Its' just a transformer blowing up.
First recorded incident was in the year 869
I hear there's a recorded incident of a dead man who got buried in a cave and then resurrected and walked out of the cave some time later.
It has the same intensity of a transformer shorting out and the color is very similar but it also looks like local lightning strikes along a line of storms.
As a geophysicist, I am in the crowd that believes that earthquake lights are a real phenomena related to the earthquake and they may eventually be utilized as a predictive tool. The technology is not mature enough to be able to understand their method of generation and the sensor network that would be necessary to use them as a predictive tool is too sparsely distributed.
I wonder whether every earthquake has a spectral signature but the only ones we are able to see directly are those that dip deeply enough into the blue end of the spectrum to fall within the limits of our own vision.
Perhaps a UV detector array deployed in an active region could help us understand whether we miss these signals because we can't see them unless the quake is powerful enough to cause a wide-bandwidth event that bleeds into the visible spectrum. This may also help explain why some animals that can see UV or spectral components outside our own visual limits are disturbed before earthquakes. Maybe they can see these signals and the combination of seeing these signals and feeling ground motion precursors could trigger the animal behavior that has been observed globally.
I think this is an interesting problem that requires further study.
An epicentre is just a human generated approximate coordinate - it has little to with the physical quake itself.
I'm not sure why you think the epicenter is unimportant in describing the "physical quake". To the geologist it helps establish rock types present in the rupture zone based on their knowledge of tectonic history and stratigraphy of the region. The characteristics of the overburden also help to understand how each wave mode in the full wavefield of the earthquake with be manifest at the surface as the energy propagates from the source (at the epicenter). The initial estimation of the magnitude of the earthquake may be revised as data from more sensors is factored in to the location determination, that is true. That does not mean that it is just a wild guess or that it is poorly constrained.
> as the energy propagates from the source (at the epicenter)
That just isn't how earthquakes work.
Let's assume a slip along a plane cutting through the ground. The earthquake waves are generated at all parts of the surface slip (all parts of the plane). At different times as the rupture propogates. There is no center - that is merely a simplification.
One feature was the asynchronicity - one of the variables had to be lagged, I can't remember which one.
Not sure what to make it but possibly related to lights in the sky.