Unruly Milky Way

empaler writes "Space.com is running a story about the movement history of the local group near our solar system. The belief until now has been that after an initial period of chaos in our galaxy, it had since 'been rather calm'. 'But this turns out not to be true. Stars have been perturbed all the time throughout the Milky Way history.'"

Comments

Mmm...

[nitrocloud]

Milky Way, mmm. Candy.

Plotting Ahead?

[glen604]

Since they seem to be able to trace the route of the Sun (and other stars) going back several million years, I would imagine they could also likewise figure out where everything is going (to a certain extent).. Sooo..I wonder if any stars are going to come closer to the Sun- ie within a visitable timeframe.

ok, yeah, one can always hope :)

Re:Plotting Ahead?

[Necro Spork]

Stars are so small and so far apart that they almost never come close to hitting. Main sequence stars are roughly separated by one parsec (3.26 light-years). You can think of it like having one golf ball ten miles away from you and hitting another and asking if they are likely to collide. Galaxies do collide often and interact because they are close together compared to their width.

Re:Plotting Ahead?

[Anonymous Coward]

And yet, oddly, people still occasionally get a hole in one at the golf course.

Stars do collide - I'm sure quite frequently given the number of them. It's just a question of how likely it is to happen to any one particular star - ie: ours.

Re:Plotting Ahead?

[Anonymous Coward]

In the cases of both stars or galaxies apporaching each other to the point of *collision*, there is almost always some orbital component, in which case angular momentum is conserved as the get closer.
If the universe were nice and tidy, even the impressive accelerations seen there would be quite predictable. But complicating matters are the realities of *open systems* such as binary - or other multiple star - systems that *shed* mass from one star to a companion, or instabilities that may result in a star assuming a new size or perhaps *blowing up*.

One wonders where the initial assumption came from: the belief until now has been that after an initial period of chaos in our galaxy, it had since 'been rather calm'.
It seems it's just a slightly modified version of the geocentric view of the universe, in the sense that *that* view was hard to shake because it threatened not only the religious views but the general sensibilities of many people. If it *looked* like things moved around the Earth - so the story goes - Interestingly, it was the observations of those willing to travel a bit, whether Eratosthenes measuring the Earth or countless mariners - along with some good logic - that built a *database* that would allow others to get a handle on cosmic movement to a greater extent.

When we have looked out through the big domes over the last 90 years or so and seen these *violent* things - supernovae, etc - occurring in distant places - because we are capable of seeing so many *times* at once - It must have seemed like our part of the universe was pretty quiet.
But in this case it was taking the time to not rush to judgement (or to disprove an assumption that had been rushed) in that these astronomers "spent 15 years making 1001 nights of observations to detail the motions of more than 14,000 stars that are currently in the solar neighborhood"
That's a lot of time and a lot of stars for a research paper that will added to the file as 'done that'.

Re:Plotting Ahead?

[ekuns]

Stars do collide - I'm sure quite frequently given the number of them. It's just a question of how likely it is to happen to any one particular star - ie: ours.

Sorry, but stars collide quite rarely, because they are so astonishingly small compared to the space around them. Even in galactic clusters -- as dense as space gets when one is looking at stars -- stars rarely physically collide.

Now, star SYSTEMS collide. That means that in a galactic cluster, two solar systems can pass closely enough to cause orbital disruption. This would be especially a risk to life, as any disruption of a star's Oort cloud would cause an incoming rain of cometary objects.

But that being said, the frequency of stars physically colliding is unimaginably small.

Even when galaxies collide, it's a gravitational "collision" and not a collision like a car hitting a wall. When galaxies collide, there are no stellar collisions. Yes, people get a hole-in-one in a golf course. But to go to the original example, if the tee was ten miles from the cup, how often would that happen? Keep in mind that stars are not aimed. :)

About the only way for stars to collide would be gravitational capture with something causing the orbit to lose energy. With that, one star would spiral into another (or more reasonably the two stars would spiral toward each other).

Supermassive black holes do not require stars collding into them. The huge quantities of dust and gas in the center of a galaxy are sufficient. And easier to capture.

Re:Plotting Ahead?

