|Artist's impression of a debris disc in orbit around a white dwarf. Over the past eight years we have seen such a disc suddenly form, then rapidly disappear around the white dwarf SDSS J1617+1620. Image Credit: NASA, ESA, STScI, and G. Bacon.|
(Note: Where I use "we " below I'm generally referring to all of the authors of the paper. I didn't actually start working on this project until January, so other people have done a lot of work on this before me.)
History of a dying star
Around four hundred million years ago, SDSS J1670+1620 was a star similar to our Sun (although with 2-3 times the mass). As it reached the end of the hydrogen fuel supply in its core, it swelled up into a red giant, then blew away its outer layers to become the tiny white dwarf that we see today.
Just like the Sun, SDSS J1670+1620 hosted a system of planets, asteroids and other objects. As the star underwent the turbulent transformation into a white dwarf its inner planets were destroyed, tumbling into the red giant.
The rest of the system may have survived relatively unscathed. The remaining planets would have started moving out into wider orbits as SDSS J1617+1620 shrank down to a tiny ball of carbon and oxygen the size of the Earth, with only around a third of its original mass.
Systems like SDSS J1617+1620 are referred to as evolved planetary systems. Over the past couple of decades astronomers have discovered dozens of systems like this, and we now know that they exist at around at least half of all white dwarfs.
Around 33 (and counting) white dwarfs we've detected rings of dust. These structures, as wide across as the Sun but only a few millimetres thick, are formed by asteroids thrown in towards the white dwarf by distant planets. When they get too close the gravity of the dead star overcomes the forces bonding the rock of the asteroids together, ripping them apart into a dusty disc.
Such a disc was observed around SDSS J1617+1620 in 2012. But as well as a dust disc, we also saw signs of a much rarer phenomenon: Another disc, this one made of gas.
How to find a gas disc
The way we find these gaseous discs is, in my opinion, rather cool- another one of those places where a seemingly complicated plot clearly shows what's actually there.
Each element in the object emitting the light leaves its mark on the spectrum, emitting or absorbing light at very specific wavelengths. When we look at the emission from calcium, we see something odd. Instead of a sharp line, the emission is smeared out into as distinctive double-peak, with one peak on each side of where the line should be.
This is a result of the Doppler Effect. If an object emitting light is moving towards us, the light waves get bunched up, resulting in a lower wavelength. The light appears to be slightly more blue than expected. As it moves away, the light is stretched out and the wavelength increases, becoming more red. The Doppler Effect has many applications, most famously being used by Edwin Hubble to show that the universe is expanding.
With this in mind we can see that the double-peak emission must be coming from a disc around the white dwarf. As the disc rotates, the side of it moving away from us emits red-shifted light, and the side moving towards us emits blue-shifted light.
Not only can we tell that there's a disc, but we can also measure its size and location. The amount that the light is shifted by depends on the speed of the material in the disc, which depends in turn on how far away the gas is from the white dwarf. Measuring the wavelengths of the inner and outer extent of the peaks can therefore tell us the exact position of the disc.
Doing this measurement has had a surprising result. Including this one, we've seen gaseous discs around just seven white dwarfs. They should have been easy to explain: As the dust gets closer into the white dwarf, the heat should sublimate it into gas.
But when the positions of the gas discs were measured, they were all found to be too far away from the white dwarf for this to happen. The question is then: if it's not the dust sublimating, how did they form? And why do we see them at only a small number of white dwarfs?
A key missing piece of evidence is variability. As the discs are formed by asteroids, which are regularly scattered in towards the white dwarf, we know that these are highly active systems. But despite being observed over several years or even decades, none of the gas or dust discs have shown any changes with time.
Slide show of some of the spectra of SDSS J1617+1620. In the space of two years double-peaked emission lines revealing a gaseous disc appeared out of nowhere. A couple of years later the emission had decreased dramatically, and by last year the disc had completely disappeared.
After discovering the first gaseous disc around a white dwarf in 2006, a search was made in the Sloan Digital Sky Survey for more white dwarfs with double-peaked calcium emission lines.
Two observations had been made of SDSS J1617+1620 one in 2006 and later in 2008. The later spectrum showed clear evidence for a gaseous disc, a huge structure as wide across as the Sun. However, in the earlier spectrum the disc was nowhere to be seen.
Intrigued by this clear, unprecedented evidence for variability, we started observing the white dwarf with ever more powerful telescopes. To our surprise, the gaseous disc dispersed almost as quickly as it has formed. Just a year after the observations first showing the disc the emission from the disc had fallen by more than half. The gas disc continued to dissipate, until by 2013 there was nothing left.
How did this happen? The video above shows one possibility. We know that SDSS J1617+1620 is orbited by a dust debris disc, formed form the shattered remains of an asteroid. A second asteroid, thrown in from the outer reaches of the system by an orbiting planet, could have impacted in this disc, creating a short-lived burst of gas.
A more intriguing possibility is that we're actually seeing the formation of the dust disc itself. Depending on how close it gets to the white dwarf, an asteroid being pulled apart by gravity might not turn entirely to dust in one go. It might take several orbits, passing close to the white dwarf and loosing a little more mass to the forming disc each time before moving away again.
Each time the debris of the doomed asteroid returned it would interact with the growing dust disc, forming a gaseous layer like the one we're seen. The intriguing possibility here is that the asteroid may return again in a few years time, perhaps creating another gas disc. This could provide a key insight into the dynamics of evolved planetary systems, and we plan to keep observing this star to find out if the gas disc returns.
However it formed, we also need to try and explain how the gas disc disappeared so quickly. Here we have a helping hand, as there has been plenty of theory work done studying the behaviour of gaseous and dusty discs.
Each particle of dust orbits the white dwarf independently, interacting very rarely with the rest of the dust in the disc. The gas, however, behaves more like one big object, with each particle mixing and colliding with those around it. This has the effect of slowing the gas (and dust) down, causing it to spiral inwards and fall onto the white dwarf. The more gas, the faster the accretion.
This accretion is easily visible as metal-pollution in the otherwise pure hydrogen atmosphere of SDSS J1617+1620. We should be able to test this theory, seeing if the accretion rate changes in time with the variability of the gas disc
However when we measured the accretion rates, they stayed the same. This could mean that our ideas about the behaviour of the gas discs are wrong, but its more likely that there was too little gas to cause a noticeable effect- this would also explain the short lifetime of the disc. Alternatively there could be a delay between the formation of the disc and the change in accretion rate as the disturbed material moves inward. More reason the keep watching!
As we were finishing off this paper, another paper was published showing variability at another white dwarf- but this time it was the dusty part of the disc changing, not the gas. With clear evidence that both types of disc at white dwarfs can change, we may be able to begin to answer some of the outstanding questions about these intriguing systems.
Studying these systems also provides an insight into the future of our own Solar System. Five billion years from now, the Sun will begin the transformation into a white dwarf. The more we learn about evolved planetary systems, the better we'll understand what happens next.
The full paper is available here. I haven't gone into all of the details about what we've discovered at SDSS J1617+1620 as that would be a really long post, but if you want to find out more feel free to ask me any questions in the comments or on Twitter.