I find the joining of galaxies interesting as the two eventually become ONE. On series I saw on a discovery channel stated we die. Everyone dies. What is it that causes us to die when the galaxies combine? Is it something colliding with the Earth from the other galaxy such as a planet? Is it the drastic element changes on the Earth such as extreme temperatures or flooding, etc. Do we have any percent of surviving on Earth?
There are two reasons why nobody will die when the Milky Way merges with Andromeda:
The merger won't take place for another 4.5 billion years, by which time solar evolution will have long since made Earth uninhabitable (and even if humans did colonize the outer planets, we'd have long since gone extinct or evolved beyond recognition).
The merger itself will be a slow and not very dramatic process - stars aren't going to collide, planets aren't going to be ejected from their solar systems, and while it will look dramatic from a distant galaxy, other than Andromeda looking larger in the night sky not much will change for any intelligent life that's around at the time.
However, there's one potential wrinkle. Both the Milky Way and Andromeda host supermassive black holes at their cores. We've never observed a merger between two supermassive black holes, and modeling the process is theoretically challenging (look up the "final parsec problem" if you're interested). When this does happen, though - long after the merger of the stellar portions of the galaxies - there would be a massive release of energy. Most of this will be in the form of gravitational waves, but if enough of it comes as high-energy radiation it could be bad news for any remaining life.
I am not a physicist. But if we take the hypothetical (suppose there is life on some planet when galaxies "collide") then I suspect the highest threat will be from impacts from Oort Cloud objects.
When galaxies "collide" the stars (and their planets) generally do not "collide" ... they merge. (There will be gases in the galaxies which do collide.)
If you pick on our own Sun and it's next nearest neighbor and you reduce the size to a scale model we can get our brains around ... we'll shrink the Sun to the size of a basketball. In that scale the Sun's nearest neighbor is the size of a foosball (1/7th the diameter of the Sun) and would be located roughly 4500 miles away.
So imagine that scale ... and a galaxy merger. You toss in a couple of sports balls to represent stars from that other galaxy. Given that a "basketball" and a "foosball" are located about 4500 miles apart... what do you suppose are the odds that any of your sports balls can manage to hit either of these objects?
That's a huge simplification to help convey the idea of how much (mostly) empty space exists between the stars in galaxies and our Sun is believed to be in a less densely packed area of our galaxy (other areas are more crowded.)
That doesn't mean there isn't anything to worry about. Our star is has an Oort Cloud of objects roughly 1/3rd of the distance to the next star. And that "next star" also has an Oort Cloud of it's own. If a star goes hurling through the space between those stars and THAT star also has an Oort Cloud of it's own... the orbits of these previously stable objects gets perturbed and you can trigger a lot of instability that could result in a new period of planetary bombardment.
So imagine now that trillions of Oort Cloud objects may come hurling into the inner solar system ... nearly all of which will miss us ... but it only takes ONE ... and there may be trillions of them.
One need not wait for a galaxy merger to worry about such an event. If a star in our own galaxy happens to come near then this could happen.
Check out (for example) Gliese 710.
Given the timescales involved, the increasing luminosity of the Sun as it ages will likely cause problems for the habitability of the Earth before the galactic merger.
If the Earth manages to avoid undergoing a runaway greenhouse, the main hazards associated with galactic collision would be supernovae as a result of star formation triggered by the gas and dust in the galaxies interacting, and the potential for an increase in the accretion luminosity of the supermassive black holes. The latter might cause problems at a range of a few kiloparsecs, at least according to Balbi & Tombesi (2017). The question of how beneficial or harmful an active galactic nucleus is, and to what range is debated, see Lingam et al. (2019) for a more optimistic assessment.
Direct collisions of planets or stars are very unlikely. Depending on how things unfold, there might be an increased frequency of stellar flybys at distances close enough to perturb the Oort cloud, I'm not sure whether or not this would noticeably increase the number of comets into the inner system.
Then again, I'm not saying we couldn't get unlucky and have something fly by close enough to destabilise the outer solar system and trigger planet-planet scattering (bad news for terrestrial planets: the gas giants tend to win in such situations), but it isn't likely to be the main concern.
If a show on the discovery channel said that, that show was wrong. There's no reason to expect the Andromeda collision to kill the Earth.
The Sun might have killed the Earth by then, but lets pretend that we've beaten the odds, we've put up a massive sun shield and we've survived on the Earth, perhaps using the energy of the sun shield to move the Earth. Discussions on why moving the Earth may be impossible are welcome on the world building site, but, a sun shield probably is possible with technology in the next 100 years or so . . . if we make it that long, but I digress.
