I'd be a lot more worried[0] if with that many satellites they were relying on "big sky, small satellite" as a collision avoidance system and never maneuvering. Their intentionally short lifespan combined with the move from xenon to krypton might just make maneuvering a kind of economic non-event.
I'd also be curious to see what percentage of those maneuvers were in response to their own satellites. Does anyone know if there is a raw data set for is available anywhere?
I know those but I was asking about the data referenced in the article. It says the consecutive FCC filings stated x and ≈2x avoidance maneuvers, but I was wondering if the data set regarding _maneuvers_ was available.
> Starlink satellites are making thousands of avoidance maneuvers as low Earth orbit becomes more crowded, feeding worries that a catastrophic impact is inevitable.
I don't understand how this is even slightly a proportionate statement, having read the rest of the article.
"SpaceX’s orbital communication satellites performed maneuvers just over 25,000 times in the six-month period between December 1, 2022, and May 21, 2023"
snip
"Of the 25,000 maneuvers, over 1,300 of those were to avoid debris generated from Russia’s November 2021 anti-satellite weapon demonstration test."
From the rest of the article, how is it not clear? 5% of their maneuvers were from avoiding something that was one solid satellite and is now thousands of smaller pieces of said satellite. If one of their birds is hit in a similar manner, it just became a larger shooting gallery. It's such a concern, there's a name for it: https://en.wikipedia.org/wiki/Kessler_syndrome
Edited to correct the percentage from comment below. Thanks!
It's not clear because they move when they calculate the probably of a collision is greater than 0.001%. So after moving, the probability is less than that. Using that as grounds for believing collision is inevitable seems disproportionate to the risk.
If they didn't move, then I'd say differently, as 25000 attempts at a 1 in 100000 is bad odds.
Starlink is in a self-clearing orbit -- if a Starlink satellite dies, it will deorbit within five years. Collisions with a Starlink only decrease that time -- collisions remove energy, the vast majority of debris will have a decreased perigee and deorbit much quicker. Those that don't will have an increased area/mass and also deorbit quicker.
Also Starlink is basically the only thing at those altitudes in volume. Every piece of debris from an exploded Starlink has two possibilities: 1 - it basically stays in its original orbit, not causing problems for other Starlinks. 2 - it changes orbit, only crossing the Starlink orbital plane twice per orbit. Given that most debris in #2 will deorbit within a single orbit a cascade is pretty much impossible.
IOW, it's basically impossible, and if the impossible does happen it will clear within days.
> If one of their birds is hit in a similar manner
Getting hit by a centimeter piece of debris is fatal to a satellite, but is not remotely comparable to being hit by a 100kg ballistic warhead, in debris-generating capacity.
>Do a search for "atmospheric drag" or anything about orbital lifetimes, gabbard diagrams [0] or the like
>nothing, instead it equates all constellations and satellites together as the same thing
Yet another click bait crummy article, with the obligatory "concerned astronomers" emotional tug on top but no serious engagement. And "big numbers!!" but no sort of context on if or why it's a problem. It's so tiresome how this gets trotted out over and over again. SpaceX's approach isn't bad for debris risks, it's good! The absolute best possible fail safe solution for sats is to have them low enough that they will passively deorbit very quickly without continuous boosts, or even better VLEO orbits where they must maintain aerodynamic orientation as well. Which in turn requires lots of sats that are actively replaced regularly, which in turn helps ensure updates can also be made rapidly based on real experience and they can become extremely refined. Sloppily throwing it all into some vague fear panic of "Kessler Syndrome" is not just wrong it's actively bad for that very reason vs clear differentiation between stuff at 300-550km vs 700+.
Also the implication that the utility function for earth-based astronomers is unlimited and they have some special right on top, ignoring the potential future paths for improvement and approaches that would ensure that (a small tax that would go directly towards a fund for space based telescopes say). And so on and so forth. There are real important discussions to be had which makes it all the more disappointing it's so very, very shallow.
Edit: Also to be clear it's not like orbital lifetimes are a new thing to think about. [1] is a classic old school webpage excerpting info from the 1987 "Satellite Orbits in an Atmosphere - Theory and Applications". It gives both very rough initial guidelines for thinking but then dives a bit more into how m/A and solar flux affect that a lot. Keep in mind when looking at the "LOW CIRCULAR ORBITS" section and graph what the m/A of diffuse debris looks like vs an original intact satellite (not that Starlink sats are big compact high weight spheres anyway).
> the utility function for earth-based astronomers
Nobody likes change. As time goes by the opportunities for astronomy will grow larger than what is lost. I'd be surprised if there isn't a backdoor plan within SpaceX to use the backside of their fleet as a giant outward facing array.
>As time goes by the opportunities for astronomy will grow larger than what is lost
I want to be fair to both sides here, as I sort of side mentioned at the end of my post I think it'd be perfectly reasonable not to take this as a given which will happen by itself. A lot of places have long had mechanisms like universal service funds, or other ways to effectively recognize that sometimes even extreme broad societal goods might create concentrated losers and that when the good is high it's also reasonable to help ameliorate that a bit. SpaceX making access to space enormously cheaper, higher mass, and higher cadence will absolutely open up incredible new opportunities for astronomy, but that doesn't mean said opportunities will be easy to take advantage of, astronomers have had to lobby and work hard for their existing installations and funding over the decades. And some of it even is also legacy from Cold War or boom times and they could very understandably fear that losing value in existing instruments will not necessarily then get an automatic replacement with new better ones in the current environment right?
But the answer there isn't to lobby to hobble Starlink (or other constellations) somehow which is likely not happening at this point anyway, but rather to quantify what's needed, form solid plans, and figure out a reasonable all around way to pay for it. Maybe governments start taking that into account directly as part of local licensing. Maybe a 0.5% (or reasonable) "replacement space instruments" fee could be added to megaconstellation services to go into a dedicated fund that will get to work planning and building Starship class instruments. Maybe SpaceX agrees to some number of launches for astronomy per year at-cost. Maybe some combo or things I'm not thinking of at all, I won't pretend to have given this massive thought. Working to reduce albedo, emissions and so on are all good anyway, but I do think existing terrestrial instruments (some more than others) will face a degraded environment. I also think that's extremely worth it, but that's no reason to be dismissive.
But I think the broad public conversation I've seen so far hasn't done a great job of engaging with this which risks the worst of both worlds, where existing interests do manage to slow things up without really helping themselves much in the end and simultaneously poison the water so that everyone is left worse off than might have happened.
> the worst of both worlds, where existing interests do manage to slow things up without really helping themselves much in the end and simultaneously poison the water so that everyone is left worse off than might have happened.
That is a really great way to put it.
> "replacement space instruments" fee
I realize you hadn't given this much thought but a downside to that approach is exactly how cell phone bills get padded. Regulatory capture fees ought to be rated as a backdoor tax like inflation.
Dodging makes it sound more dangerous than it is. They also have getting close to 5,000 satellites. Each one makes less than 1 maneuver a month.
Starlink is also way more conservative
> SpaceX’s satellites move when the probability of a collision is greater than 1 in 100,000, while NASA and other industry firms use a threshold of 1 in 10,000.
They’ve got a lot of satellites up there. I wonder if that’s just an adjustment to keep the odds of the company being responsible for a collision similar to everybody else.
Yes, all satellites are designed to do this, but it's the frequency that makes it interesting. If geo-sync satellites had to maneuver this often, you'd be hearing about it as well.
Compare that to the frequency of a geosync bird's maneuver requests. Probably doesn't see that much action in a life time. I'm sorry you're so bored with extremely large number of satellites like it's a normal thing, but it's clearly not. It's new. It's a complicated logistical problem that is interesting. There's way more going on than just speed up/slow down decision. If you slow down too much, then you have another bird come up behind pretty quickly as well as lose altitude. you then have to know how much fuel will be required to perform the maneuver. the sooner you decide to make the maneuver the more fuel efficient it is.
I'm sorry that you are unable to comprehend that I'm bored by the simple observation that more satellites in similar orbits means that the chance of collision is higher.
The logistical argument is something new you brought up because your original post claimed nothing but a simple frequency claim, so now you're trying to pivot
how many of these maneuvers are manually approved vs allowing the bird to make that decision internally? if manually approval, how many people are on staff to do that? also curious of how much they actually need to speed up/down to make the avoidance and for how long.
This is extremely old news. Here's SpaceX's explanation (included at the end of the Forbes article), which I think is quite cogent:
> Our Starlink team last exchanged an email with the Aeolus operations team on August 28, when the probability of collision was only in the 2.2e-5 range (or 1 in 50k), well below the 1e-4 (or 1 in 10k) industry standard threshold and 75 times lower than the final estimate. At that point, both SpaceX and ESA determined a maneuver was not necessary. Then, the U.S. Air Force's updates showed the probability increased to 1.69e-3 (or more than 1 in 10k) but a bug in our on-call paging system prevented the Starlink operator from seeing the follow on correspondence on this probability increase – SpaceX is still investigating the issue and will implement corrective actions. However, had the Starlink operator seen the correspondence, we would have coordinated with ESA to determine best approach with their continuing with their maneuver or our performing a maneuver.
I agree, specifically the part where they note that probability of collision was greater than 1 in 10k but that they were unable to address the situation (forcing an Earth observation satellite to maneuver).
edit: Maybe I'm being unnecessarily salty, but it's worth noting on their track record.
I guess this is just a matter of perspective. Nothing that you say is false, but it seems relevant to predicting future issues that this was ultimately due to a bug in their on-call paging system that was detected and remedied three years ago. The thousands of maneuvers they've executed since them without incident seem much more probative.
A related scenario to collisions that may be of concern is that, if LEO gets too crowded, the orbits may ossify and become finite real estate trading on the market. If players anticipate this happening, they might flood the orbits with the satellite equivalent of the parking pages that domain name traders use, and simultaneously push for a favorable version of orbit regulations/property rights. Essentially, an enclosure of the LEO commons.
This could prevent smaller players from working their way up and lead to a monopolization of space by a handful of private entities, imposing a tax on innovation/exploration in space.
> if LEO gets too crowded, the orbits may ossify and become finite real estate trading on the market
Increased orbital precision manufactures more space. And unlike spectrum, LEO is a volume. We're far from LEO getting full, though I agree there is need for an international allocation framework. (It would require bilateral talks between the U.S. and China, and so seems unlikely for the near term.)
As an initial matter, if the Commission updates its rules, it should adopt operational information sharing requirements that would enhance space situational awareness and collision avoidance for all satellite systems. For example, the Commission should require operators to register themselves and publish contact information on Space-Track.org to ensure that other operators can reach them in the event of a potential conjunction. The Commission should also require all operators to publish their ephemerides and reasonable covariance data...
Evenly applying requirements across operators would also curtail the regulatory gamesmanship that is unfortunately rampant in the Commission’s satellite dockets. This gamesmanship is encouraged when interested parties believe that they can force burdensome or restrictive conditions on their competitors while avoiding having these conditions applied to their own operations.
Unlike an effective performance-based metric, applying aggregate collision probability would arbitrarily and needlessly cap the size and capacity of responsible systems to the detriment of American consumers with no corresponding benefit for sustainability. These arbitrary caps would further drive satellite operators to license overseas while punishing U.S.-licensed operators and limiting their ability to effectively compete.
Interesting to note that a lot of the objects in those orbits are probably debris from their own launches because they have been very active launching stuff in those orbits. Also, they probably aren't the most stable orbits so that means most of the debris is recent too. Which points the finger at them as well.
> Starlink satellites are making thousands of avoidance maneuvers as low Earth orbit becomes more crowded, feeding worries that a catastrophic impact is inevitable.
By "catastrophic", do they mean the loss of 1-2 satellites, or are they alluding to the well-known debris chain reaction scenario?
Rather than say, the article switches to talking vaguely about some other concerns, and then wraps up with the final word being a PR statement that some concern is being taken care of in some unspecified way.
I suspect that the reader is left with the muddle-headed idea that there are some unclear and fairly innocuous concerns involving something about satellites, and a particular company is being very responsible in addressing these, in cooperation with experts who raised some issue, but other entrants to the market/space could be more a problem.
Also, the article doesn't say whose "a press release" is the one that the article uses to tell the reader what to think.
Whoever wrote this is a trying to get me upset that a company is being more cautious than it needs to be.
Instead, we should be encouraging companies to be cautious, rather than use the evidence of their cautiousness against them in a sad attempt to get rage-clicks.
Trying reading it again. It does no such thing, and instead is a cautionary warning that this will only become more of a problem for future constellations.
> SpaceX’s satellites move when the probability of a collision is greater than 1 in 100,000, while NASA and other industry firms use a threshold of 1 in 10,000.
The article also implies that the uptick is primarily due to a 15% increase in the size of the installation, which is bit silly.
There's loads of satellites up there, not just Starlink. How do rocket launches miss them? Yea, there's lots of room but everything is going really fast.
This is concerning to me, but not quite for the reason the article suggests. SpaceX is doing the right thing here by maneuvering and using Very Low Earth Orbits to cap the amount of time a failed satellite can stay up before falling. What if the maneuvers stop though? Say they run into financial issues and cease operations.
25,000 1-in-100,000 collision probability events in 6 months is actually pretty high. Without actively controlling the satellites to avoid you get:
1 - (1 - 10^-5)^(50e3) = 0.39...
chance of collision per year.
SpaceX says Starlink satellites should deorbit in about 5 years if they don't keep actively boosting their orbits. That still means a 90% chance of a collision before deorbit if they stop maneuvering. Those debris would be in low orbits and also decay quickly. You could still to some real damage to any of your neighbors in that time though.
From a regulatory perspective, I sincerely hope the FCC and others are thinking about what contingency plans need to be in place if SpaceX became financially unstable. They've basically guaranteed themselves a government bailout at this point. The UK did that to OneWeb for purely economic reasons. If SpaceX goes under, they're taking LEO with them.
> Say they run into financial issues and cease operations.
Then the last thing they'll do will be to deorbit the satellites.
You also missed a major factor in your calculations -- the odds that a collision will raise the odds of further collisions. And that's zero. The vast majority of collisions are with centimeter size objects, which aren't going to shatter a Starlink. And if you collide with something large enough to shatter a Starlink, then it will remove enough energy from the orbits of most of the objects that they'll deorbit within a single orbit.
It's true that the bigger question is "what is the impact of a collision?". For many of them, the answer is probably nothing. In an industry where 1 in 10,000 makes people uncomfortable though, I think it's enough be to be potential headache. Increasing density of LEO would make that worse. It doesn't need to be a runaway effect to be economically or especially politically problematic.
Proactively deorbiting at EoL is definitely the correct move, but requires enough runway to continue short term operations. I can absolutely picture this being used as political leverage.
I imagine if SpaceX was financially toast and trying to ransom LEO for a bailout then we'd probably just Defense Production Act them and give them a loan for only enough to pay the staff to deorbit the constellation.
It's probably true that energy is removed from the whole system, but I don't think it's correct to say that each piece of resulting debris has lower energy, some do gain energy, or lose very little.
Just to contrive a counterexample, what about a collision from a higher eccentricity object near its perigee? That could be adding a lot of energy I think, no?
Probably its time to declare the co-ordinates of starlink satellites as America's sphere of influence.
Maybe in future UN can come up with some rules for satellite orbits like they have for sea/oceans (EEZ).
Slightly side topic, but would a telescope on the dark side of the moon be shielded from all of earth’s radio and visible interference?
And if not on the surface, is there a Lagrange type point beyond on the far side of the moon that’s a stable observation point? I feel like the moons orbit is tilted enough that this point would also get permanent sunlight, while being shielded from all earth based interference.
58 comments
[ 2.8 ms ] story [ 99.5 ms ] threadI'd also be curious to see what percentage of those maneuvers were in response to their own satellites. Does anyone know if there is a raw data set for is available anywhere?
[0] and not _entirely_ surprised
https://platform.leolabs.space/visualization
https://www.n2yo.com/database
I don't understand how this is even slightly a proportionate statement, having read the rest of the article.
snip
"Of the 25,000 maneuvers, over 1,300 of those were to avoid debris generated from Russia’s November 2021 anti-satellite weapon demonstration test."
From the rest of the article, how is it not clear? 5% of their maneuvers were from avoiding something that was one solid satellite and is now thousands of smaller pieces of said satellite. If one of their birds is hit in a similar manner, it just became a larger shooting gallery. It's such a concern, there's a name for it: https://en.wikipedia.org/wiki/Kessler_syndrome
It's not clear because they move when they calculate the probably of a collision is greater than 0.001%. So after moving, the probability is less than that. Using that as grounds for believing collision is inevitable seems disproportionate to the risk.
If they didn't move, then I'd say differently, as 25000 attempts at a 1 in 100000 is bad odds.
Things having names isn't proof of much[0] (-:
[0] https://en.wikipedia.org/wiki/Dragon
Also Starlink is basically the only thing at those altitudes in volume. Every piece of debris from an exploded Starlink has two possibilities: 1 - it basically stays in its original orbit, not causing problems for other Starlinks. 2 - it changes orbit, only crossing the Starlink orbital plane twice per orbit. Given that most debris in #2 will deorbit within a single orbit a cascade is pretty much impossible.
IOW, it's basically impossible, and if the impossible does happen it will clear within days.
Getting hit by a centimeter piece of debris is fatal to a satellite, but is not remotely comparable to being hit by a 100kg ballistic warhead, in debris-generating capacity.
>nothing, instead it equates all constellations and satellites together as the same thing
Yet another click bait crummy article, with the obligatory "concerned astronomers" emotional tug on top but no serious engagement. And "big numbers!!" but no sort of context on if or why it's a problem. It's so tiresome how this gets trotted out over and over again. SpaceX's approach isn't bad for debris risks, it's good! The absolute best possible fail safe solution for sats is to have them low enough that they will passively deorbit very quickly without continuous boosts, or even better VLEO orbits where they must maintain aerodynamic orientation as well. Which in turn requires lots of sats that are actively replaced regularly, which in turn helps ensure updates can also be made rapidly based on real experience and they can become extremely refined. Sloppily throwing it all into some vague fear panic of "Kessler Syndrome" is not just wrong it's actively bad for that very reason vs clear differentiation between stuff at 300-550km vs 700+.
Also the implication that the utility function for earth-based astronomers is unlimited and they have some special right on top, ignoring the potential future paths for improvement and approaches that would ensure that (a small tax that would go directly towards a fund for space based telescopes say). And so on and so forth. There are real important discussions to be had which makes it all the more disappointing it's so very, very shallow.
Edit: Also to be clear it's not like orbital lifetimes are a new thing to think about. [1] is a classic old school webpage excerpting info from the 1987 "Satellite Orbits in an Atmosphere - Theory and Applications". It gives both very rough initial guidelines for thinking but then dives a bit more into how m/A and solar flux affect that a lot. Keep in mind when looking at the "LOW CIRCULAR ORBITS" section and graph what the m/A of diffuse debris looks like vs an original intact satellite (not that Starlink sats are big compact high weight spheres anyway).
----
0: https://www.youtube.com/watch?v=mQT5aMa_7iI
1: https://www.spaceacademy.net.au/watch/debris/orblife.htm
Nobody likes change. As time goes by the opportunities for astronomy will grow larger than what is lost. I'd be surprised if there isn't a backdoor plan within SpaceX to use the backside of their fleet as a giant outward facing array.
I want to be fair to both sides here, as I sort of side mentioned at the end of my post I think it'd be perfectly reasonable not to take this as a given which will happen by itself. A lot of places have long had mechanisms like universal service funds, or other ways to effectively recognize that sometimes even extreme broad societal goods might create concentrated losers and that when the good is high it's also reasonable to help ameliorate that a bit. SpaceX making access to space enormously cheaper, higher mass, and higher cadence will absolutely open up incredible new opportunities for astronomy, but that doesn't mean said opportunities will be easy to take advantage of, astronomers have had to lobby and work hard for their existing installations and funding over the decades. And some of it even is also legacy from Cold War or boom times and they could very understandably fear that losing value in existing instruments will not necessarily then get an automatic replacement with new better ones in the current environment right?
But the answer there isn't to lobby to hobble Starlink (or other constellations) somehow which is likely not happening at this point anyway, but rather to quantify what's needed, form solid plans, and figure out a reasonable all around way to pay for it. Maybe governments start taking that into account directly as part of local licensing. Maybe a 0.5% (or reasonable) "replacement space instruments" fee could be added to megaconstellation services to go into a dedicated fund that will get to work planning and building Starship class instruments. Maybe SpaceX agrees to some number of launches for astronomy per year at-cost. Maybe some combo or things I'm not thinking of at all, I won't pretend to have given this massive thought. Working to reduce albedo, emissions and so on are all good anyway, but I do think existing terrestrial instruments (some more than others) will face a degraded environment. I also think that's extremely worth it, but that's no reason to be dismissive.
But I think the broad public conversation I've seen so far hasn't done a great job of engaging with this which risks the worst of both worlds, where existing interests do manage to slow things up without really helping themselves much in the end and simultaneously poison the water so that everyone is left worse off than might have happened.
That is a really great way to put it.
> "replacement space instruments" fee
I realize you hadn't given this much thought but a downside to that approach is exactly how cell phone bills get padded. Regulatory capture fees ought to be rated as a backdoor tax like inflation.
Starlink is also way more conservative
> SpaceX’s satellites move when the probability of a collision is greater than 1 in 100,000, while NASA and other industry firms use a threshold of 1 in 10,000.
Compare that to the frequency of a geosync bird's maneuver requests. Probably doesn't see that much action in a life time. I'm sorry you're so bored with extremely large number of satellites like it's a normal thing, but it's clearly not. It's new. It's a complicated logistical problem that is interesting. There's way more going on than just speed up/slow down decision. If you slow down too much, then you have another bird come up behind pretty quickly as well as lose altitude. you then have to know how much fuel will be required to perform the maneuver. the sooner you decide to make the maneuver the more fuel efficient it is.
The logistical argument is something new you brought up because your original post claimed nothing but a simple frequency claim, so now you're trying to pivot
> Our Starlink team last exchanged an email with the Aeolus operations team on August 28, when the probability of collision was only in the 2.2e-5 range (or 1 in 50k), well below the 1e-4 (or 1 in 10k) industry standard threshold and 75 times lower than the final estimate. At that point, both SpaceX and ESA determined a maneuver was not necessary. Then, the U.S. Air Force's updates showed the probability increased to 1.69e-3 (or more than 1 in 10k) but a bug in our on-call paging system prevented the Starlink operator from seeing the follow on correspondence on this probability increase – SpaceX is still investigating the issue and will implement corrective actions. However, had the Starlink operator seen the correspondence, we would have coordinated with ESA to determine best approach with their continuing with their maneuver or our performing a maneuver.
edit: Maybe I'm being unnecessarily salty, but it's worth noting on their track record.
This could prevent smaller players from working their way up and lead to a monopolization of space by a handful of private entities, imposing a tax on innovation/exploration in space.
Increased orbital precision manufactures more space. And unlike spectrum, LEO is a volume. We're far from LEO getting full, though I agree there is need for an international allocation framework. (It would require bilateral talks between the U.S. and China, and so seems unlikely for the near term.)
Some interesting quotes:
As an initial matter, if the Commission updates its rules, it should adopt operational information sharing requirements that would enhance space situational awareness and collision avoidance for all satellite systems. For example, the Commission should require operators to register themselves and publish contact information on Space-Track.org to ensure that other operators can reach them in the event of a potential conjunction. The Commission should also require all operators to publish their ephemerides and reasonable covariance data...
Evenly applying requirements across operators would also curtail the regulatory gamesmanship that is unfortunately rampant in the Commission’s satellite dockets. This gamesmanship is encouraged when interested parties believe that they can force burdensome or restrictive conditions on their competitors while avoiding having these conditions applied to their own operations.
Unlike an effective performance-based metric, applying aggregate collision probability would arbitrarily and needlessly cap the size and capacity of responsible systems to the detriment of American consumers with no corresponding benefit for sustainability. These arbitrary caps would further drive satellite operators to license overseas while punishing U.S.-licensed operators and limiting their ability to effectively compete.
By "catastrophic", do they mean the loss of 1-2 satellites, or are they alluding to the well-known debris chain reaction scenario?
https://en.wikipedia.org/wiki/Kessler_syndrome
Rather than say, the article switches to talking vaguely about some other concerns, and then wraps up with the final word being a PR statement that some concern is being taken care of in some unspecified way.
I suspect that the reader is left with the muddle-headed idea that there are some unclear and fairly innocuous concerns involving something about satellites, and a particular company is being very responsible in addressing these, in cooperation with experts who raised some issue, but other entrants to the market/space could be more a problem.
Also, the article doesn't say whose "a press release" is the one that the article uses to tell the reader what to think.
Instead, we should be encouraging companies to be cautious, rather than use the evidence of their cautiousness against them in a sad attempt to get rage-clicks.
The article also implies that the uptick is primarily due to a 15% increase in the size of the installation, which is bit silly.
25,000 1-in-100,000 collision probability events in 6 months is actually pretty high. Without actively controlling the satellites to avoid you get:
1 - (1 - 10^-5)^(50e3) = 0.39...
chance of collision per year.
SpaceX says Starlink satellites should deorbit in about 5 years if they don't keep actively boosting their orbits. That still means a 90% chance of a collision before deorbit if they stop maneuvering. Those debris would be in low orbits and also decay quickly. You could still to some real damage to any of your neighbors in that time though.
From a regulatory perspective, I sincerely hope the FCC and others are thinking about what contingency plans need to be in place if SpaceX became financially unstable. They've basically guaranteed themselves a government bailout at this point. The UK did that to OneWeb for purely economic reasons. If SpaceX goes under, they're taking LEO with them.
Then the last thing they'll do will be to deorbit the satellites.
You also missed a major factor in your calculations -- the odds that a collision will raise the odds of further collisions. And that's zero. The vast majority of collisions are with centimeter size objects, which aren't going to shatter a Starlink. And if you collide with something large enough to shatter a Starlink, then it will remove enough energy from the orbits of most of the objects that they'll deorbit within a single orbit.
Proactively deorbiting at EoL is definitely the correct move, but requires enough runway to continue short term operations. I can absolutely picture this being used as political leverage.
And if not on the surface, is there a Lagrange type point beyond on the far side of the moon that’s a stable observation point? I feel like the moons orbit is tilted enough that this point would also get permanent sunlight, while being shielded from all earth based interference.