenix took 10 years because the SETI-ologists had to borrow observing time, a week here and a month there, on other people’s radio telescopes, including the 1,000-foot-diameter Arecibo dish in Puerto Rico and the 210-foot dish at Parkes, Australia. Tarter’s new array of telescopes will do SETI operations all day, every day. It will also be used by scientists for conventional radio astronomy—Berkeley is partnered with the SETI Institute on the project—but this time, they’re the ones who will be along for the ride.
Phoenix listened for signals over a range of three gigahertz—a wide swath of radio spectrum by SETI standards. The ATA will monitor 10 gigahertz continuously, targeting 1 million stars with enough sensitivity to detect an Arecibo-size transmitter broadcasting from 1,000 light-years away. That is, if the array grows from the current 42 to the planned 350 telescopes. Any radio telescope’s sensitivity to faint signals depends on its total collecting area, whether it’s a giant single dish like Arecibo or lots of little ones. The ATA philosophy is to build many small dishes as cheaply as possible, then rely on sophisticated software to process the signals.
Economy is a necessity for the project, which has only seven people working at the Hat Creek site. They buy off-the-shelf when they can, invent when they must. The small secondary reflector attached to the front of each dish, along with the telescope’s electronics, is covered with a shroud that’s fabric on top, aluminum below. The project’s engineers had to test all kinds of fabrics before they found one that kept out water but let in radio waves.
Tarter opens the shroud from the aluminum bottom and we poke our heads up inside, where we’re hit by a blast of hot air. Temperatures in the valley routinely top 100 in summer, and the sensitive electronics have to be kept cool (mini-refrigerators used in cell phone towers turned out to work nicely).
Inside the shroud is a spiky, silver-gold device that looks like an artificial Christmas tree. This is the telescope feed, which Tarter says reminds her of something from Flash Gordon. This particular design is another ATA innovation, and a crucial one. The feed is where radio energy collected by the dish is focused and converted to a signal containing the multi-frequency SETI data, which is then sent via fiber-optic cable to computers for processing.
The brains of the telescope array are inside the computer building. That’s where digital signals from the individual telescopes are sorted and manipulated, turning this field of small dishes into a large and powerful phased array (see “How Things Work: Phased-Array Radar,” June/July 2006). Because radio waves from a target star reach any two telescopes at slightly different times, the peaks and troughs of the waves are slightly out of synch, or phase. The SETI computers can artificially shift the phases to match up, effectively combining the waves and boosting the signal. Or, equally useful, a signal can be canceled out by phasing up the troughs in the waves. That enables the team to filter out, for example, an annoying, beeping satellite known to pass over Hat Creek every night, which otherwise might be mistaken for a broadcasting alien. For a computationally intensive program like SETI, that’s huge.
The more antennas, and the more random their pattern (hence the deliberately scattered placement), the better the technique works. “It’s a lot of vector algebra, that’s all,” says Tarter. Her bachelor’s degree from Cornell was in engineering physics, and she’s very much a hands-on experimentalist. I ask her if, despite the workload, she likes all the tinkering and testing. “Sure,” she says. “The frustrating thing is, I want it now. There are observations I want to make.”
In fact, now that the first section of the array is in place (the 42nd antenna was installed in February), the ATA will make its first observations this summer. The dishes will be pointed toward the center of the Milky Way galaxy, which is thick with stars. The bad news is that most of them are extremely far away—the most distant extraterrestrials would need a transmitter 20,000 times more powerful than Arecibo’s to be heard. It’s a long shot, but worth trying while the array is still under construction, partly because such a broad survey has never been done.
Meanwhile, the catalog of candidate stars (those most likely to be orbited by habitable planets) has grown to 250,000, and will eventually reach 1 million by the time the array is finished. As for when that might be, it depends on money.