Random mass-loss kicks from the dying Sun eject a giant planet in nine of ten simulations

The new model keeps the Sun's total mass loss close to the figure the older calculation used. In the work Konstantin Batygin and Jim Fuller of Caltech and Fred Adams of the University of Michigan published in The Astrophysical Journal Letters ^[1], each of 4,600 ejections carries off about a ten-thousandth of the Sun's mass, roughly 33 Earths ^[6][7]. Multiplied out, the kicks remove 0.46 of the Sun's mass ^[17], close to the "about half" assumed by the calculation that gave the outer planets 100 billion undisturbed years ^[3][2].

So the comparison changes one thing: how the mass leaves. Each ejection shifts the Sun slightly and alters its pull on every planet ^[4]. Symmetric bursts might cancel out. Random ones leave small offsets that add up to large disturbances in the outer orbits ^[4]. Per kick, the Sun's speed changes by about 7 metres per second ^[7]. "The surprise is what happens when smoothness gives way to granularity: break the mass loss into discrete ejection events and the picture changes wholesale," Batygin told ScienceAlert ^[16].

The kick size was not picked freely. The team inferred kicks from Gaia observations of wide binaries containing white dwarfs ^[5], then ran almost 700 simulations at several kick levels, concentrating on the 48 with the most realistic mass loss ^[8]. In 37 of those 48, about 77 percent, the outer planets began crossing each other's orbits before the Sun had finished shedding its outer layers ^[9][18]. The earliest crossing came after only about 10 percent of the mass had gone ^[10]. Saturn may be ejected within a few million years in some runs ^[12].

"We lose them. In nine out of ten of our simulations, at least one giant planet is hurled into interstellar space," Batygin said ^[13]. According to ScienceAlert, in 90 percent of the models the system comes apart within three billion years of the Sun becoming a white dwarf, which puts the end less than 10 billion years from now ^[15].

ScienceAlert describes this as disruption 100 times faster than previously estimated ^[19]. I'd treat that number with caution. Measured from today, 100 billion years against under 10 billion is a factor of about 10 ^[20]. The report gives the 90 percent figure for "the models" without saying whether that means the 48 realistic runs or all of the nearly 700, and it does not show how the factor of 100 was reached.

The thing this doesn't tell you is what the Sun itself will do. These are simulations, and the kick strength in them comes from white-dwarf binaries around other stars ^[5][8]. A different reading of those binaries would mean a different input to the runs. Batygin said the result "dovetails with microlensing surveys, which suggest there may be as many free-floating planets in the galaxy as there are stars," and that dying suns "are plausibly a major source" ^[14].