How planets die: When good Jupiters go bad

“Whether they ever find life there or not, I think Jupiter should be considered an enemy planet” — Deep Thoughts by Jack Handey

This series is about how planets die — it is introduced here.

In the sky, Jupiter doesn’t look too different than other bright planets like Mars or Venus or Saturn. But Jupiter  has the power to destroy the Solar System. And many of its cousins (around other stars) are planetary murderers (eek!).

Jupiter is the Solar System’s most massive planet. It is 300 times more massive than Earth. Its gravity dominates the present-day Solar System (apart from the Sun of course).  And during its early years it shaped the entire Solar System.

Let’s go through the main ways in which Jupiter has influenced the Earth, starting before Earth was fully-formed:

  • Jupiter protected the inner Solar System from icy invaders. Once Jupiter grew large enough to become a gas giant it carved a gap in the Sun’s planet-forming disk and blocked the migration of more distant icy bodies.  Jupiter held back these large icy bodies, which instead grew into Saturn’s core, Uranus, and Neptune.
From this Nautilus article from 2015.  Technical details here. See also MOJO video 9.
  • Jupiter is the reason Earth has water but not too much.  In addition to holding back the invasion of icy bodies, Jupiter scattered primordial asteroids and comets toward the growing Earth, seeding it with enough water to have both oceans and land (see this post or MOJO videos 5 and 6 for more details).
  • Jupiter cleared out the asteroid belt.  It removed 99.9% of the asteroids and almost emptied out the belt, which could in principle have held another planet. Some of the details are uncertain (see this post or MOJO video 4 for more) but Jupiter is almost certainly responsible.
  • Jupiter controls the dynamics of comets and influences whether they collide with Earth.  Jupiter’s gravity determines how comets enter the inner Solar System and how long they spend near the planets (with the potential to crash into Earth) before launching them into interstellar space. [It’s worth mentioning that this argument is circular: Jupiter only actually protects Earth from comets that are only a threat because of Jupiter.  Still, Jupiter doesn’t seem to be doing Earth any harm.]

To sum up: Jupiter has really helped out the Earth.  Its early influence cleared a path for our Earth to grow.

Jupiter was like Earth’s rich uncle, who got Earth into the best schools, kept Earth out of any sticky situations and generally had Earth’s back. It’s easy to wonder whether Earth would even be here without Jupiter.

But Jupiter’s benevolence is out of character. 

Most gas giants have a destructive influence on rocky planets. They are more like a drunken uncle who shows up out of the blue on a bender with gangsters and machine guns (and the story does not end well….).

Gas giants are destructive because they are unstable.  Planet-forming disks that grow one giant planet usually grow two or three or four.  The gas disks have a stabilizing influence: by dissipating energy they keep the planets in line. But disks evaporate after a few million years… That’s when good Jupiters go bad.

Gravitational nudges between the gas giants add up and stretch out their orbits.  When two planets’ orbits cross things get even crazier.  Because their gravity is so strong, the planets rarely collide.  Instead, they give each other gravitational kicks.  Eventually, one planet is kicked hard enough to escape into interstellar space! Check out this animation by Eric Ford:

Notice how the outer surviving planet in the animation has a stretched-out, eccentric orbit?  That is a scar of the instability. When we discover gas giants on eccentric orbits around their host stars, it tells us that those systems underwent instabilities like this.

When a gas giant goes bad its whole system suffers.  Because giant planets are so massive, nearby small planets are kicked all over the place during gas giant instabilities.

Here is a simulation of a planetary system in which three Jupiters go unstable. It starts off looking a lot like our own young Solar System. Close to the Sun rocky bodies are growing (with colors that correspond to water content) by colliding with each other. Far out there is a disk of comet-like leftovers (dark blue).  In between there are three gas giants instead of our two (Jupiter and Saturn).

Here the giant planets went unstable late (41 million years into the simulation).  During the instability, one gas giant — and all of the outer icy comets — were ejected into interstellar space.  The giant planets’ moons can sometimes survive.

When the instability hit the largest rocky planets had already grown to Earth-size.  During the instability, the rocky planets were driven into the star: the wandering gas giants stretched the rocky planets’ orbits so much that they ended up inside the star.

What actually happens when a planet falls onto its star?

Most of the time it’s probably not too dramatic.  The planet just gets swallowed up and then vaporizes inside the star.