[ekuns]

The can figure out where things are going about as well as they can figure out where things came from. It depends critically on our knowledge of where everything else is that the stars can gravitationally interact with. It can be highly chaotic when stars pass at all close to one another. (Using the appropriate definition of "close," which does not include any likelyhood of collision.) If there's a dense gas cloud we're not aware of in the wrong spot, that would invalidate much of this kind of simulation forward or backward.

Local stars are all moving slowly enough (relative to us) that there are no surprises likely. The only kind of surprises we could get would be a brown dwarf or the like, something not visible until it's close. In terms of human lifetimes, this isn't changing very rapidly at all.

Remember, the animation from the article covers 250 million years! Two such rotations ago, there was no life on land and fish were only just beginning to appear. One rotation ago -- the beginning of the simulation -- was the time of the Permian mass extinction. That was a freaking long time ago! That predates roaches! And mammals.

If we're lucky, we'll still be around in another 250 million years.

Re:Plotting Ahead?

[QuantumFTL]

Stars are so small and so far apart that they almost never come close to hitting. Main sequence stars are roughly separated by one parsec (3.26 light-years). You can think of it like having one golf ball ten miles away from you and hitting another and asking if they are likely to collide. Galaxies do collide often and interact because they are close together compared to their width.

Very true! Two things though:

Plotting ahead is useful because even though stars don't collide, they do come close enough to each other to disrupt planetary systems. Even just preturbing the Oort cloud a bit would cause destructive impacts on earth that could wipe out all life.

Second of all, galaxies "collide" however it's kinda like two clouds of smoke "colliding" because there's really not much actual physical contact (although black holes might eat quite a bit during this period).

Good to point this out though! Galactic dynamics is quite fascinating, and much more complex than you'd expect!

Cheers,
Justin

Re:Plotting Ahead?

[Ayaress]

Even in globular clusters, where star densities are thousands of times greater than the rest of any given galaxy, star collisions are impressively rare. In fact, I know of no observed instance of two stars colliding that weren't formed together in a binary system. The size of the hole compared to the size of the golf course is pretty impressive compared to the scale or stars in space. Two stars colliding would be more like firing a golf ball out of a cannon in London and hitting a hole-in-one in an unmaintained golf course in Scottland. Yes, its possible, but there are a lot of other things that will happen long before that.

Re:Plotting Ahead?

[TMB]

Stars do collide - I'm sure quite frequently given the number of them.

Rather than just asserting that blindly, why don't we work out the numbers?

The only place in the Milky Way that stars collide are inside globular clusters, where they form "blue stragglers". There are about 3000 known blue stragglers in the Milky Way (Piotto et al. 2004). The rate is probably higher now than in the past (it should increase with time as more clusters undergo core collapse), but averaging over the 13.7 billion year history, that's an average rate of one every 4.5 million years

Now, whether you consider that "quite frequently" or not depends on your perspective. ;-)

[TMB]

Re:Plotting Ahead?

[TMB]

It's possible to trace the orbits for the stars they have forward about as well as they can trace them backwards.

The problem, though, is that getting good enough kinematical information to do this is hard for stars that aren't near the Sun. You notice that all of the stars end up right near the Sun - that's not a coincidence! Those are the ones whose orbits we can get a handle on precisely because right now they're close enough that it's easy to get their kinematics.

You can imagine that if you waited 220 million years (one orbit) and did the same exercise, you'd find the same thing - the stars came from a broader range of radii "220 million years ago" (i.e. now). So right now, most of them are too far away for us to get good proper motions and/or good distances, and therefore good spatial velocities.

[TMB]

Re:Plotting Ahead?

[Anonymous Coward]

Actually roaches and mamals already existed back then. Expecially roaches I bet existed in large numbers. Two orbits ago there wern't any mamals I think, might still have been roaches around though. Some things just don't die. ^_-

Quickshot

Re:Plotting Ahead?

[ekuns]

Actually roaches and mamals already existed back then. Expecially roaches I bet existed in large numbers

Sorry, but roaches and mammals are both younger than 250 million years. Not by a whole lot, but still, younger is younger. And like I said, two orbits ago there was no life on land. Probably not even algae on rocks. Only oceanic life.

In comparison to the 4.5 billion years age of the Earth, multicellular life is very young.

