The Solar System Has a New Life Expectancy, and It’s a Billion Times Shorter

Isaac Newton suspected that the gravitational dance between the Sun and Jupiter would eventually destabilize the solar system, but modern computational research has shown that its long-term stability is, statistically, pretty secure. However, new simulations with more realistic solar mass-loss conditions make a strong case for the disassembly of the outer solar system — and on timescales much shorter than previously thought.

Kick It Up a Notch

A plot showing final orbits of the outer planets for two simulations.

Final outer planet orbital configurations for two simulations — one stable and one unstable. t=0 at the onset of white dwarf formation. Click to enlarge. [Adapted from Batygin et al. 2026]

Several studies have investigated what will happen to the orbits of planets in our solar system as the Sun sheds its mass during its transition to a white dwarf. These studies have found that while the semimajor axes of the planets will expand by approximately a factor of 2, their orbits remain stable on timescales longer than the age of the universe (approximately 1018 years). This expanded orbital architecture will make the solar system more prone to jostles from stellar flybys, but even accounting for this, the giant planets’ orbits wouldn’t disassemble for another 30–100 billion years.

However, all of these studies modeled solar mass loss as a smooth function, so the dismantling of the solar system was left up to external forcings. In reality, stars on the red giant branch undergo mass ejections that are random in time and direction, corresponding to velocity “kicks” that can alter the dynamics of the planets. 

A team led by Konstantin Batygin at the California Institute of Technology conducted numerical experiments to analyze the orbital dynamics of the outer solar system under different solar mass-loss conditions. First, they tested the classic case of smooth mass loss and validated previous findings that, despite planetary orbits doubling in size, the dynamics are pretty tame. This holds true even when mass is ejected anisotropically: if the timing is smooth, the asymmetries cancel out.

Outer Solar System, Your Days Are Numbered

Next, they simulated stochastic kicks and found that the behavior of the outer planets was largely tied to the ejection mass of each kick. For very small ejection masses (≤ 10-7 M☉), dynamics mimic the smooth mass-loss case. For intermediate values (10-7 to 10-6 M☉), there’s a roughly 20% chance Jupiter–Saturn and Uranus–Neptune enter mean-motion resonances: an orbital configuration where planets’ periods are a ratio of small integers. These resonant configurations don’t protect the giant planets, however: resonant simulations disassemble slightly more often than nonresonant simulations. For larger ejection masses (≥ 10-5 M☉), the outer solar system disassembles on the timescale of billions of years. 

Plot showing the instability onset times as a function of ejection mass for all simulations.

Instability onset time — or how long after white dwarf formation the outer solar system disassembles — as a function of ejection mass. The larger the ejection mass, the greater the chance of instability. Click to enlarge. [Batygin et al. 2026]

Which ejection masses are the most accurate for our Sun? Well, unfortunately for our beloved outer planets, ejection models and observed kick amplitudes suggest a future ejection mass of 10-4 M☉. With kicks like these, not only is there a considerable chance of orbit destabilization, but it could happen before the white dwarf even forms. About 40% of simulations lead to disruption or violent scattering while the Sun is in its red giant phase, and roughly 90% break down within 3 billion years. 

This new discovery reduces the life expectancy of the solar system from one billion billion years to a meager one billion years after white dwarf formation. 

Citation

“Terminal Instability of the Solar System Triggered by Stochastic Solar Mass Loss,” Konstantin Batygin et al 2026 ApJL 1009 L22. doi:10.3847/2041-8213/aea290