Surprise! Bet you didn’t know that.
In 1999, Annis performed a simple “search” for alien life. He hypothesized (probably correctly) that a type III Kardashev civilization, one that controls power comparable to an entire galaxy, would be an outlier on R-I-T relations, where R is the size of the galaxy, I is the intensity, and T is the temperature. Both spiral and elliptical galaxies have such relations, Tully-Fisher for spiral and the fundamental plane for elliptical galaxies. He also theorized that such galaxies would have an excess of IR radiation and would have low surface brightness. Using data on 106 elliptical galaxies and 57 spiral galaxies taken more than a decade before he published his paper, Annis looked for any outliers in the above relations. He defined his outliers to be objects more than 1.5 magnitudes dimmer than the emission predicted by the respective relation. This seems fairly safe, and like a sound argument for detecting life. Personally, I don’t see any civilization becoming a type III civilization, so I consider this and related searches to be a decent waste of time (for SETI; some of the searches lead to important science).
Not surprisingly, Annis found no objects that met his outlier criteria. I say this isn’t surprising for two reasons. First, I feel that a galaxy with anomalously low emission would have been flagged prior to this search given how long the data were available. Second, this search was of so few galaxies. (A weak third argument would simply be that a civilization capable of eating up energy on the order of their galaxy would have ways of compensating for this fact, like Kipping 2015 suggested covering transits with lasers.)
This search to me just seems a bit useless and like he had a free afternoon and wanted to just quickly write up and publish *something*. Even if he were to find a galaxy with too low emission, it would require additional data in some other field to confirm it had anything to do with ETs. That being said, I would be interested in seeing someone do this with all the data we have nowadays.