Passing Stars Shake Up Simulations of Earth’s Past Orbit and Climate

The chaotic interactions between solar system bodies prevent us from knowing Earth’s precise orbit billions of years in the past or future. New research shows that passing stars introduce even more uncertainty, shortening the time out to which we can reliably predict Earth’s past orbit by 10%.

Uncertain Future, Uncertain Past

Though the solar system may seem stable and predictable, with the planets moving calmly along well-defined orbits, this stability may only be temporary. Planets and asteroids gravitationally nudge one another, introducing an element of chaos into the calm and making it possible — though improbable — that one or more planets in our solar system will be ejected or destroyed in the next 5 billion years.

This chaotic element doesn’t just introduce uncertainty into the future behavior of the planets. It also means we can’t look infinitely far into their past motions, either. For Earth, the limit seems to lie around 60–70 million years ago, meaning that we cannot extrapolate Earth’s orbital history beyond that point. Because changes in Earth’s orbit influence our planet’s climate, researchers interpreting the ancient climate record contained in ice cores and sediment consider the behavior of Earth’s past orbit, and it’s critical for them to know how far back our orbital simulations are reliable.

Previous studies have focused on how planets and asteroids affect simulations of Earth’s past orbit. But the solar system doesn’t exist in a vacuum (metaphorically speaking) — could passing stars affect our ability to study Earth’s orbital history as well?

Chaos Caused by Passing Stars

In a research article published today, Nathan Kaib (Planetary Science Institute and University of Oklahoma) and Sean Raymond (National Centre for Scientific Research, France; University of Bordeaux) used numerical simulations to understand how passing stars impact our ability to trace back Earth’s orbit. Kaib and Raymond added a tiny amount of scatter in the positions of the planets and large asteroids and allowed the system to evolve backwards over 150 million years of simulation time.

Some of the simulations contained stars placed at random positions about 3.4 light-years from the Sun and given random velocities. (Technically, all of the simulations contained passing stars, but in the star-free runs the stars’ masses were set to minuscule values. This prevents the results from depending on the computational methods being used rather than the presence of the stars.) By marking the moment at which Earth’s orbital eccentricity, or how circular its orbit is, diverged by more than 10%, the authors determined how far back their simulations of Earth’s orbit were reliable.

Close Encounters of the Stellar Kind

Kaib and Raymond found that passing stars interfere with our ability to trace back Earth’s orbit, shortening the time we can reliably look back by about 10%. The simulations suggest that stars act on Earth by nudging the giant planets, which then nudge Earth; when the giant planets were removed from the simulation, passing stars had little effect.

To look more closely at a specific event, Kaib and Raymond considered the star HD 7977. HD 7977 currently sits about 250 light-years away, but its trajectory suggests that it passed within about 13,200 astronomical units (au) of our solar system 2.8 million years ago, making it possibly the closest recent stellar encounter. The team showed that the star’s passage likely didn’t impact Earth’s orbit much, unless the star came as close as the data allow — there’s a 5% chance that the star came within 3,900 au.

If HD 7977 did approach the solar system so closely, the encounter may have introduced a large amount of uncertainty into simulations of Earth’s past orbit. This allows for a broader range of past orbital eccentricities, opening up new possibilities to consider for Earth’s orbital and climatic past.

Citation

“Passing Stars as an Important Driver of Paleoclimate and the Solar System’s Orbital Evolution,” Nathan A. Kaib and Sean N. Raymond 2024 ApJL 962 L28. doi:10.3847/2041-8213/ad24fb