Death from the stars!

Isaac Newton thought that our Solar System was doomed.  He suspected, but could not prove, that gravitational kicks between the planets would eventually cause their orbits to become unstable. The 18^th century mathematicians Laplace and Lagrange showed that the planets’ orbits, at least to first order, seemed to oscillate in a sinusoidal (wave-like) pattern, leading them to conclude that they would remain stable ad infinitum.  But along came the next generations of mathematicians (including Gauss and Le Verrier), who showed that the sinusoidal pattern in the planets’ orbital evolution eventually breaks down.  Poincaré threw a monkey wrench into this whole business, showing that the general gravitational problem with 3 or more bodies simply cannot be solved analytically. After that, we had no choice but to wait until fast computers were available.

The only way to reliably determine the long-term future of the Solar System is to simulate it directly.  That is, to put the planets into a computer and evolve them forward under the effects of gravity.  The problem is that we don’t know the positions of the planets to infinite precision.  Using satellites, radar, and other astronomical techniques, we can pinpoint the position of the rocky planets to within a few meters, and the outer (giant) planets to within a few tens of meters (for gory details, see here).  That’s awfully precise, given that the planets are thousands of kilometers in size and hundreds of millions to billions of kilometers away. These simulations have been done by a number of groups across the world, spearheaded by Jacques Laskar of the Paris Observatory.

Full simulations of the Solar System have revealed two dramatic results. First, the planets’ orbits are chaotic. This does not mean that the planets’ orbits randomly jump around, but rather that we cannot project them indefinitely into the past or the future.  There is a chaos-dictated `horizon’ beyond which any estimation of Earth’s exact orbit is no longer reliable.  This is especially important when trying to connect Earth’s climate and orbit, in order to understand the exact cause and pattern of ice ages (so-called Milankovitch cycles).  Beyond 60-70 million years into the past, chaos makes it impossible to perfectly reconstruct the history of the planets’ orbits.

The second result is that the Solar System may not be stable forever.  Simulations show that there is about a 1% chance that Mercury will fall into a state in which its orbit tracks Jupiter’s orbit (called a secular resonance; illustrated in the image below).  This will cause the shape of Mercury’s orbit to become more and more stretched-out.  Mercury will either collide with the Sun, or have a close approach with Venus, which would lead to either a collision with Venus or a complete destabilization of the rocky planets’ orbits, likely involving a planet or the Moon crashing down on Earth. 

In about five billion years, things will change dramatically. The Sun will become a red giant, swallowing Mercury and Venus (and maybe Earth), and pushing the other planets outward.  The 1% odds of the Solar System being chaotically disrupted are for the next 5 billion years, before the red giant phase. 

Almost all of the work on Solar System stability and chaos has assumed that our Sun is isolated in space.  Of course, this is not true: the Sun orbits among a hundred billion stars in the Milky Way galaxy, and stars fly past the Sun all the time.  The typical approach distance is far enough away that, most of the time, stellar flybys have no measurable effect on the planets’ orbits. But that is not always true. If you throw randomly at a dartboard, the odds of getting a bullseye are pretty low – that’s like a star passing close enough to a star to have a direct impact on the planets.  But if you throw enough darts, one will eventually hit the bullseye. (Statistically-speaking, for a standard dartboard, it takes 199 randomly-thrown darts to hit the outer bullseye and 1261 to hit the inner bullseye). Over the Sun’s history, there have been about 100 thousand ‘darts’ thrown our way in the form of stars passing by. Even though the effect of most of them was probably tiny, it is statistically improbable that none of them had an impact. Over the past couple of years, Nate Kaib and I have taken a closer look at the effect of flybys, and it turns out that they matter more than you might think.

In a 2024 paper, we explored the role of stellar flybys in the chaotic evolution of the planets.  We showed that flybys shorten the horizon inside which the orbits of the planets can be reliably estimated – from about 70 million years into the future or the past, down to about 60 million years.  This puts the limit of reliability right on the edge of a peculiar time in Earth’s past called the “Paleocene-Eocene Thermal Maximum,” a period when Earth’s average temperature was 5 to 8 degrees Celsius higher than immediately before or after.  This thermal maximum took place about 56 million years ago and only lasted for about 200 thousand years. Its cause is debated, and a number of ideas have been proposed, including an episode of strong volcanism, a comet impact, a change in ocean circulation, a strong release of methane from clathrates in sea-bottom sediments, as well as orbital forcing. The cycle of ice ages is well-known to correlate with Earth’s orbital shape and spin alignment, and it is possible that Earth’s orbit was especially stretched-out during the Paleocene-Eocene Thermal Maximum.  However, we showed that it is very difficult to trust the exact phasing of Earth’s orbit that long ago, as stellar flybys may well have ‘scrambled’ the signal.  There are specific flybys of stars characterized by the Gaia space telescope that may have played a role in this scrambling.

In a new paper, we look to the future and show how stars flying by the Sun affect the long-term stability of the Solar System. The first thing that we found is that Pluto’s orbit is surprisingly unstable.  Previous work had found that the odds of Pluto becoming unstable were basically zero.  However, we show that there is about a 5% chance that Pluto’s orbit will be perturbed by a passing star.  The most likely outcome is that Pluto will be gravitationally scattered by Neptune, then passed inward between the giant planets until it is ejected from the Solar System by Jupiter.  (Kind of like what happened to comets when they were kicked out of the early Solar System). 

Our most dramatic result is that stellar flybys make the Solar System more than 50% less stable than if the Sun were isolated in the Galaxy.  In our simulations, most of the time it was the same planet getting knocked out of its orbit: Mercury.  Even though it is the closest planet to the Sun – and, therefore, unlikely to be directly perturbed by a passing star – Mercury’s precarious orbit is vulnerable to small changes in the giant planets’ alignments.  One way this can happen is a small kick to Jupiter’s orbital position, leading Mercury to fall within its unstable secular resonance (the same one that dominates the instability when stellar flybys are neglected). But even a small shift in Uranus’ orbital position has a similar effect, and Mercury’s orbit is eventually destabilized in a trickle-down fashion.  It’s like Mercury is at the end of several chains of dominos – more than one can eventually knock it over (or into the Sun).

Stellar flybys are the main potential driver of Solar System instability in the near-term.  The odds of an internally-driven instability increase slowly in time, and peak 4-5 billion years in the future (before the Sun goes red giant in about 5).  In contrast, the odds of a flyby-driven instability are roughly constant in time. As you can see in the image below, for the next 4-4.5 billion years, stellar flybys pose the greatest threat to our Solar System. Thankfully, the odds of a star flying extremely close to the Sun are very small (and we explored their dramatic consequences in a recent study). 

There is also a chance that flybys could affect Earth’s climate without destabilizing the planets’ orbits.  This would happen if Earth’s orbital shape were stretched-out enough to increase the total amount of solar energy reaching our planet’s surface.  Of course, Earth only has 1-2 billion years left before the Sun brightens enough to render it uninhabitable. There is about a one-in-a-thousand chance that Earth’s orbit will be stretched out by a flyby before that happens.  Sadly, the odds are that the flyby will actually heat up the Earth, and probably accelerate its demise.

None of us is truly alone, not even the stars. And while we humans get massive health benefits from social interactions, the same cannot be said for planetary systems.  Passing stars are like darts being thrown at planetary systems – best to stay as far away as possible!

---

References/resources:

• Our new (2025) paper, entitled “The Influence of Passing Field Stars on the Solar System’s Dynamical Future”
• Our 2024 paper, entitled “Passing Stars as an Important Driver of Paleoclimate and the Solar System’s Orbital Evolution”
• A great review about studies of Solar System stability by Jacques Laskar, entitled “Is the Solar System Stable?”
• The End of the Solar System — a blog post about how our Solar System will end, including the effects of chaos and the Sun going red giant (from the Solar System’s story)
• Could a stellar flyby save Earth from impending doom? — A blog post about the effect of extremely close (and unlikely) encounters on the Solar System.