General Relativity may tear the solar system apart

[BoJangles2]

Article i read this morning, unfortunatly i dont have a link to the research, though sciencedaily and such will likely run it soon, if you have a link to the actual paper please paste it.

A FORCE known as orbital chaos may cause our solar system to go haywire, leading to a possible collision between earth and Venus or Mars, according to a study released today.

The good news is that the likelihood of such a smash-up is small, around one-in-2500.

And even if the planets did careen into one another, it would not happen before another 3.5 billion years.

Indeed, there is a 99 per cent chance that the sun's posse of planets will continue to circle in an orderly pattern throughout the expected life span of our life-giving star, another five billion years, the study found.

After that, the sun will likely expand into a red giant, engulfing earth and its other inner planets - Mercury, Venus and Mars - in the process.

Astronomers have long been able to calculate the movement of planets with great accuracy hundreds, even thousands of years in advance. This is how eclipses have been predicted.

But peering further into the future of celestial mechanics with exactitude is still beyond our reach, said Jacques Laskar, a researcher at the Observatoire de Paris and lead author of the study.

"The most precise long-term solutions for the orbital motion of the solar system are not valid over more than a few tens of millions of years," he said.

Using powerful computers, Mr Laskar and colleague Mickael Gastineau generated numerical simulations of orbital instability over the next five billion years.

Unlike previous models, they took into account Albert Einstein's theory of general relativity. Over a short time span, this made little difference, but over the long haul it resulted in dramatically different orbital paths.

The researchers looked at 2501 possible scenarios, 25 of which ended with a severely disrupted solar system.

"There is one scenario in which Mars passes very close to earth," 794 kilometres to be exact, said Mr Laskar.

"When you come that close, it is almost the same as a collision because the planets get torn apart."

Life on earth, if there still were any, would almost certainly cease to exist.

To get a more fine-grained view of how this might unfold, Mr Laskar and Mr Gastineau ran an additional 200 computer models, slightly changing the path of Mars each time.

All but five of them ended in a two-way collision involving the sun, earth, Mercury, Venus or Mars. A quarter of them saw earth smashed to pieces.

The key to all the scenarios of extreme orbital chaos was the rock closest to the sun, found the study, published in the British journal Nature.

"Mercury is the trigger, and would be be the first planet to be destabilised because it has the smallest mass," said Mr Laskar.

At some point Mercury's orbit would get into resonance with that of Jupiter, throwing the smaller orb even more out of kilter, he said.

Once this happens, the so-called "angular momentum" from the much larger Jupiter would wreak havoc on the other inner planets' orbits too.

"The simulations indicate that Mercury, in spite of its diminutive size, poses the greatest risk to our present order," said University of California scientist Gregory Laughlin in a commentary, also published in Nature.

 

[MeteorWayne]

Hmm, well part of the opening isn't new news, since due to the intractability of any 3 body (or more) problem we already knew that the solar system orbits could not be predicted for more than a billion years anyway. However, they have added a new twist in taking relativity into account. I'll keep an eye out for the original research.

Here's the SF Chron article, which states the study in in today's Nature, so I'll get to read it in a few days

http://www.sfgate.com/cgi-bin/article.c ... pe=science

And for those who want to panic, please read this again from the first article

And even if the planets did careen into one another, it would not happen before another 3.5 billion years.

Indeed, there is a 99 per cent chance that the sun's posse of planets will continue to circle in an orderly pattern throughout the expected life span of our life-giving star, another five billion years, the study found.

For a different take, here's the SDC article:

http://www.space.com/scienceastronomy/0 ... llide.html

 

[MeteorWayne]

Having read the article in Nature early this morning, I thought I'd add some fascinating details about how the study was conducted. The science nerds among us should enjoy this.

First, I should point out that they ran the simulation in a more basic form, without relativity, and ignoring the Moon, and the results were much more dire for the solar system. They ran 200 simulations with the semimajor axis "a" of Mercury offset in 3.8 cm steps in an increased distance, and 3.8 cm in a decreased distance, 100 steps in each direction.
(The semimajor axis of Mercury is only known to a precision of a few meters). In about 60% of the solutions large scale disruption of the inner solar system would result in ejection of planets, or collisions between the planets, or planets and the sun during the 5 Gy integration time.

Adding relativity and the moon really calms things down, only 20 of 2501 solutions lead to problems of that scale (0.8%)

I'll put that in my next post; have to keep them short today due to some issues.

 

[MeteorWayne]

From the article:

"Owing to the chaotic behaviour of the Solar System, the distance between two initially close orbits increases by a factor of 10 every 10 million years. It is thus hopeless to search for a precise solution for the motion of the SS over 5 Gy, that is, over a time comparable to it's age or life expectency...The most precise long-term solutions for the orbital motion of the SS are not valid for more than a few tens of millions of years. A numerical integration of the SS's motion over 5 Gy can thus only be considered as a random sample of it's possible evolution. Statistical studies are then required to search for possible changes in the planetary orbits that lead to collisions or disruption of the system"

So how do they accomplish this? First they are using the most recent ephemeris of the solar system INPOP06. The model is then run in steps of about 9 days, unless the eccentricity of any of the planets exceeds 0.4, then the step size is reduced to preserve numerical accuracy. To generate the random orbits, the start with the nominal "a" (semimajor axis) of Mercury and integrate it 2500 more times, 1250 in each direction (+ and -) in steps of 0.38 millimeters! for a total range of about a half a meter in each direction.

Next up, a discussion of these 2501 integrations.

 

[MeteorWayne]

"Among those 2501 solutions that are compatible with our best knowledge of the Solar System, in 20 the eccentrcity of Mercury increased beyond 0.9"

At the time of the writing of the article, 14 were still running, and will probably do so for a few more months. This is due to the reduced step size I mentioned in the last post.

Of the 6 remaining, Solution S(-947) {that means the nominal orbit minus (947 X 0.38 mm)} reached 5 Gy without collision although a close encounter of Mercury and Venus occurred at 4.9 Gy.

In 3 Solutions, S(-915), S(-21), and S(+33), Mercury collided with the sun between 4.218 and 4.314 Gy.

In S(-812) Mercury collided with Venus at 1.763 Gy

The most interesting is S(-468) in which a close encounter of Mars with the Earth (CA distance 794 km) occurs at 3.3443 Gy.

"Such a close approach would be disasterous for life on Earth with a possible tidal disruption of Mars and subsequent multiple impacts on the Earth, but we also searched for more direct collisions."


This one was analyzed in more detail; that I'll discuss in the next post.

 

[MeteorWayne]

So what they did was ran this simulation forward in time from 3.344298 Gy, 100 each with an plus and minus offset of 0.15 mm steps in the semimajor axis of Mars.
This means 201 integrations, the nominal, and 100 steps plus 0.15 mm and 100 steps minus 0.15 mm.

Of the 201 solutions 5 resulted in Mars being ejected from the solar system. The other 196 resulted in the following collisions:

Sun-Mars 48 cases
Mercury-Venus 43
Sun-Mercury 33
Earth-Mars 29
Venus-Mars 23
Venus-Earth 18
Mercury-Earth 1
Mercury-Mars 1

Fascinating!!!

Wayne

 

[BoJangles2]

Thanks for that MW great read, indeed very fascinating that the Mars - Earth collision is fairly high up on table. It will intersting to see what the other solutions hold when the research is finished.

 

[MeteorWayne]

Thread bumped for origin

 

[origin]

Thread bumped for origin

Thanks Wayne.