That makes the very big assumption that the density of other things to hit doesn’t massively increase.

That makes a big assumption we can't remove the junk before the density is high enough

It is worthy of note that SpaceX is pitching Starship for active debris removal in space.

It is worthy of note from https://spacenews.com/upper-stages-top-list-of-most-dangerou... that in a list of the 50 most concerning objects in orbit, 78% are rocket bodies, and 80% were launched before 2000 when various mitigation strategies for space junk began to be improved.

There is therefore a Pareto principle at work here. We are tracking over 25,000 objects, but only a few of them contribute most of the risk, and we aren't adding as much to the risk as the number of launches would make you think.

Infinitesimal over 100,000 years? I'm not losing any sleep over it personally, but that's a long time.

The surface area of the sphere at 1400km is far greater than the earth's surface area. And satellites at that altitude are traveling much slower. Put together, the probabilities would akin to a handful of birds flying on earth, at random altitudes/places/directions, and two of them colliding.

> The surface area of the sphere at 1400km is far greater than the earth's surface area

It's only about 50% larger (1+1400/6371)^2 = 1.49

This! How likely would collisions be on a sphere the size of the earth? How do the probabilities change from low earth orbit to higher orbits?

"much slower" is still >7000 m/s

Yes but at that altitude their orbital period is about 1/2 that of low orbits, meaning that even if two orbits do cross paths, there would be fewer potential collisions in a given time. They are 'slower' in relation to size of the sphere they occupy.

90 mins at 400km vs 110 mins at 1400km

I guess that makes sense since the Earth’s radius is around 6400km so from 400 to 1400km is really a change from 6800 to 7800km which is proportionally similar to a change from 90 to 110 minutes.

Edit: I just looked up the formula for orbital period and it turns out period is proportional to radius^(3/2) (See: https://en.m.wikipedia.org/wiki/Orbital_period)

Yeah, that statement makes a lot of assumptions about the next 99,950 years.

That's true, but the chances of quite a lot of things happening over 100,000 years are fairly big.

Almost everything we do carries a small but non-zero risk of something which, if you stretched it over a hundred thousand years, would almost certainly be realized.

On the other hand, the average damage from a collision in space within 100,000 years is also pretty small.

I wonder how the odds stack up against the odds of a human being hit by a meteorite? Those kind of odd's/statistics I can get my head around.

Suppose there are just two satellites orbiting at the same altitude, a bit higher, let's say 5 orbits a day (ISS orbits 18 times a day, roughly). The two satellites cross each other's path twice per orbit. Over the course of a year, that's 3650 potential collisions. Given the length of their orbit might be 50,000 km, then the average distance between these two crossings is 13 km. Over 10 years, that reduces down to 1.3 km. Over 100, 130m. Over 100000, it's 0.130m. They are gonna collide.

Of course, I am assuming a few things, like their orbits don't decay, they aren't exactly synchronized (i.e. they drift, and will eventually cross every point on the other's), but also that the uneven gravity of Earth doesn't corral them into narrower orbital paths (which it absolutely will, increasing the chances of collision).

That's for only two satellites. There are literally thousands. And when they smash apart, they can each generate hundreds more killer-sized fragments.

Kessler syndrome is a very real possibility on a long timescale.

We should be more careful the junk we put in space.