The other day, I watched a video on Facebook about Elon Musk’s plan to launch tens of thousands of satellites into Low Earth Orbit (LEO) to provide a low-latency satellite internet service to the world. Part of the latency inherent in current satellite-based internet is that the satellites must be in geosynchronous orbit, which is about 26,000 miles above the Earth. Simply waiting for the signal to get to the satellite, sent back to Earth, back to the satellite, and back to the user takes too long, even at light speed.
To make service universal from LEO, Musk’s company (Starlink, a division of SpaceX) has begun launching satellites that will form a mesh above the planet. Eventually, this mesh will surround the Earth. Amazon and others are planning similar services. Eventually, we could have 100K satellites in LEO.
The problem is that a failure or destruction of one satellite could result in a chain reaction failure (think Gravity) that would leave the Earth surrounded with high-speed debris. This would make space travel too dangerous to be attempted as any object passing to or through LEO would likely be impacted by this debris.
Is this a valid concern? Are we at the cusp of making our planet not only unlivable (via climate change) but also unleavable due to this debris cloud. I’m proceeding with the expectation that, if such a failure can happen, it most certainly will happen at some point.
[I’m starting in GQ because it seems to me that this has a factual answer, but I can see it moving into GQ or even P&E. The video, which I can no longer locate, says that there needs to be more government regulation of networks such as Musk’s. I’m not so sure regulation will work.]
Low Earth orbit (where Starlink will be - because of this concern) is where debris will still experience drag from the atmosphere, and thus be cleared out pretty quickly. Higher Earth orbit
would be a bigger issue.
At that said, there are other reasons to dislike Starlink
Given the massive failure of many companies and governments to provide decent rural Internet service, I like the idea of something like Starlink. I hope they can find a way to make it less obtrusive. I think these things should be included in agreements since so many tech companies are better at breaking things than fixing them.
A Kessler scenario isn’t triggered just by a satellite “failing”. If, say, the electronics in a satellite get fried by a freak solar event, it’ll still just keep on orbiting, in accord with Newton’s laws. What would trigger a Kessler scenario would be a satellite blowing up, which is not generally something that happens accidentally.
So, if I’m understanding correctly, debris in LEO is really not a problem because they fall to Earth fairly quickly. Thank you for that bit of information. I actually knew that, but it escaped my attention.
Let’s say Musk gets his network up and running. Along comes Amazon. Would they need to be careful to coordinate their planned orbits with Starlink’s orbits? Would they really have an obligation NOT to “accidentally” take out their competitors sats?
In a related notion, consider the scenario in the movie Gravity. The Chinese intentionally destroy one of their own satellites in a weapons test. The resulting debris cloud takes out a shuttle-type orbiter. It also destroys the ISS. Play that out for a while. How long would this hypothetical debris be an issue?
Despite that, collisions between satellites have occured. Notably, one in 2009 between an active Iridium satellite and a derelict Russian military satellite:
That one was at 789 Km altitude, which is not especially low. That’s about twice as high as ISS orbits, for example. As I understand it, Starlink and Amazon’s constellations will be at different altitudes to reduce the possibility of collisions.
Kessler Syndrome is a cascade of collisions causing increasing amounts of hazardous and often too small or irregular to track debris that can eventually deny access to or through the affected azimuths. It doesn’t have to be initiated by a satellite spontaneously “blowing up”; the collision of an inactive satellite (that has not been moved to a retirement orbit or deorbited) or even a piece of debris from a launch like an empty stage or separation hardware with another satellite could trigger a chain of collisions, especially in crossing orbits.
I’ve played around a bit with orbital mechanics and satellite position estimation to have a sense of the probability of this occurring naturally (without redirection) and given enough time the likelihood of a cascade collision is statistically almost inevitable. There is a lot of unexpected variability in orbits as they are perturbed by various influences and over a period of years the potential variance can be significant. The Cubesats and other pico- and nano-sats deployed in very low Earth orbit are very different because drag will cause small, lightweight debris to quickly deorbit so it isn’t a great concern. Debris in higher orbits will last for centuries or millennia and as of now there is no practical way to remove dead satellites or push them off into retirement orbits…
How much it will affect the ability to launch spacecraft or propagate to other orbits depends upon the extent and energies involved, but at the 550 km altitude of Starlink, that debris will be up there for centuries. I did speak with someone who does orbital estimation analysis for the then-Air Force and he expressed shock and dismay that this Starlink being approved. I didn’t get into details but the gist was he thought an impact cascade would occur sooner rather than later and the effects may be dramatic.
As a revenue generation endeavor for SpaceX, their posture on risk seems to be the typical, “Money now, worry later” approach which does not offer confidence in the protection of a common resource, the effects upon optical astronomy notwithstanding. This is the very “tragedy of the commons“ which plagues the desire to remove all regulation in the name of unfettered competition.
When some simulation reports e.g. 2% “max probability” of collision with Iridium-33 debris, 0.75% probability between a couple of Starlinks, etc., is it as bad as it sounds? Or are those methods too crude to predict the true frequency of collisions?
Given enough time, a collision is likely inevitable if their are significant perturbations in an expected orbit. How much time this is matters. However, given enough time and incentive, technology for removing clutter will likely improve. I’m not saying that this justifies the “money and que sera sera” approach seemingly favoured by the corporacracy and government enablers. But there are also benefits to a system, especially for a big country like Canada. I am not persuaded even a several collisions amount to a “danger to humanity”. However, the number of proposed satellites is of greater concern. Why not a cooling off period to monitor things and develop better methods of removing clutter? How many satellites are needed to try things out?
Space law is definitely pretty light at the moment. I don’t think that we currently have traffic laws or regulations in place.
That said, it is probably an area that will be growing. It would be necessary for people to coordinate their orbits in order to avoid collisions. We would probably need to deploy technology to clean up space junk and de-orbit defunct satellites.
The movie was a bit dramatized, and very little of it actually paid any attention to orbital mechanics as they actually are.
Even the worst case kessler syndrome doesn’t mean that anything that goes up there gets immediately bombarded with debris. What it means is that anything you put up there for a while will likely get hit eventually by something.
In any case, low earth orbit would be pretty clear in a few years or decade at most. It’s higher orbits, more than a few hundred miles, that the debris will stay a problem for much longer.
OTOH, the higher the orbit, the more area that orbit encomapses. It would take a few orders of magnitude more debris to make geostationary orbit unsafe than low earth orbit.
A few orders of magnitude? I was expecting something like hundreds of times more based on this comment, but according to my calculations, even when using the minimum distance for a low Earth orbit:
Surface area of low Earth orbit = 4 π (6371 km + 250 km)2 = 5.51 x 108 km2
Surface area of geostationary orbit = 4 π (6371 km + 35786 km)2 = 2.23 x 1010 km2
The latter is only ~40 times larger. What am I missing?
What I’m wondering is whether Starlink and similar projects are a problem for astronomers? Or can they easily ignore the satellites? (Incidentally, a friend recently moved to rural Worcester County, Massachusetts and his best or only choice for reasonably fast internet might just be Starlink. Seems odd that no terrestrial services are available there.)
Although those may not sound bad to a layman, those are huge probabilities in terms of orbital estimation, even though they are based on Monte Carlo perturbative methods that tease out edge cases. Orbital space is large, and while it is described as being increasingly “crowded”, that is more in terms of frequency spectrum allocations rather than physical space that they take up; however, if you had a 2% chance of impact per annum, that would be almost 19% over five years or almost 87% per century, which I think everyone would agree are relatively concerning and very likely, respectively. At a 550 km circular orbit the mean lifetime is about 17.8 years but there is obvious huge variability and it could vary by factor of 4 either way depending on solar activity and the satellite configuration. The more eccentric the orbit, the more pronounced thermospheric activity (like the kind that brought down the NASA Skylab prematurely) will have on orbital lifetime.
