So What Are We Looking For?

It's impossible to predict how bright the next nearby supernova will be or how long we will have to wait for it to pop off. But we can get a feel for the answers by looking at the questions in several different ways.

As a starting point, we can create a list, largely from Oriental and Arabic records, of supernovae that have been seen during the last two millenniums.

Two nearby supernovae that should have shone brightly but apparently didn't are omitted from the preceding list. SN 1680?, also called Cassiopeia A, is one of the strongest radio sources in the sky and was probably glimpsed by John Flamsteed at 6th magnitude. As shown in the images below, extensive dimming of its light by interstellar gas seems very unlikely. However, according to Thomas Dame (Harvard-Smithsonian Center for Astrophysics), one "can't rule out the possibility that the supernova went off behind a very small, very dense clump of gas."

The recently discovered X-ray remnant of SN 1320± lies at a distance of only about 650 light-years, making it the closest known supernova to Earth; its light could have equaled that of the full Moon! So why wasn't it seen? Again, extinction by interstellar gas and dust is "extremely unlikely," says Dame. Perhaps both SN 1320± and SN 1680? mark a hitherto unknown class of supernovae that are optical duds (S&T: April 1999, page 22).

All the entries in the list predate the invention of the telescope: seven conspicuous naked-eye supernovae in 1,400 years, or one every couple of centuries, on average. So why haven't we had another in 400 years? Whether because of bad statistics or bad luck, it seems we're overdue by a factor of two.

Determining how often supernovae explode in our Milky Way is fraught with uncertainties, the estimate being confounded particularly by the gas and dust that pervade the galactic plane. The rate can be judged in many ways, but all involve surrogate evidence or initial assumptions that are subject to observational bias. These methods include our galaxy's inventory of heavyweight stars (which blow up 10 million years or so after being born); the number of pulsars (spinning neutron stars, the progeny of supernovae); counts of expanding, wreathlike supernova remnants; and the determination of supernova rates in galaxies kindred to our own.

The Milky Way's supernova rate was estimated in 1994 by Richard G. Strom (Netherlands Foundation for Research in Astronomy). By comparing supernovae observed over the past two millenniums with supernova remnants of comparable age, he concludes that a star blows up near the Sun (within 5 kiloparsecs [kpc] or 16,000 light-years) every 175 years, on average. By extrapolating this rate to the whole galaxy, Strom predicts a supernova every 20 years or so.

On the other hand, a team from the University of Western Australia published a paper in 1999 that joins evidence from extragalactic sightings, stars in our galaxy, and the historical record of supernova explosions within 4 to 5 kpc (13,000 to 16,000 light-years) of the Sun. According to coinvestigator Ronald Burman, "One cannot reliably extrapolate from the rate of historical supernovae to obtain a rate for the galaxy as a whole, since we appear to live in a region of the galaxy with an enhanced event rate." Such would be the case if we were located adjacent to active star-forming regions, where supernova progenitors are most likely to be born. The bad news is that this team finds the most likely rate for Milky Way supernovae to be only about two per century. The good news is that the vast bulk of these dying stars will spit out neutrinos.

So how bright might the next Milky Way supernova be? In 1975 Sidney van den Bergh (now at Dominion Astrophysical Observatory) made a careful estimate. In preparing this article, I did my own calculation, using somewhat different rules, and got similar answers. So I combined both results in the following table.

If this distribution is accurate, it implies that pretelescopic observers logged only a third of the supernovae that exploded in our galaxy. Historians have pointed out that a new star had to be really bright, perhaps exceeding magnitude +1.5, to stand a good chance of being noticed by ancient astronomers.

The fact that no galactic supernova has been recognized on 20th-century sky-patrol photographs seems to confirm the robustness of the table. Combined with the best-guess frequency of supernovae, it suggests that only three or four supernovae would have brightened enough to be recorded on patrol plates. And even if a maverick image had been spotted, it would likely have been dismissed as a "Kodak comet" or other defect.

Maybe those plates are worth checking again by someone armed with modern radio, X-ray, and other ledgers of supernova suspects. As van den Bergh wrote: "Very red 'novae' that exhibit a relatively slow rate of brightness decline are prime supernova suspects."

What's the chance that a supernova will jolt the neutrino detectors in the coming year? Odds of about 1 in 30 would probably satisfy Las Vegas bookmakers. The chance that the stellar fireworks will actually be seen drops to about 1 in 70, in my opinion. So, if you want instant gratification, you had better look elsewhere. But what intrigues me is that a champagne cork could pop tomorrow!