A group of astronomers has honed a new way of searching for alien civilizations in the stars

NASA's Perseverance rover landed on Mars last month in a mission to, in part, search for signs of ancient microbial life on our neighboring planet. If any evidence of such extraterrestrial organisms — likely long dead — is found, it will be a massive development in astrobiology and humanity's understanding of its place in the universe. But when people are honest, they tend to admit that fossilized remains of ancient microorganisms aren't what gets them excited in the search for alien life. What really fires the imagination is the search for intelligent alien life and extraterrestrial civilizations.

A recent paper, though, flagged by astronomer and writer Phil Plait at SyFy, offers surprisingly promising developments in that branch of space exploration. Working at UCLA, a group of astronomers published findings on a new method of searching for signs of technological civilizations in the stars.

"We selected 31 Sun-like stars," they explain, "because their properties are similar to the only star currently known to harbor a planet with life."

And to be upfront about it, they unequivocally found no sign of civilization. But Plait argued that they showcased a promising method for pursuing the search.

Basically, the technique entailed using the Green Bank Telescope in West Virginia to scan for radio signals. Many scientists hopeful about searching for intelligent alien life think that the most likely way we'll uncover it is by receiving radio transmissions from non-human sources. Radio signals are easy to create, they can travel the long distances between solar systems, and they can encode detailed information. And if there are any civilizations out there listening, they may have already picked up many of our own transmissions.

But the universe is filled with radio signals naturally, so the key is to detect those that appear to be artificial. And inconveniently, most of the artificial radio signals we detect while looking at space have a mundane source: humanity.

That's where the UCLA researcher's novel work comes in. They've developed an algorithm to weed out extraneous signals in the data, pinpointing those most likely to be caused by aliens. Plait explained:

Finding a sharp, narrow radio signal isn't that hard. But if there are a lot close together in frequency some search algorithms can get confused, and only find the brightest one in the pack.
What they did instead was search using what is called topographical prominence. This is a concept used when mapping mountains, and it quantifies how tall a mountain is relative to the terrain around it. You can get a tall mountain, for example, that's easy to spot if it's relatively isolated (like Kilimanjaro, for example) but more difficult to ID if there are lots of slightly smaller mountains all around it.
By measuring the prominence of narrow radio signals they were able to find signals that were still interesting but unlikely to be found using standard search methods. They wound up finding something like 200 times as many candidate signals as other methods this way.

After scanning each of the stars for five minutes twice one year apart, they found more than 26 million signals. But the algorithm let the researchers screen out more than 99 percent of those signals as human-made. Another 90 percent were also clearly from radio signals, and after all that, 4,539 were left over. Checking through them one-by-one, the researchers were able to conclude these, too, were from human sources and not evidence of alien technology. A disappointment, to be sure — but the real point was to hone the technique.

They also did a neat trick to check their methodology. In a separate part of the study, they injected fake technological signals into their raw data to see if their algorithm could pick them out. And they found:

We make a first step towards the implementation of this tool and show that our current pipeline detects 93% of the injected signals over the usable frequency range of the receiver and 98% if we exclude regions with dense RFI. In addition, our pipeline correctly flagged 99.73%of the detected signals as technosignature candidates. Although our current implementation requires additional work to fully capture the end-to-end pipeline efficiency, it can already illuminate imperfections in our and other groups' pipelines and be used to calibrate claims about the prevalence of other civilizations...

If aliens are hanging out in the galaxy communicating by radio signals, "this process would find them," said Plait. "I don't know if we'll ever find aliens, let alone how. Unless faster-than-light travel is possible (and I'd bet against it right now) it's really hard to go from star to star. Not impossible, just hard, and it's far easier to just send radio signals instead. So this does seem like a decent bet."