Galactic Futures: Milkomeda and Beyond

Andromeda, also called M31, lies roughly 2.5 million light-years away. Its light is blueshifted: after correcting for the Sun’s motion, its radial approach toward the Milky Way is about 110 kilometers per second. That measures the component along the line joining the galaxies.[3]

At Andromeda’s distance, the angular changes are extremely small. Researchers must separate the galaxy’s overall motion from internal stellar motion and instrumental offsets. Updated Gaia analyses can therefore change the inferred orbit even though the basic observation—Andromeda is approaching—stays the same.[18]

A forecast starts with measured positions, velocities, and masses, each with an uncertainty. Researchers run many realizations of those inputs through a dynamical model. The fraction that merges is a conditional prediction, tied to that model and its starting measurements.

These percentages should not be averaged: the studies adopt different inputs. “No merger within ten billion years” also does not mean “no merger ever.”

The Large Magellanic Cloud (LMC) is massive enough to shift the Milky Way’s motion. Its effect on the Andromeda encounter depends on the orbital geometry: it suppresses mergers in the 2025 reference model but promotes them in the revised 2026 model. Triangulum, or M33, also changes the dynamics.[1], [2]

Both panels use the same reference positions. The curves sketch the companion’s motion relative to the galaxy on the left: solid teal for a closer passage, dashed violet for a wider one. They are not calculated orbits; panel areas do not represent probabilities, and a close passage does not guarantee capture.

A galaxy merger unfolds through gravity acting across enormous distances. Individual stars are tiny compared with the spaces between them, so direct star-on-star impacts are rare. Yet stellar orbits can change dramatically as the combined gravitational field changes.[4]

A spheroid resembling an elliptical galaxy is a familiar outcome of major-merger simulations. But the final shape depends on the orbit, the amount and distribution of gas, and the angular momentum that survives. A disk component can remain or regrow under suitable conditions.[5]

“Milkomeda” is therefore a name for a possible combined galaxy, not a precise prediction of its appearance. Its future gas reservoir is uncertain, so neither a spectacular starburst nor a completely quiescent remnant should be promised.

For the mechanics in more detail, see Collisions and Mergers: Drivers of Galactic Growth.

A new orbit does not require a collision with another star

In a galaxy merger, the Solar System could move onto a substantially different orbit around the remnant. Its location billions of years ahead cannot be predicted precisely, and some material can be thrown into distant regions or escape.

The important distinction is between changing the Sun’s galactic orbit and disrupting individual planetary orbits. A large-scale tidal disturbance can strongly reshape the galaxy while many compact planetary systems remain intact. Earlier merger simulations illustrate possible solar orbits; they do not establish a guaranteed destination.[4]

The Milky Way hosts Sagittarius A*, while Andromeda contains a substantially more massive central black hole. If their galaxies merge, the black holes may eventually form a bound pair. Reaching coalescence requires several stages of orbital evolution.[8]

A galactic merger therefore does not automatically imply an immediate black hole merger. The supply of stars on suitable orbits, gas dynamics, and the binary’s properties influence the delay.[9]

Gas reaching either black hole could also power an active galactic nucleus. Its radiation and outflows would affect nearby gas, but whether a bright episode occurs—and how strongly it changes star formation—depends on the available fuel and how the energy couples to the surroundings.[10]

The standard ΛCDM model combines cold dark matter with a positive cosmological constant, Λ: a form of dark energy whose density stays constant as space expands. If that description remains valid indefinitely, accelerated expansion increasingly separates structures that are not bound together.

A galaxy or a sufficiently bound group can remain together under its own gravity. It does not simply stretch in proportion to the expanding cosmic background. Bound clusters elsewhere can also survive as separate islands; the Local Group is not the only one.[11]

Light emitted earlier can still arrive from a galaxy after newly emitted signals become unable to reach us. Its image becomes increasingly redshifted and faint, rather than disappearing in one sudden moment. A recession speed greater than light’s speed is also not, by itself, the definition of an event horizon.[12]

Star formation requires gas that can cool and become dense enough to collapse. Over time, gas is incorporated into stars and remnants, heated, expelled, or held in a diffuse atmosphere. Aging stars return some material, so the supply is recycled rather than used only once.

Future star formation depends on that recycling, cooling, fresh infall, and feedback. Simulations find that heating by accreting black holes can strongly influence late star formation. Cosmic expansion alone does not set a universal date when every galaxy stops making stars.[14]

A declining birth rate can therefore coexist with a vast population of living, faint stars. Later, white dwarfs, neutron stars, black holes, and brown dwarfs become increasingly prominent among the surviving compact objects. A white dwarf shines mainly from stored heat; a quiet black hole does not provide ordinary starlight.[16]

Calling an old galaxy “dead” is convenient shorthand for little current star formation. It does not mean that all its stars have died, that all its matter has become black holes, or that its gravitational evolution has ended.

The source of a galaxy’s light can change long before its stellar system disperses. Black hole evaporation belongs to a far more remote regime still. The ranges below are illustrative scales under stated assumptions, not a single timetable for Milkomeda.

1. Billions of yearsLocal encounters and solar evolution

   A changing neighborhood

   Major encounters remain possible while the Sun ages into its later evolutionary stages. Their clocks are independent, and the Andromeda outcome remains sensitive to orbital inputs.[2], [7]

2. Tens to hundreds
   of billions of yearsIncreasing separation under persistent Λ

   A more isolated view

   Unbound structures become increasingly remote and difficult to observe. There is no shared disappearance date: distance, emitted light, and observational sensitivity all matter.[19]

3. Trillions of yearsThe longest stellar lifetimes

   The faintest stars endure

   The smallest red dwarfs can continue hydrogen burning on roughly trillion-to-ten-trillion-year scales. A quiet galaxy can retain living stars long after vigorous star formation declines.[15]

4. Around 10¹⁹ yearsAn illustrative galactic evaporation scale

   Gravitational encounters redistribute remnants

   Repeated encounters can give some objects enough energy to escape while others move inward. This classic order-of-magnitude estimate depends on the system; rapid ejections can also occur much earlier during mergers.[16]

5. Around 10⁶⁷ years
   and far longerIdealized black hole evaporation

   A theoretical, much later process

   A simple blackbody estimate based on Hawking’s temperature gives a nonrotating, solar-mass black hole an evaporation timescale of order 10⁶⁷ years. The lifetime grows approximately as mass cubed: a billion-solar-mass hole would take order 10⁹⁴ years under the same assumptions.[17]

10¹⁹ means a 1 followed by 19 zeros. These are approximate process durations or future scales; adding the universe’s present age makes negligible difference at the largest values.

The assumptions become more consequential farther ahead

Hawking radiation is a theoretical prediction; it has not been directly detected from an astrophysical black hole. The estimates above idealize an isolated, nonrotating hole, neglect continued feeding, and do not resolve the final quantum-gravity stage of evaporation.[17]

The ultimate survival of stellar remnants also depends on unknown physics. Proton decay, for example, has not been observed.[20] Dark matter’s long-term behavior and the nature of dark energy add further uncertainty. We cannot confidently draw an intact elliptical galaxy surviving unchanged into every later era.[16]

What can better observations improve?

More accurate proper motions and better measurements of halo masses can narrow the Local Group forecast. Improved treatment of measurement offsets is especially valuable because Andromeda’s sideways motion is so small on the sky.[18]

Studies of real galaxy mergers test how stars, gas, and black holes respond to encounters. Cosmological observations test expansion. Together they strengthen different parts of the forecast, while keeping a clear distinction between measured conditions and extrapolated futures.