Earthquake lightsflickerings and flashes that occasionally accompany earthquakeshave always been a stepchild topic among seismologists, like any subject that relies on human reports and perceptions. Therefore it's a notable occasion when a journal publishes a paper on this elusive phenomenon, as Seismological Research Letters did this month (see the introduction here, but it's not informative). I haven't read the paper yet, but I have a long familiarity with the subject (coauthor John Derr was once my supervisor around 1980) and wrote it up here in 2006. The journal summarizes the paper in this press release, and it's gotten a buzz among those who are new to the subject (follow "earthquake lights" on Google News).
The study itself is not a breakthrough in any way. It's a rigorous catalog of well-documented sightings of earthquake lights, along with a tentative conclusion that the lights appear to favor vertically oriented faults. This is something we might expect from pure geometry, as the source of the lights' energy would naturally be nearest to the ground in a vertical fault. But science can only advance by first testing "what we might expect," because often nature fools us. So far so good.
To me, the significance of this paper lies in its publication, period. As more and more cameras (both surveillance and smartphone cams) record earthquake lights, it's essential to have a working hypothesis in place as well as a good catalog. That should be familiar from other scientific quests, like earthquakes themselves (an ongoing success) and UFOs (an ongoing enigma).