Not only this, but they fall in a randomly-distributed fashion. Most satellites are still under control and steered to a predetermined area for reentry over the South Pacific.
> Most satellites are still under control and steered to a predetermined area for reentry over the South Pacific.
This is not true, most satellites do not perform controlled re-entry. In order to perform controlled re-entry, you need to have enough thrust to perform the final de-orbit maneuver in a short amount of time before you get so low that you start to lose attitude control of the spacecraft and wouldn't be able to guarantee you are firing your thrusters in the correct direction. To do this in a short amount of time means you need a high thrust system, in other words a chemical propulsion system and not electric propulsion.
All of the Starlink satellites use electric propulsion to do their de-orbit so they are performing uncontrolled re-entry. NASA's Orbital Debris Mitigation Standard Practices (ODMSP) say that you can perform uncontrolled re-entry if the probably of casualties on the ground is less than 1 in 10,000. There are NASA analysis tools that are used to do that assessment of what kind of debris would survive re-entry. This is likely getting updated in the near future to be more conservative though as there was recently a piece of an old ISS battery they thought would burn up hit someone's house in Florida:
I thought I recently read about a law requiring a de-orbit plan for anything launched. I would assume if most of the mass doesn't burn up (Starlinks mostly burn up), it would require them to hit the South Pacific patch or similar.
That is closer than I expected, and is specifically 47% of the re-entry mass (which is what we care about here) is from controlled. I would expect if you were just looking at # of satellites the numbers would be much more skewed as lots of little satellites would probably burn up and its the big ones that we need to worry about surviving to the ground.
