Astronomical Mysteries #2: The Blinking Star
What exactly is going on with Epsilon Aurigae?
Epsilon Aurigae (call it EA for short) itself is a fairly normal F0-type giant star – okay, it’s a giant, so not very normal at all, but normal enough for its type of object. It has a temperature of about 7,000 K and it’s located something like 2000 LY away. So far so unexceptional.
Except for what happens every 27 years.
EA basically goes out.
Well, it’s not quite like throwing a switch. The star isn’t plunged completely into darkness. Rather, the brightness that we see from Earth drops by about 60%. It stays dim for between 640 and 730 days, before brightening back to where it was beforehand. Then, 27 years later, it repeats the performance.
Now, variable stars exist. Some of them vary because something’s changed physically inside the star – pulsating variables are one example, where the star expands and contracts. (That alters its radiating area and temperature, in turn altering its light.) Some variables, like Algol, vary because they are periodically eclipsed by a companion – a dimmer object ‘gets in the way’, essentially.
Thing is, none of these models really seem to fit EA.
Although the star dims, its spectrum doesn’t seem to change much during the dimming. This suggests that its surface temperature hasn’t changed, which would be inconsistent with a pulsating variable. This would tend to suggest that it’s dimming by being eclipse by something. But there are problems here too. First of all, eclipsing binaries are very, very regular – the Newtonian clockwork doing its thing. So why does the period of eclipse wobble back and forward between 640 and 730 days? That’s more than a 14% difference, which is quite significant. If there was a third body in the system, perturbing the second’s orbit, we could perhaps explain it – but why don’t we see the third body?
Also, where is the secondary’s spectrum? We’d expect to see features from it superimposed during the primary’s eclipse – instead all we see is the primary, dimmed. One model attempted to work around this by actually suggesting a transparent secondary – surely one of the weirdest ideas in stellar astronomy?
Also, there’s another big problem – right in the middle of its eclipse cycle, EA actually briefly gets brighter before fading again. That really makes no sense in terms of the eclipse hypothesis – the only way to make it work would be assume a secondary star with a hole in the middle!
We don’t see many* donut-shaped stars in the night sky, needless to say. It’s hard to conceive of how such an object could even exist – one would expect its own gravity to collapse it inwards, quickly filling any hole.
An alternative suggestion has been that the eclipsing secondary is not actually a star but actually a disk of gas, with a hole in the middle. This can resolve some of the above problems, but has issues of its own – like what exactly is holding the disk of gas together? A current hypothesis suggests a disk of gas with a B5-type star embedded in its middle, and with a hole, and oriented so that we don’t actually see the B5 star, and, and….
Yes. The more people look at this object, the more confusing it gets. I think the honest answer is that we currently just don’t know what’s going on with EA.
*And when I say ‘many’, I mean ‘any’.