In the name of getting on to more, um, current topics, here’s my summary of the last day of the Caltech Fireworks Meeting:
Day 4 was devoted to everything that didn’t fit into the first three days. A big chunk was devoted to the theory and observations of “overluminous” supernovae, much too bright to fit comfortably into one of the existing channels and spectroscopically also unique. There was also some attention given at the end of the day to weird-looking underluminous, short-timescale transients.
There are two conceptual classes of peculiar overluminous SNe, one driven by theory and the other by observation. On the one hand, the desire to understand the first stars and how they lived and died drives the search for pair-instability SNe (PISNe), the explosions of extremely massive stars — between 150 and 250 times as massive as our Sun. These are in general predicted to be extremely bright. On the other hand, new wide-area transient searches have been uncovering an increasing, though still small, number of rare supernovae of (literally) unprecedented luminosity. The Texas Supernova Search, which formed the UT-Austin Ph.D. thesis of Dr. Robert Quimby, continually set new records for the most luminous SN ever discovered. The resulting list of events, and newly discovered events that resemble them spectroscopically, are now unofficially (but widely) referred to as “Quimbies”.
PISNe: explosions of the biggest stars EVER
In general, thermonuclear SN explosions are expected to arise from white dwarfs, the slowly-cooling corpses of stars like our Sun, set into the rigor mortis of electron degeneracy. These stars never had enough mass to fuse the nuclei of atoms in their cores to elements heavier than oxygen. And they usually take a long time to die — at least a couple hundred million years, so dinosaurs were still contentedly roaming the earth when the stars which were to become the youngest of white dwarfs first formed. Core-collapse SNe, in contrast, arise from younger stars as massive as a few tens of stars like our Sun; these stars live fast and die young, after only a few millions or tens of millions of years. These stars are sufficiently massive that the pressures in their cores become high enough to produce elements as high up the periodic table as iron before those cores collapse.
Pair-instability supernovae are another class of thing entirely. In the carbon/oxygen core of a star more than 150 times as massive as our Sun, the temperature — higher than 10 billion degrees Kelvin — is equivalent (first law of thermodynamics, “heat is energy”) to the rest mass of a pair of electrons. Well, applying Einstein’s E = mc2 and a little bit of quantum mechanics, this means that pairs of electrons and anti-electrons (or positrons) are spontaneously created, using up the heat energy in the process.
The problem with this is of course that the pressure from hot gas is what’s holding up the star! Once the creation of electron-positron pairs starts sucking up all the heat into the star’s core, the gas loses pressure and has a harder time keeping the star from collapsing. What’s more, the process tends to reinforce itself and run out of control: rather than cooling off, the core contracts, becomes hotter, creates more pairs, etc. You might think the end result would be collapse to a black hole, much as in a typical CC SNe. However, what tends to happen instead is that the core eventually becomes hot enough to explosively ignite the carbon and oxygen in the star’s core, and the entire star blows itself apart. So you have a thermonuclear explosion, but of a super-massive star, not a white dwarf.
The resulting event looks like a weird cross between a SN Ia and a CC SN. The underlying thermonuclear nature of the event creates silicon in the atmosphere, much as in a SN Ia. However, the star’s outer layers haven’t yet been blown off, so the expanding SN blast wave will dump energy into those layers and cause them to glow — increasing the luminosity and producing telltale signatures of hydrogen and/or helium in the SN’s spectrum, as in a SN II. The events are usually very long-lived, and depending on how much material is burned to nickel, can be very bright. The most convincing candidate observed to date is SN 2007bi (see astro-ph link here), although there are still some detractors who believe it’s possible to explain that event as a CC SN instead.
The theoretical framework surrounding the explosion itself is robust, and we believe PISNe must surely happen out there somewhere. But the original motivation for looking for these stars was the belief that the only stars which could be so large would be stars made initially of only hydrogen and helium, with negligible amounts of carbon, oxygen, etc. The reason for this is that carbon, oxygen, etc. are chemically complex atoms which can emit and absorb light at many wavelengths; quantum mechanics ensures that hydrogen and helium have comparatively limited options for interacting with light. This has two effects: “Metal-free” hydrogen/helium gas should have a hard time cooling off by emitting radiation, meaning it should form monolithic dense clumps instead of many smaller clumps which cool and contract quickly to form stars of more moderate mass. At the same time, stars produce radiation that can push material with heavy elements mixed in outwards away from the star. In a “metal-free” star, the outer layers don’t experience as much pressure from radiation, so an initially massive star will tend to stay massive and eventually be subject to the pair-instability mechanism.
Theory has gotten better since then, and we were reminded of this: Modern simulations of metal-free stars show that there are other cooling mechanisms involving deuterium which allow clouds of primordial hydrogen/helium gas to fragment and form small stars. It turns out that although some large stars certainly will form in primordial hydrogen/helium clouds, the big monolithic clumps in earlier simulations stuck together because the resolution in those simulations wasn’t good enough to represent the processes underlying the fragmentation — sound familiar? Similarly, some doubt has been cast on the perceived necessity that very large stars be “metal-free”; though they should have significantly less heavy-element content than our Sun, their compositions may not be as extreme as we used to think. Finally, recent observations of clouds of gas in the early universe (see astro-ph link here) show that they have a chemical composition consistent with that expected from a primordial hydrogen/helium cloud “polluted” by only a handful of SNe. The composition of these clouds appears to favor enrichment by CC SNe, suggesting that the role played by PISNe is probably much smaller than we used to think it was.
Quimbies: big, bright, and ???
Although the race may still be on to unambiguously identify a PISN in nature, there are plenty of really luminous explosions out there which we don’t really know how to explain. Various mechanisms have been trotted out, none of which are entirely satisfactory: They are very bright at first, but they fade away faster than SNe Ia, so they can’t be powered by the same power source (radioactive decay of 56Ni created in the explosion). They don’t seem to have traces of hydrogen in their spectra, meaning they probably aren’t powered by the glow from hydrogen gas which has been heated by the SN blast wave. (Some colleagues of mine suggest that you could probably heat up a lot of helium gas and cause it to glow without emitting helium lines, but they aren’t willing to commit to an answer to the question of how much.)
These SNe, like many other weird-looking SNe, happen preferentially in very small galaxies which are forming stars at a ridiculous rate. The most successful supernova searches up until only a few years ago were targeted searches, which looked preferentially at a few bright galaxies waiting for one or more of their stars to explode. While these searches turned up a lot of SNe, one might worry that they would miss SNe going off in “anonymous” faint galaxies which hadn’t appeared in any catalog before — as, indeed, we find more and more to be the case.
Other unexplained, mostly short-timescale transients
Near the end of the day, everyone got their chance to come forward and talk about their favorite weird-looking transient. The largest number seemed to come from the PTF survey. Some of these were difficult to identify because their spectra bore passing resemblance to any one of a number of totally different physical things, making it difficult to figure out exactly where to put them. But many remained unidentified simply because they were short, and were gone before adequate follow-up could be obtained. It’s hard to know yet how many of these things are out there, although more will be found the deeper we look — as good an argument for a helium-cooled space pony as any I could think of.
While I’d like to describe the weirdnesses of these transients in more detail, time permitting, the experts are themselves very much divided and so they don’t fall into a tidy story lending itself well to blog posts. So I’m going to have to forbear, except to say that even the experts have a lot to learn about what’s out there. In words popularly ascribed to Einstein, “if we knew what we were doing, it wouldn’t be called ‘research’.”