So, Andromeda approaches the Milky Way. It grows slowly larger, in time it becomes the brightest star in the sky, but with a telescope there's a visible haziness to it, it's more of a white blob than a star. We're only seeing it's central core, like we see now, only it's a dim star now. With a telescope we could see more.
As it gets close enough we may see it's faint early glow over more and more of the sky. It'll never look like it looks on photographs. Those are re-created. It may, eventually, look a little like the white milky band of the milky way, though probably less white and more spread out, as we probably won't see it from the flat side.
As the main stars in the spiral arms begin to tangle, 3 things will happen.
First, keep in mind that the shape of the stars in the Milky way, outside of some more distant straglers is basically the shape of a pizza with a golf ball in the center. The Milky Way is a flat circular collection of stars. The gravity and tidal forces will begin to tug on and stretch out the flat pizza of stars that we're used to and initially the stars we see will get slightly more spread out and dimmer as Andromeda gets larger and somewhat milky in the sky.
The night sky, 5 billion years from now will probably be less bright than the night sky today due to a gradual decrease in the number of large stars and the fact that nearly every star in the night sky that we see is a large star. The more abundant small stars are much harder to see.
As the merger begins, we'll see an increase in star formation. The Milky Way and Andromeda are is full of gas clouds, and while stars won't collide with stars, gas clouds will collide with gas clouds and that will create star formation. As the galaxies merge, there will be a significant increase in star formation. Exactly what that will look like is hard for me to say. I don't think it'll change the night sky that much, as the stars will still far outnumber any star formation, so, many collapsing and coalescing and combining gas clouds will be visible by telescope, it probably wouldn't be that impressive by the naked eye.
We can't see the Milky Way's bright central core, even though it's dense with stars. We may be able to, at times, see Andromedas, which, when we are positioned right and passing under it as it approaches, it'll probably be, by far the brightest object in the night sky (after the Moon ofcourse). Astronomers will track the bright center of Andromeda, as we wouldn't want to fly through it. It would put on quite a show if we did fly through it, but it probably wouldn't be good for us.
Another thing that would happen is, today, stars fly close to the Earth, say, 1 light year or less, every million years or so. The thing about the Andromeda merger is, Andromeda will be moving towards us (or we'll be moving towards it), very fast. Much faster than stars usually move, relative to us, in our spiral arm now.
Stars move so slowly, relative to their distance that Kepler was able to look at 1700 year old star charts and only find the tiniest of differences. When Andromeda and teh Milky way's spiral arms begin to merge with each other, we'll see stars whizzing by us at about 10 times the speed that we see now, so it would only take 200 years or so, for people to need to update their star charts instead of 1700 years. :-) We won't see stars move across the sky, but the higher velocity of the andromeda stars will be of interest to astronomers.
As noted in Tim Campbell's answer, if we collide with an andromeda star Oort cloud (which could happen every million years or so), we'd want to keep an eye out for very high velocity Oort cloud objects.
With all these hypervelocity stars, interstellar travel becomes easier. Why fly lightyears to a star when we can just wait for a star to pass less than 1 light year from us. The andromeda merger might provide us with the best means to leave Earth and find a new solar system. . . . or maybe I'm getting ahead of myself.
A final point is that, while stellar collisions will be rare to non existant, stellar ejections will be common. All these stars flying around at such high velocities, a LOT of stars, perhaps our own, will get tossed away from the new Milkdromeda galaxy.
Lots of things could happen. A high velocity impact from another stars Oort cloud, or a close enough pass to a massive object to disrupt Earth's orbit, or we could pass too close to the center of Andromeda and get more UV radiation than we'd like, or we could get tossed from the merged galaxies entirely, or maybe none of those things. It's too hard to make any real predictions, and 5 billion years from now, our sun won't be a friendly star to be near anyway, so we should think about finding a new home, probably before then, anyway.
Still, I think the merger will be more fun then problematic, all things considered. It'll be a cool event for our two aging galaxies, assuming any of our distant offspring are still around to see it.
I have no idea if the gravitational wave ripples from the merger of the two supermassive black holes will affect us. That's above my pay grade.
Nobody will die when the Andromeda galaxy eventually collides with ther Milky way. Collisions of galaxies may look rapid and violent when you watch computer simulations, but these simulations are speeded up many trillions of times. The collision between our galaxy and Andromeda won't take place until billions of years hence, long after humankind has gone to join the dinosaurs. When it does eventually take place, it will last for billions more years and won't be a violent process. There is so much space between the stars that collisions will be extremely rare. Collisions between clouds of gas and dust will be a bit more common, but nothing for us to worry about.