Dar is at the Technion and De Rujula at CERN. Their paper looks at what happens when a gamma ray burst (GRB) goes off in our galaxy. A GRB, visible somewhere in the cosmos about 3 times every day, is thought to be either the dying blast of a high mass supernova, in which the core collapses to either a neutron star or a black hole or the merger of two neutron stars. In either case, the combination of high angular momentum, trapped magnetic field and extreme kinetic energies - comparable to the entire rest mass of the star or stars - results in a few-second long blast of energy beamed along the opposite poles of the rotating system. We see a fraction of this energy as gamma ray photons. These two beams are tightly collimated by a combination of magnetic fields and relativistic effects, with opening angles thought to be no more than a few milliradians. That's how we can see them all the way across the universe -- an appreciable fraction of the supernova's total energy goes into these tightly focused beams. When one of them happens to point in our direction, it's bright enough to outshine everything else in the sky at gamma ray energies, even from billions of light years away.
If one of these things goes off in our own galaxy, though, we could be in trouble. Anything sufficiently close - and the center of the galaxy, 25,000 light years away, is apparently, close enough -- gets blasted with radiation. For us, it's not the direct gamma rays that are the problem. Our atmosphere blocks those pretty effectively. GRBs weren't even noticed until the 1960s, when, during the cold war, the US launched satellites designed to detect nuclear bomb blasts. (Above ground tests were banned by treaty in the early '60s.) Being above the atmosphere, they could see the GRB gamma rays.
However, once the gammas (and any other cosmic ray stuff -- protons and heavier nuclei swept up in the blast) hit the upper atmosphere, their interactions with atmospheric atoms produce a copious shower of high energy particles. The result of this air shower is a flood of high energy muons. Muons are heavy enough to zip by atomic electrons without being significantly scattered or deflected, not affected by the strong interaction, making them highly penetrating, and sufficiently long-lived for most of those created in the upper atmosphere to reach the surface. In fact, several (created by ordinary cosmic rays) go through you every second as you read this. They pass easily through the lower atmosphere and hundreds of meters of water or rock, and deposit all of their energy along the way. This puts them in a really nasty "sweet spot" of particle properties to cause trouble. Protons, which are heavier, interact strongly with atomic nuclei and lose energy before reaching the ground. Electrons are so light that they get scattered before going very far. Neutrinos just zip right through everything without losing any energy. Other particles like pions (aside from being strongly interacting) are too short-lived to reach the surface.
The net effect from the muons is a ~10 second zap of high energy radiation deposited in anything on the earth's surface or in the shallower parts of the ocean. According to the paper, a GRB at the distance of the center of the galaxy would, if pointed right at us, have enough energy to sterilize the planet's surface. Thanks to associated cosmic rays, which would follow the initial gamma pulse for a few days, no part of the planet would be safe. Dar and De Rujula calculate that a person at sea level would receive hundreds of times the fatal dose of radiation. Only the hardiest, smallest or deepest ocean life forms would have a chance of survival. Also, just for extra fun, there'd be a global firestorm from the thermal pulse in the upper atmosphere, like in the asteroid strike scenario.
Because the gamma burst comes from a supernova, it should also be accompanied by a neutrino pulse, and these might precede the gammas by seconds or minutes, because they are produced early in the collapse and escape almost immediately. The neutrinos would be imperceptible except to a few large experiments like IceCube. It seems unlikely that, even if correctly interpreted, the warning would help anyone very much. The only place to ride out the initial radiation blast would be deep underground. And the surface that anyone who happened to be down in one of these mines (such as physics grad students minding an experiment) would return to would be sterilized by the blast and the firestorm. Life would survive in the deep oceans and also among smaller, hardier critters (like these) that are able to withstand and repair cellular damage.
The paper's authors calculate an expected rate for getting caught by one of these blasts at roughly once per 100 million years. The nearest star likely to undergo this kind of supernova is Eta Carinae, but apparently its spin axis isn't pointing at us, though it's expected to blow sometime in the next million years or so. There's been some speculation that one or more of the handful of mass extinctions seen in the fossil record was due to one of these events, with arguments recently made for one about 450 million years ago as a candidate. (Oddly, while there are several later papers on the same general subject, no one seems to reference this paper. I'm not sure if that means it's wrong or if this is some petty turf thing.)
Anyway, this is another interesting way for the world to end, up there with the solar system wandering into a black hole or metastable vacuum decay. Always fun to think about!
UPDATE: A new paper suggesting that GRBs make most galaxies inhospitable to complex life. http://physics.aps.org/articles/v7/124. Doesn't mention the muon shower effect though. In fact I haven't found a later paper that references the one at the top. Maybe there's something wrong with the calculations, but de Rujula is a big name, so I doubt it's egregiously wrong.
