I am partial to gamma-ray bursts.
First, they are simply too cool: the titanic explosion of a supermassive star triggers twin beams of gamma rays that contain as much energy as the Sun will generate over its entire lifetime. These cosmic blowtorches chew their way out of the exploding star and tear across the Universe, and, in the first few thousands light years, destroy anything they touch. Eventually, weakened by distance, we see them only as a sudden flash of gamma rays (extremely high-energy light) which fades rapidly, never to be repeated.
Second, I worked for several years on the education program for the Swift satellite, which targeted GRBs. For quite some time I kept track of every burst seen, and how they behaved.
GRB 070125 -- a burst seen on January 25, 2007 -- was unusual. That was obvious even when I wrote about it at the time: it was intense, long-duration, but had no features in its spectrum. Usually, a spectrum taken of a burst shows some features that allow astronomers to get a distance, or determine various other characteristics of the burst. But this one was peculiarly empty of such things. Eventually, a faint feature was found that allowed a distance determination: at least 9.4 billion light years, about average for GRBs, but still a mind-numbing distance.
But the big surprise was to follow. Observations made after the glow of the burst had faded revealed no underlying galaxy! This is extremely weird. Massive stars which trigger GRB explosions die when they are still young, having consumed their fuel at a furious rate. Therefore we expect to see these types of bursts happening not only in galaxies, but specifically in locations in galaxies where stars are being born.
But GRB 070125 is an orphan. Deep images of the region where it went off yield no galaxy. There are two somewhat nearby, but they are at least 80,000 and 150,000 light years way, far too distant for either to be the host galaxy.
Cool! A mystery!
Astronomers speculate that it's possible that the two galaxies are interacting, colliding. When that happens, long streamers of stars and gas get drawn out of the galaxies, and stars do tend to form there. If this is what happened, then GRB 070125 isn't an orphan, it was just born in an exotic locale; if so the "tidal stream", as it's called, is too faint to be seen in the images. Even deeper images are being planned to see if any tendril of stellar stuff can be seen. If so, then the mystery is solved. If not, well, then, GRBs have a history of being odd, of throwing monkey wrenches into our ideas, and of revolutionizing the way astronomers perceive the Universe.
In that way, at least, GRB 070125 is fairly run-of-the-mill. But that's only because of the extraordinary company it keeps.
Oh, and if such things concern you, don't panic: the odds of a GRB going off in our Galaxy and damaging the Earth are extremely remote. I'll be talking about this quite a bit more on the blog as the time of my book publishing approaches. I have a whole chapter on GRBs; the damage they can do when nearby is terrifying, but happily for us there are no progenitors lurking nearby. This is one cosmic danger that we're relatively safe from.