Where does the solar system end?

The solar wind streams outward until its pressure no longer exceeds the pressure of the interstellar medium it is pushing into. The surface where the two balance is the heliopause, and inside it the environment is governed by the Sun rather than by the galaxy.

That is a physical boundary rather than a cartographic one, which is why it can be detected by an instrument rather than agreed by a committee. Pressure balance requires the plasma density to jump by a factor of twenty to fifty across it, which is what makes it something a spacecraft can measure.

In August 2012 Voyager 1 was about 18.5 billion kilometres from the Sun. Its particle detectors recorded a sharp boundary being crossed five times in roughly thirty days: ions from inside fell away, and galactic cosmic rays from outside rose.

The magnetic field, though, did not change direction at any of those crossings, which is what a passage into interstellar space was expected to look like. For a year the same data supported two readings: the spacecraft was out, or the spacecraft was in an unexpected layer of the bubble where particles could flow both ways.

For most of the field it was settled by a different instrument. When the Sun threw out a disturbance, the plasma around Voyager 1 rang, and the frequency of that ringing gives the density of the plasma directly. Two episodes of it, in late 2012 and again in 2013, measured a density characteristic of interstellar space rather than of the solar wind.

Voyager 2 crossed on 5 November 2018 at 119 astronomical units, about six years after its twin. This time a working plasma instrument sampled the interstellar plasma directly. Voyager 1’s had failed in 1980.

The plasma wave instruments on both spacecraft told the same story: a steep jump in density, about a factor of twenty at Voyager 2. Voyager 2’s first interstellar density was read at 119.7 astronomical units, Voyager 1’s at 122.6.

Two independent crossings of the same surface is the strongest thing anyone can say about the heliopause, and it is the entire in-situ dataset.

If the solar system is the region where the Sun’s wind dominates, it ends at about 120 times the Earth’s distance from the Sun, and two spacecraft have left it.

If it is the region where the Sun’s gravity dominates, it extends perhaps a hundred to a thousand times further, because comets that fall in from that far share the Sun’s motion and are bound to it. Voyager 1 is not close to leaving that, and will not be for tens of thousands of years.

If it is simply where the objects are, the answer depends on which objects, and the outer catalogue is still being filled in.

Two crossings sample a large three-dimensional surface at two points. The flows and magnetic field at Voyager 2’s crossing fit a blunt front, but two points cannot settle the shape of the whole. Models disagree about whether the heliosphere has a long tail like a comet, two lobes, or something closer to a croissant: the Boston University group’s model gives a short, croissant-like tail, and the Moscow group’s a long, comet-like one.

The disagreement is not aesthetic. The models also differ on how large the bubble is, and the bubble is what stands between the planets and the interstellar medium. One simulation finds that if the Sun passed through a cold, dense cloud two to three million years ago, which maps of the galaxy’s hydrogen allow, the bubble shrank inside the Earth’s orbit and left the planet in direct contact with interstellar gas.

No object has ever been observed in the outer Oort cloud. Oort inferred it in 1950 from a score of well-observed original orbits: the “new” long-period comets, he found, generally come from regions between about fifty thousand and a hundred and fifty thousand astronomical units, so the sun must be surrounded by a general cloud of comets with a radius of that order. Nineteen orbits carried the argument: twenty-two had been computed accurately, and three were set aside because their errors were too large.

He took the random distribution of their orbital planes and perihelia, and the preponderance of nearly-parabolic orbits, as consequences of that cloud being stirred by passing stars rather than as signs that the comets came from interstellar space.

Two bodies far beyond the planets, Sedna and 2012 VP113, are read as members of an inner part of it. That is a reading of two orbits, not a sighting of the cloud.

The inference is strong and the observation is absent. It is one of the largest structures anyone claims is part of the solar system, and it is known entirely by what leaves it.