April 28, 2008

The Fermi "Paradox"

If you haven't heard of it before, Fermi's so-called paradox asks why, if the universe is billions of years old, and contains billions of billions of star systems, we haven't found any evidence of alien life elsewhere in the universe. In theory, even if an intelligent civilization only colonized one new star system every thousand years, they could have colonized the entire galaxy in a few tens of millions of years.

Endless argument goes on about this subject. Science fiction authors enjoy writing about the most dramatic possibilities: most civilizations wipe themselves out, or, there are aliens out there who actively wipe out upstart competitors whenever they find them. Many academics instead argue that the series of events leading to intelligent civilizations is so improbable that we are likely the only one in the observable universe.

Permeating all these discussions is the assumption that if intelligent life were out there, we would know about it. (I'm not talking about the theory that we're in a nature preserve, and all the aliens out there are carefully leaving us undisturbed.) It's blithely assumed that because we've been broadcasting radio since the 1930s, any alien civilization within 70 light years should have noticed us by now. Conversely, it's assumed that we would have noticed radio broadcasts from ancient alien civilizations.

These assumptions eventually bothered me enough for me to look around for some hard numbers. The SETI@home FAQ has some nice detailed equations and tables of numbers about this very subject. Summary: the power of a radio broadcast drops with the square of the distance from the transmitter and receivers have to be able to distinguish the transmission from natural background noise (e.g., the sun, the center of the milky-way galaxy, etc.)

What it comes down to: with the Arecibo radio telescope, one of the most sensitive radio telescopes on the planet, we would be hard-pressed to detect our entire civilization beyond a distance of 0.3 light-years. The number of star systems potentially containing alien civilizations within that distance? Zero. We couldn't detect a civilization of our level if there were one in the nearest star-system, Proxima Centauri, 4.3 light-years away.

Detection of alien civilizations through radio broadcasts at ranges of ~100 light-years (a volume which contains only a few thousand star systems, out of ~250 billion in this galaxy) only becomes possible if the radio transmitters are using more than 25 terawatts of energy. For comparison, nuclear power plants produce about 1 terawatt. Can you think of any reason to pump the output of 25 power plants into a single radio broadcast?

Narrow-band transmissions make detection possible at ~100 times greater ranges, but are unlikely to be pointed in the correct direction. Even if there's a narrow-band transmission pointing directly at us, we can only reasonably expect to detect it at a ranges of a thousand light-years or so, or about 1% of the diameter of this galaxy (and a considerably smaller fraction of the volume). That's ignoring the problems of what frequency to scan, and how to recognize a signal. (Can you distinguish compressed data from random static? I can't.)

In this case, the absence of evidence is not evidence of absence, because we're effectively blind. We're still discovering new stars within 100 light years. And it's likely that nobody knows we're here, either. The chances of there being an alien civilization with a receiver sensitive enough to detect our transmissions being in one of the nearest 2,000 star systems (max 80 light years away) is pretty low, even if there are thousands of such civilizations evenly distributed through our galaxy.

None of this addresses the problem of why we don't have evidence of expansive alien civilizations, though. It wouldn't (theoretically) be too difficult to station a robotic probe in every star system of the galaxy, so why aren't there a few alien star-probes here already? Maybe that, too, is harder than we think, but really, the only reasonable answer is "I don't know."

This post (or rant, if you insist) was triggered by this article.