A “partial” Dyson sphere would periodically eclipse or transit its host star, creating distinctive photometric signatures.  Geoff Marcy is looking for such signals in Kepler data, and my search for waste heat is looking for the glow of these mega-engineering projects.

A provocative and smile-inducing paper was published by Luc Arnold that investigated Kepler’s ability to find such things.  Arnold’s angle was actually at the border between artifact SETI and traditional communication SETI.  Arnold first pointed out that Kepler would be able to distinguish between spherical planets and other shapes — say, triangles — if the transit signal was strong enough.  He argued that artificial objects the size of planets placed in orbit around stars would have a low-baud-rate but be fantastically efficient — essentially you would get around the problem of how find the power to communicate over very large distances by using the star itself as your beacon.  It’s a brilliant idea:  you don’t need to collect the power of your star to use it to power big radio transmitters, you just need to block enough of it that other civilizations notice it flickering, and you’ve essentially got a one-solar-luminosity transmitter!

After calculating sensitivities to non-circular cross-sectioned planets, Arnold demonstrates that by building a huge sheet with moveable louvres, or by creating complicated and obviously non-natural swarms of objects one could vary the transit depth observed and send simple signals.  For instance, prime numbers:

Or, using the louvres:

Clearly, the message here is that a “cheap” way for aliens to attempt communication would be to build these devices that might last for a very long time with little maintenance and send signals visible across the Galaxy.  We should always be on the lookout, it follows, for transits from planet-sized objects with highly variable depths.  After all, it would be very difficult to come up with any natural explanation for such a phenomenon (*ahem*).

Which brings me to KIC 12557548, a bizarre star in the Kepler field of view with an apparently transiting object whose eclipse depths vary by a factor of six (sound familiar?):  

The authors go to valiant lengths to find a natural explanation for this, and come up with a Mercury-sized object that is evaporating, and the cometary cloud behind it has turbulence that generates highly variable opacity to the starlight.  

Now, I don’t know what this is.  Maybe it really is an evaporating planet (the best guess, I’d say).  Maybe it’s KH15D all over again, but with some high-frequency twist of some kind.  I’d bet my house on it not being aliens.

But given that he went way out on a limb and predicted almost exactly this sort of thing, don’t you think Luc Arnold at least deserved a citation?

(To be fair, I should be generous and assume that the authors weren’t aware of Arnold’s paper.  But I’m not at all sure they would have cited it if they had.  And, who knows?, perhaps the referee wouldn’t have taken it seriously if they had.)