Sometimes planets are shredded on their way to being eaten. This only happens when planets fall onto stars that are very dense (side note: the interstellar object ‘Oumuamua may be a fragment of a larger body that was disrupted — see here).  Before the planet can drop below the surface, tidal forces tear it into pieces and create a disk of dust that drains onto the star over thousands of years…

White Dwarf Shredding A Planet Painting by Lynette Cook
Artist’s impression of a planet being torn to shreds as it falls onto a very dense, white dwarf star.  Credit: Lynette Cook.

Instabilities like this are how gas giants destroy rocky planets.

Rocky planets are usually driven into their host stars to die fiery deaths. On occasion they collide with a gas giant or are ejected into interstellar space, but usually it’s a fiery death.

Depending on the timing, unstable giant planets may either destroy fully-formed rocky worlds or their building blocks. Gas giant planets form fast. Despite being smaller, some rocky planets take longer to finish forming. We know that Earth took 50-100 million years to reach its final size whereas Jupiter’s was complete within a few million years.

There are two kinds of triggers for giant planet instabilities: internal and external.

Systems often go unstable on their own. The gaseous disk has a calming effect.  It’s like a teacher at recess, keeping the kids (planets) in line.  When the teacher (gas disk) goes away, the kids (planets) usually go nuts immediately.  But sometimes they take their time, and tiny changes in the planets’ orbits accumulate until they reach a point of no return.  And some well-behaved systems never go unstable on their own.

External factors can also make planets go unstable.  A star can pass too close and kick the planets out of alignment.  In wide binary star systems, changes in the binary’s orbit can bring the second star close enough to trigger instability (animation here).  Even belts of planetary leftovers can trigger instability (and trigger a rain of comets and asteroids).

When giant planets go unstable early they destroy rocky planets before they are fully-grown. It’s like planetary abortion.

When giant planets go unstable late it’s a galactic tragedy.  It takes time for a planet like Earth to fully form and longer for life to develop and take hold.  When giant planets go unstable afterwards and destroy their life-bearing world it’s a serious bummer.  Luckily, we think that most of the time giant planets go unstable early.

Let’s put Jupiters going bad on our Planetary Death Scale.  It’s bad.


Not all bad Jupiters kill off their Earths, but it’s very common.

Now let’s figure out how many planets have been destroyed when their Jupiters have gone bad.

About 10% of stars like the Sun have one or more gas giants like Jupiter.  But only about 10% of stars are Sun-like stars.  Most stars are puny red stars, which host far fewer Jupiters.  Averaged over all types of stars, about 3% have gas giants.  There are a few hundred billion stars in the Milky Way.  That makes about 10 billion stars with gas giants.

To match the orbital shapes of gas giants, about 75% of the systems we see must be survivors of instabilities. And about 75% of instabilities are strong enough to destroy their rocky planets. Put together, about half of all stars with giant planets have destroyed their rocky worlds.

An evil gas giant. Credit: Phil Plait (from this article). Photos by NASA, ESA, and A. Simon (Goddard Space Flight Center)/Shutterstock & Tribalium.

That makes about 5 billion cases of planetary murder in our Galaxy by Jupiters going bad. 

And while most Jupiters probably destroyed their rocky worlds before they were fully-formed, a fraction were unstable late.  If just 1% of systems went unstable late then that is 50 million systems in which inhabited worlds were tossed into their stars.  Ouch!

And now, a short (tragic) story about an alien named Bob….

Bob lived on an Earth-like planet in a different Solar System.  In Bob’s system there were three gas giants, each roughly like Jupiter.  They are a bit closer and brighter in the sky than our Jupiter, and are symbols in the main religions on Bob’s home world.

Bob was a student interested in astronomy. His professor gave him a small research project, something she thought was of little consequence.  Bob simply had to measure the relative positions of the three giant planets over the course of the school year in order to map out their orbits.

The project plodded along.  It was less dynamic than Bob had hoped but he are dedicated.  Each night (weather permitting) Bob measured the planets’ positions just after sundown and and just before sunup. Every week Bob entered the planets’ positions into a spreadsheet.

Finally, a few weeks before the end of the year Bob computed the orbits of the planets. He also compared with the well-known, “true” orbits of the planets from the textbook. They didn’t match.  Bob’s measurements diverged from the textbook orbits early in the winter and had been veering farther off course since.

Bob felt like he must have made a mistake.  He double checked each date and each number.  He redid the calculation over and over.  He couldn’t find a mistake.

Bob went to the professor and showed her his work.  She smiled the slightly condescending, unsurprised smile of a teacher whose student made the usual dumb mistake.  But she checked Bob’s work and couldn’t find his mistake.

