In four billion years, anyone left in our solar system will witness the most spectacular event to take place in the history of the night sky as the Andromeda Galaxy plows headlong into our own Milky Way.

But will that be the very last night sky our solar system witnesses?

See that fuzzy blob on the sky, the one just left of the Milky Way center in the constellation of Andromeda?

That's M31, the Andromeda Galaxy.

It's two and 1/2 million light years away and host to a trillion stars.

It has a beautiful spiral structure, spanning its gently rotating disk 220,000 light years in diameter, and a central bulge that hides a giant black hole that contains the mass of well over 100 million suns.

Andromeda is also racing towards our galaxy at 110 kilometers per second.

That faint blob will slowly grow to around half again its current size over the next two billion years.

Then its growth will accelerate.

At three billion years, it'll be two and 1/2 times bigger.

At three and 3/4, it'll fill half the sky.

At around four billion years from now, it'll crash through the Milky Way, and both galaxies will be utterly disrupted in the monumental collision.

This much we know for sure.

But what about the sun, the solar system, the Earth?

The Andromeda Galaxy was our first clue that there existed a universe outside the Milky Way.

We've known about it forever.

On the dark sky, it's visible to the naked eye as a faint smudge.

But being so far away, you can't see individual stars in Andromeda without a good size scope.

Because of this, there was originally no way to know whether Andromeda was a much smaller cloud of gas, a nebula inside our galaxy, or whether it was a galaxy in its own right at a much greater distance.

The distance and fundamental nature of "the Great Andromeda Nebula" was the subject of long debate, beginning with a Immanuel Kant.

In the mid 1700s, he hypothesized that Andromeda was an island universe, a vast sea of stars distant to our own.

It was a guess, albeit a very good one.

Milky Way philosophers living a few billion years from now won't have to speculate.

The galactic nature of Andromeda will be clear to the naked eye.

That galactic nature is also clear when we train modern telescopes on that faint smudge.

The first incontrovertible evidence came when Edwin Hubble calculated its distance by watching the pulsation of stars in Andromeda.

He observed Cepheid variables, which have a pulsation rate that depends on their energy output.

Time the pulsation rate, and you know how luminous the star is.

Those Cepheids appeared extremely faint in Edwin Hubble's observations due to the galaxy's great distance.

But knowing their intrinsical luminosity allowed Hubble to calculate that distance.

It was finally clear that Andromeda was, after all, an island universe far outside the Milky Way.

Hubble went on to combine distance measurements to many galaxies with measurements of their velocities to discover the expansion of the universe.

Those velocities were found by another astronomer, Vesto Slipher, by measuring Doppler shifts of spectral lines.

Almost all of Slipher's galaxies seem to be moving away from us.

Andromeda was a striking exception.

It's close enough that the mutual gravitational attraction between it and the Milky Way overcomes the outward expansion, allowing them to fall together.

But Doppler shift measurements only gives the line of sight velocity, the component of the galaxy's motion directly towards or away from us.

That doesn't tell us whether Andromeda will actually hit the Milky Way.

If the galaxy has enough sideways or transverse velocity, then it could miss us completely.

For a long time, we had no idea about Andromeda's transverse velocity.

It's actually very hard to measure.

The galaxy is so far away that its motion relative to background galaxies is almost imperceptible.

Even with a transverse velocity equal to its line of sight velocity, Andromeda's motion over several years, in terms of angle on the sky, would be minuscule, a fraction of a percent of the angular width of one of the Hubble Space Telescope's tiny pixels.

So how do we measure Andromeda's transverse velocity?

Well, we use the Hubble Space Telescope over several years, of course, with a heavy dose of being extremely clever.

A team of researchers, led by Roeland van der Marel of the Space Telescope Science Institute, did just this.

They mapped the locations of thousands of stars in Andromeda between 2002 and 2010 and compared them to background galaxies.

Then they averaged the observed motion of all of those stars and removed the effects due to the rotation of Andromeda and the motion of the sun.

They calculated a transverse velocity of 17 kilometers per second.

Even taking uncertainties into account, Andromeda is racing towards us much faster than it's moving to the side.

A head-on collision is inevitable.

Van der Marel and team also ran a computer simulation to study the consequences of this collision.

They used simulations of the gravitational interactions of millions of particles representing groups of stars and dark matter.

In other words, they made a little Andromeda and a little Milky Way in their computer and watched them smash together.

They also included the Triangulum, or Pinwheel Galaxy, the third-largest member of the Local Group.

This is an animated representation of the predictions of that simulation.

The giant spiral galaxies fall together, and the little Triangulum Galaxy joins the party.

The first impact in around four billion years completely disrupts the spiral structure of both galaxies, creating these amazing tidal tails.

We see these in other distant galaxies, like the Antennae, which are currently in the process of collision.

After slamming through the Milky Way, Andromeda's core travels on for a bit before falling back, and the two galaxies merge into a vast football-shaped elliptical galaxy in around six billion years.

Both galaxies contain a supermassive black hole, which will fall towards the center of the new merged galaxy.

They do that through a process called dynamical friction.

Gravitational interactions with stars slingshots those stars into larger orbits or even completely out of the galaxy.

Meanwhile, the black holes lose angular momentum and fall towards the center.

When those black holes are around a light year apart, they'll start losing orbital energy to gravitational waves.

Then they spiral towards each other and merge.

The resulting super supermassive black hole may briefly power a new quasar as it consumes any gas that also ended up in the core.

There's also a chance that gas throughout the galaxy will be shocked into a storm of new star formation.

This isn't completely clear because, in four billion years, both the Milky Way and Andromeda will have burned through a lot of their remaining gas reserves.

But what about the sun and the earth?

Well, for one thing, we don't expect any collisions between stars.

The average distance between stars is around 100 billion times greater than the average size of a star.

They'll slide right past each other.

There's a higher chance of another star passing inside Neptune's orbit, which might cause some gravitational disruption.

But that chance is still low, at something like one in 10 million.

No, our planetary system will probably survive this encounter.

One big question is where we will land when the new uber galaxy settles.

Van der Marel, et al's, simulation follows several candidate suns, simulation particles with similar orbits and masses to our sun, and they track their final locations.

Most end up in the outer parts of the merged galaxy, but many have orbits that periodically plunge them through the central regions.

And some even travel far enough from the center to make dashes through the Triangulum galaxy before that galaxy also gets gobbled by the giant elliptical.

There's also a small chance that the sun will encounter one of the supermassive black holes as they descend to the core.

And that could slingshot our solar system into intergalactic space.

But most likely, we'll remain within the system with front row seats.

So what will this look like to us?

Well, for around two billion years after the initial impact, our sky will be full of a galactic train wreck as the two galaxies settle down.

Finally, the giant orb of Milkdromeda will fill much of the sky.

At this point, the sun will already have expanded into a red giant.

And so we best be watching from the warm oceans of Enceladus or Europa.

Earth will long ago have been roasted by our own brightening and then expanding sun, which we talked about in earlier episodes.

I sometimes think how lucky we are to live in the time before our collision with Andromeda, a time when we have such a clear view of our dynamical evolving universe, when we have a neighbor whose visible stars revealed its great distance, and whose spiral structure helped us guess the shape of our own galaxy.

Astronomers in the distant future will see only a single featureless orb in the sky, and the next nearest galaxies will be very far and fast receding.

Will those astronomers ever figure out that there are countless other island universes stretching across a much vaster space time?