To bring it back on topic, and talking about choatic orbits, we are so far from the core of the galaxy that even the likelihood of a star passing close enough to cause severe disruption of our solar system is unlikely. And things travel so slowly out here that we'd see it coming LONG before it got here if it was at all luminous.

The most likely interaction in this part of the galaxy is that some star would pass closely enough to throw a few comets in from the Oort cloud. That's about it.

Now that's some history !

[cyberchondriac]

I wonder if they'll find anything to indicate dramatic climate changes, as has been speculated, due to radiation I think. Right now we're in the arm of Orion, which is a relatively calm place to be, but we (or should I say Sol) tends to drift a bit.

How common is orbital disruption?

[BerntB]

.. star SYSTEMS collide. That means that in a galactic cluster, two solar systems can pass closely enough to cause orbital disruption. This would be especially a risk to life, as any disruption of a star's Oort cloud would cause an incoming rain of cometary objects.

I've wondered about this.

What percentage of solar systems in the spriral arms would have experienced a larger disruption than a bunch of measly dinosaur killers?

Any life on a planet in such a system would at least be "reset" quite a bit, yes? Is this a major influence on the Drake equation?

Re:How common is orbital disruption?

[Ayaress]

Probably not many. You still have to get fairly close to disrupt the orbits of planets more than just throwing a few comets into the inner solar system. Odds are good that too much disruption would more than just "reset" life, but actually render the system unable to support life. For example, a large gas giant planet could be disrupted into an elliptical orbit that would bring it dangerously close to inner planets. Close interaction with this planet could eject life-friendly planets from the system completely, send them spiraling into the parent star, or even cause them to collide with the gas giant and/or its moons. This sort of thing might offer some explanation for the "hot jupiters" that have been found around many stars (large planets orbiting so close to their parent star that they shouldn't have been able to form in their current orbit).

Extrapolations

[dan42]

It seems too fantastic to me that they can extrapolate the data more than 7 orders of magnitude into the past (15 years into 250 million years). At this extreme, usually the error terms and list of assumptions swamp the usefullness of the projections.
They imply they are measuring the change in both relative position and brightness of the stars. From their conclusion and simulation showing that stars appear closer now (than from 15 years ago), I guess this means the change in relative brightness was the dominating statistic in the extrapolations. I seem to remember that this is very difficult to measure accurately, let alone precisely - especially with land-based telescopes.

Re:Extrapolations

[TMB]

It seems too fantastic to me that they can extrapolate the data more than 7 orders of magnitude into the past (15 years into 250 million years). At this extreme, usually the error terms and list of assumptions swamp the usefullness of the projections.

That's definitely true... and if you asked what the error in the trajectory of any individual star is, you'd find it's pretty high. But the aggregate distribution should be pretty well constrained... you know what fraction are moving outwards, what fraction are moving inwards, what fraction are moving up, what fraction are moving down, etc...

They imply they are measuring the change in both relative position and brightness of the stars. From their conclusion and simulation showing that stars appear closer now (than from 15 years ago), I guess this means the change in relative brightness was the dominating statistic in the extrapolations. I seem to remember that this is very difficult to measure accurately, let alone precisely - especially with land-based telescopes.

You're mis-interpreting the article... what you're suggesting would be impossible to measure, as you suggest (quick numbers: let's say a star 1pc away was moving towards us at 100 km/s, which are ridiculously extreme numbers... then in 15 years it would be 0.15% closer, so 0.3% brighter. About the best photometric accuracy you can get is 1% and that requires a hell of a lot of work).

The way they get the velocity toward/away from us is using the Doppler shift of lines in the spectrum of the star... that's actually quite a bit easier than getting the proper motion (change of position). You only need one observation to get it! (though it sounds like they have multiple observations if they've catalogued the fraction of multiple systems).

[TMB]

Re:Extrapolations

[dan42]

Ahhh red shift - of course...
I also can see now that the conclusions do not imply that our neighboring stars are all moving towards us - but instead they are moving so fast that they must have been very far away a long time ago.
...And combining the measurments with the estimated mass and rotational velocity of the galaxy reveals that the sun traverses a perfect circular orbit while every other star follows a complex epicycle every 250 million years.

Humph!

[Tukla]

I'm perturbed all the time, but they don't write any articles about me!