Interstellar exchange of massive objects is difficult across several light years. But as the stars orbit the galaxy the distances between them change. I don't find data for neighbor star distances which covers more than a couple of a hundred thousand years. I can imagine that it is difficult to trace their orbits further given todays uncertainties in distance measurements. But I'm interested in a purely statistical/geometrical estimation of how frequent nearby Sun-star passages have been.

  • How many times has a star come closer than, for example, 1 light year from the Sun since it formed?
  • How well mixed have the nearest stars become over time? What fraction of them have followed the Sun since formation and made multiple passages?

I suppose that 1 light year's distance is enough to cause disturbances in the Oort cloud and likely encourage comets.

From Wiki


2 Answers 2


TL;DR There is an encounter within 1 light year about once every half a million years. So about 9000 close encounters during the lifetime of the Sun

The answer to the second part of your question is that the stars near the Sun are extremely well mixed, with a range of ages from a few million years to 12 billion years. None of them are identifiably "following the Sun".

Some details

Using the re-reduction of the Hipparcos astrometry, Bailer-Jones (2014) has integrated orbits for 50,000 stars to look for objects that might come or might have come close to the Sun.

The K-dwarf Hip 85605 is the winner on that timescale, with a "90% probability of coming between 0.04 and 0.20pc between 240,000 and 470,000 years from now".

The next best is GL710 a K-dwarf that will come within about 0.1-0.44pc in 1.3 million years.

On a statistical basis, some work has been done by Garcia Sanchez et al. (2001). They estimate, using the Hipparcos data, that encounters within 1pc occur every 2.3 million years. However, the Hipparcos data isn't complete for low-mass, faint objects. Making a correction for this, the authors estimate a $<1pc$ encounter every 100,000 years. The probability for closer approaches scales as the inverse of the square of the separation. i.e. the timescales for approaches closer than 0.1pc is 100 times longer.

There are plenty of big error bars in all this work - recall that an error of 1km/s in velocity leads to a 1pc error in position after 1 million years. All this should be nailed down really well by Gaia results in the next 2-3 years.


There is a new paper based on Gaia DR3 astrometry by Bailer-Jones (2022). The star GL710 is now much more firmly constrained, coming within 0.060-0.068 pc of the Sun in 1.26-1.33 million years.

A new contender on the scene is HD7977, a G3 dwarf that may have come even closer at 0.02-0.12 pc about 2.72-2.80 million years ago.

There are a total of about 50 stars that will or have passed within 1 pc of the Sun in the last $\pm 6$ million years - so about one known encounter every 240,000 years. But note that this cannot possibly be a complete list and the figure is a lower limit even over this period, because faint stars and high velocity stars can move in and out of Gaia's sensitivity range on these sorts of timescales and thus would not be in the current catalogues.

Bailer-Jones (2018) (using Gaia DR2 data) attempted to correct for this incompleteness and estimated an encounter rate within 5 pc of $545 \pm 59$ per million years. With the inverse-square scaling mentioned above (and also advocated by Bailer-Jones), this becomes about 1 encounter within 1 pc every 50,000 years or about 1 encounter within 1 light year every 0.5 million years.

Assuming this rate is roughly correct over the whole lifetime of the Sun then this would yield 9000 close encounters over its 4.5 billion year lifetime (arguably there might have been a slightly smaller number of stars earlier in the Galaxy's history, but this might be outweighed by the Sun probably gradually migrating outwards from a radial position with a higher stellar density.)

Projecting this further, then if you are prepared to wait 50 million years, then you might expect an encounter within 0.1 light years, which is in the Oort cloud and really starting to get interesting as far as the timescales of mass-extinction events go.


I was curious about the same things. I believe it was in the astronomy stack exchange I was referred to an online data base that gives position and velocity vectors for neighboring stars. From those I put together a spreadsheet. Here's a screen capture:

enter image description here

I only entered 48 of the closest stars so it's by no mean an exhaustive list.

It looks like your graphic matches my estimates which is reassuring. I don't know why the Ross Stars aren't in my list, possibly the omission is an error on my part when I was entering data to the spreadsheet.

Looks like the closest approaches are around 3 lightyears.

If each star has an Oort cloud, I believe the comets' velocity with regard to our sun would be pretty close to the star's relative velocity. Slowest star wrt our solar system seems to be Gliese 729 which is moving ~14 km/s wrt the sun.

If, for example, some of Van Maanen's Oort cloud came within a light year of our sun, they would have been moving 270 km/s. Those snowballs would have zoomed in and out of our neighborhood.

With these distances and relative velocities I don't see much opportunity for swapping comets.

It's speculated that our sun swapped comets with neighboring stars when our solar system was being formed. From Wikipedia:

Recent research has been cited by NASA hypothesizing that a large number of Oort cloud objects are the product of an exchange of materials between the Sun and its sibling stars as they formed and drifted apart, and it is suggested that many—possibly the majority—of Oort cloud objects were not formed in close proximity to the Sun.

  • 1
    $\begingroup$ On the scale of 0.0002 billion years, movements of stars near the Sun is available. And interaction with siblings since the beginnings of earthly time is only one part of the equation. The other part is how the Sun might have interacted with comets (or spaceships?) from unrelated much older or younger stars, during the 16 or so times the Sun has orbited the Milky Way like once every 0.25 billion years. How frequently should mass exchanging encounters have occured, is really my question. (Wiki's lack of dates makes terms like "recently" mean little in quickly developing fields of astronomy). $\endgroup$
    – LocalFluff
    Jun 2, 2014 at 15:03
  • $\begingroup$ But I think you argue that stars and their comets are most likely to have higher than capture velocity relative to the solarsystem? And for that reason they are unlikely to stick around. $\endgroup$
    – LocalFluff
    Jun 2, 2014 at 15:07
  • 1
    $\begingroup$ That's my thinking. Even if perihleion were 1/2 a light year, 14 km/s Vinf would be a hyperbola of very high eccentricity -- the path would look like a straight line. That's how I modeled the paths -- as straight lines. $\endgroup$
    – HopDavid
    Jun 2, 2014 at 15:12
  • $\begingroup$ It's my understanding they get the stars' distance by Parallax -- which is much more error prone as the stars go further. Radial velocity is measured by blue shift. And then distance plus angular velocity are used to get the other velocity components. So estimates are more prone to error with more distant stars. $\endgroup$
    – HopDavid
    Jun 2, 2014 at 15:18
  • $\begingroup$ So distances don't say everything, it's the delta-v that matters for exchanging stuff like comets. It might as well be easier for a more distant star to happen to launch a comet our way that our Sun is more likely to trap. $\endgroup$
    – LocalFluff
    Jun 2, 2014 at 15:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .