Circumstellar Disk Disappears

[Enthalpy]

A surprise to astronomers...

 

The young star TYC 8241 2652 is considered to have a circumstellar disk that radiates the observed thermal infrared. This IR radiation uses to vary moderately and slowly, but not so TYC 8241 2652 : between 1983 and 2009 it fluctuated by a factor of 3, but then it dropped by a factor of 10 within two years though the star remained quiet.

 

A general-press paper: http://www.gemini.edu/node/11836

 

The science paper (30€): http://www.nature.co...re11210.html#t2

 

Free excerpts: one figure, two tables and a rationale that dismisses usual explanations

http://www.nature.co...re11210_F1.html

http://www.nature.co...re11210_T1.html

http://www.nature.co...re11210_T2.html

http://www.nature.co...ure11210-s1.pdf

Not a coalescence, not a collapse, not blown away nor molten.

 

-----------------

 

I'd like to suggest a different scenario.

 

The disk has not changed in the event, but its heat source did, which is not mainly the star.

 

A cloud of gas impinging on the disk (or maybe very fine dust) could heat it and explain the previous luminosity. When this source disappears, the disk cools down quickly.

 

A local gas cloud may not be exceptional in a stellar system, even old. I feel it's an explanation to our comet 17P/Holmes, which increased its luminosity by 15 magnitudes without breaking apart.

http://en.wikipedia....ki/Comet_Holmes

 

The disk, cooler under normal circumstances, would only need to reside farther from the star. I suppose its diameter is not observed: the authors suggest 0.4 AU diameter at a distance of 456 L-y, converting to 0,0015" apparent radius. A 2m telescope observing at 10µm has its first diffraction zero at 2*1.2" and a 10m telescope observing at 1µm has them at 2*0.025".

 

Marc Schaefer, aka Enthalpy

[D H]

Arxiv preprint: http://arxiv.org/abs/1207.1162

[Enthalpy]

The necessary amount of gas can't be a comet tail, but a mass equivalent to a planet may fit.

 

As the disk was active, it radiated 0.3 Jy (0.3*10^-26 W/m2/Hz) between roughly 6 µm and 20 µm (or 35 THz) at a distance of 456 L-y (4.3*10¹⁸ m, surface 2.3*10³⁸ m²), or a power of 2.4*10²⁵ W.

 

This power was observed for 25 years; let's imagine it lasted for 40 years, then the emitted energy equals 3.0*10³⁴ J.

 

As a circumstellar disk is an early phase of a planetary system, I imagine many objects are still out of the equatorial plane, even massive ones and rather near to the star.

 

Let's say the heavy disk had 30km/s where the collision happened, and the gas 30km/s as well but 30° out-of-plane. Then, it takes 2.7*10²⁶ kg of gas to bring the kinetic energy that heats the disk: that's half a Saturn - remember planets still aren't formed, so these amounts of gas are available.

 

But if the gas orbits at 30km/s and 90° out-of-plane, then 3.3*10²⁵ kg suffice, or half a Uranus.

 

And if the gas had 15km/s above the star's gravitation well and 180° inclination, then 1.1*10²⁵ kg suffice or two Earths.

 

Such an event wouldn't be anecdotal. It would be an important step in the formation of a planetary system - a step needed, by the way. Which implies it wouldn't happen often - not necessarily a bad thing, since many circumstellar disks are observed and TYC 8241 2652 is the first showing this behaviour.

 

Marc Schaefer, aka Enthalpy

 

Arxiv preprint: http://arxiv.org/abs/1207.1162

Thanks D H! From that paper:

 

- The disk emitted 11% as much infrared power as its star over the total spectrum. This is more than usual, with only V488 Per exceeding it;

- It emitted at a wavelength around 11µm when it was active, but 22µm now.

Both are compatible with a disk heated by collision.

Should they observe V488 Per over time?

[Enthalpy]

Trying to check how probable it was to observe such a collision...

 

From a list of 900 stars which had a debris disk in 2002:

http://www.roe.ac.uk...tification.html

I assume that 1000 stars have a circumstellar disk which is monitored every few years, a first bold assumption.

 

Let's say this phase of the star evolution lasts for 10 million years.

Then, I suppose that (no reason for it!) 30 collisions between gas and the circumstellar disk happen and end in that time, or one in 300,000 years.

 

The 1000 stars would let one such event end every 300 years.

If precise infra-red surveys have been made over the last 30 years, astronomers had 1 chance in 10 to see one collision end.

 

Similarly, if each collision lasts for 30 years, astronomers had 1 chance in 10 to see one in action (V488 Per).

 

Not bad. So many figures here are unknown and can be twisted that the desired probability is easily cheated.

Edited July 13, 2012 by Enthalpy