Indeed, it does not. There is the line of thinking that the problems we create today will somehow be solved by future innovation or otherwise just go away, and this persists in everything from industrial pollution to carbon emissions to nuclear fission power, and it inevitably creates a situation where we end up against a wall trying to figure out ad hoc solutions. The problem of mitigating orbital debris, regardless of method, is one of basic physics; you have objects moving at ~105, which is way faster than any bullet, and with momentum and specific energy that are enormous, enough to destroy any normal structure you might use to catch or deflect such an object. I’ve read through several dozen detail proposals (not PowerPoint decks, actual plans with analysis backed by simulation and estimations of technology readiness) on orbital debris mitigation, and not one of them would be capable of being implemented in the near term with extrapolations of existing technology. Even a crash effort to develop this kind of technology and system would be the work of a couple of decades or more, minimum, notwithstanding any political, funding, or technical roadblocks.
What consists of “space law” in its entirety are a handful of international agreements including International Telegraph Union (ITU) regulations on frequency spectrum allocation, International Civil Aviation Organization (ICAO) rules governing airspace, a few mostly ignored or not comprehensively ratified treaties, and national laws and regulations like those imposed by the FAA and FCC which have little practical effect other than imposing fines for violations. There is not a lot of effort to develop space law and engage in international treaties, especially since the rising major player of China has zero interest in cooperation (they are not a signatory to the Treaty on Principles Governing the Activities of States in the Exploration and Use of Outer Space, including the Moon and Other Celestial Bodies, i.e. the “Outer Space Treaty of 1967” which all signatories agree not to deploy nuclear weapons in space and to not use celestial bodies for other than “peaceful purposes”, as if we’re suddenly going to become pacifists in space) and no real means of imposing restrictions or claiming an overarching jurisdiction outside of the United Nations Security Council or the International Criminal Court, both of which are fairly arbitrary in how they can impose authority.
They can be compensated for, and it’s not too bad at this time, but with hundreds of thousands, it’s not going to be easy.
As it is, they have satellite trails often enough in their images. It’s not something that they don’t know how to deal with, it’s just that they will have to deal with a whole lot more.
It costs a lot to run broadband capable infrastructure. Rural areas are generally not profitable to do so.
That’s where the upside of starlink is. It brings broadband internet to places where it previously was not. A great benefit in this is developing areas, where instead of having to build a data link from a small town to connect to the global network, they just need to put up a little reciever.
It remains to be seen whether Starlink can actually provide the kind of bandwidth to a large user population (which has been a problem with other, higher orbit satellite Internet systems) but the argument that “it costs to much” to provide broadband Internet to rural areas is kind of nonsensical and exists only because we currently treat Internet access as a commercial service rather than the essential necessity it has become. We have electrified and provided telephone access to all but the most remote areas because it is deemed an essential service, so unless you live in a remote mountain cabin or out in Slab City, CA, you can get access albeit at some cost if your house is far back from a public road. The idea that we “need” satellite Internet is borne out of a public infrastructure failure to provide it, and furthermore puts such access in the hands of a for-profit corporate entity that could raise prices or discontinue service at a whim, notwithstanding the potential for space weather to degrade or destroy the network as I doubt the inexpensive Starlink sats have the same protections and S-rated components as the large telecom birds.
Well, it seems odd that one of those “developing areas” where Starlink might prove useful is the second-most populous county in Massachusetts, in a town with median family income north of $70,000.
You are missing nothing, rather the opposite: Geostationary is actually smaller, because LEO can cover the whole sky from pole to pole, while geostationary is only really useful at the equator, otherwise it is not geostationary, but from the ground seems to describe a line in the sky wandering north - south - north - south every 24 hours.
On the other hand, satellites in LEO move fast and in different directions, satellites in geostationary orbits hardly move with respect to each other, so they can be better calculated and placed closer together.
BTW: I believe high altitude balloons would be a much better solution, but they seem too easy for the ego of someone like Mr. Musk and his ilk.