It must be that Bob’s measurements are bad.  They looked okay but the professor didn’t trust them. Over the next two months the professor met Bob periodically at the telescope to make the measurements together.  But the measurements keep following the path that Bob had already mapped out.

After countless nights the professor admitted that it seemed like Bob’s calculations were right and that the textbook orbits of the gas giants were wrong.

The professor ran a computer simulation to project the planets’ orbits forward in time. According to the computer, the innermost gas giant’s orbit would quickly become stretched-out and in just a few hundred years, a blink of an eye in the history of Bob’s civilization, things were looking bleak. The innermost giant planet was going to gravitationally scatter off of the middle planet. This would trigger a chaotic phase of gravitational kicks between the giant planets that would stretch and tilt their orbits. Sixty thousand years later the inner gas giant would be ejected from the system and the two other giant planets would survive on wider, eccentric orbits. The bad news: while the giant planets were scattering each other, Bob’s home planet crashed into its host star, along with its two neighboring rocky planets.  It looked like this:


Bob and the professor ran thousands more simulations. They don’t know the orbits of the planets perfectly and the process is chaotic.  They couldn’t know precisely how things would turn out because if the starting positions of the planets were off by just an inch the planets’ behavior would be different.

The computer chugged along overnight. In the morning they had the answer: their home planet was doomed.  There was a 72% chance of their planet colliding with the star, an 18% chance of impacting another planet and then colliding with the star, an 8% chance of being ejected into interstellar space, and just a 2% chance of survival.  But even if the planet survived its orbit would be stretched-out and tilted, which would dramatically change the climate.

Things were about to change on Bob’s world.  Within a few hundred years the planet was doomed. But there at least a century before things got out of control and the planet’s orbit started to change drastically.

The professor talked to a colleague on the government’s science advisory board.  Within days she presented their findings. Astronomers across the planet scrambled to observe the planets and within a month had confirmed Bob’s and the professor’s findings. To this point everything was kept hidden from the public.

Then the president made the announcement: our planet is doomed.

Everyone went bonkers. It’s not easy to stay in control when your whole world has no future. Riots and general unrest lasted for weeks. Public services were disrupted. Crops started to rot. Things were looking bleak.

What was next?  Should their civilization race to achieve interstellar flight before it was too late?  Or should they just accept their fate, as a few religions were advocating that this was a sign from God that the end was near?

Politics took over. Bob’s civilization took a giant leap toward space flight.  But would they make it in time?  (Cue dramatic music…)

Additional resources:


18 thoughts on “How planets die: When good Jupiters go bad

  1. Is the high population of eccentric Jupiters actual, or is there some observation bias? And how often are poor unfortunate rocky planets captured as moons? Is that rare?

    1. There is some uncertainty in the eccentricities of exo-Jupiters. The radial velocity signal created by two planets in resonance can sometimes mimic that of a single planet on an eccentric orbit, so it’s possible that there is some contamination (probably not more than about 1/4, but it’s not too well known). In terms of capture of rocky planets as moons — this is possible but very rare. I don’t actually know of any numbers on this but I never saw it happen in a few thousand simulations…

  2. Let’s not wait for something to happen, we should have interstellar travel now, but were too busy destroying ourselves, our solar system is ours to explore, and so is interstellar space.. let’s do it. We can easily do it..

  3. I love that you included a story with this! That makes all this seem far less academic and far more personal. Poor Bob. I hope his planet’s space program worked out.

  4. “I‘m just so curious about the discoveries we can make, and learning about our beginnings, and what can Jupiter tell us about how the 00004000 solar system was made,” says mission principal investigator Scott Bolton of the Southwest Research Institute.

  5. I have my doubts about world-wide panic triggered by the discovery of the world being doomed in a few centuries. Panicking when something in your environment turns out to be significantly different than expected is not particularly useful. So I doubt such a reaction would be common in a sentient species. For comparison, only ten percent of humans panic in case of disaster.

    1. Also, I think there would be multiple astronomers simultaneously discovering changes in the orbits of other planets in the same solar system. A civilisation possessing computer simulations would have at least tens of thousands of people having as their job to systematically watch the sky. Sensibly, some of them would be studying the three largest planets of their system and would have noticed if their orbits were off. As an astronomer I think you would have realised that.

  6. Sorry for calling you an astronomer instead of an astrophysicist. (I misunderstood what you do for a living.) What do you think could destabilize the orbits of giant planets billions of years after they formed? Would all those factors be noticeable? Or would the change be too slow for that?

Leave a Reply to Gregrox Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s