Starlink: Good Night to the Sky as You Know It

Elon Musk’s latest “world changing” new invention might well be the first one to really change the world – and fast. Our night sky is at risk.

What is Starlink?

This blog post is not going to shower Elon Musk in praise, I know people get riled up about that online, so there’s a warning about that.

Starlink (Official Site) is a low-orbit satellite internet service that has started launching. It is operated by Elon Musk. Now that it has launched, some people are beginning to realise some of the concerns that this kind of system might bring.

I may not be the best night photographer, and I don’t know all the constellations off by heart, but I love being able to look at stars. Having an unsullied view might not be possible much longer, if the Starlink project keeps going ahead.

An astronomer’s view

First off, I would suggest reading this thread of tweets by a NASA Engineer. It was what prompted me to write this post, and I think gives a good background on the main issues that have cropped up. I know it looks like a lot of reading, but it raises a lot of good points. I’ll summarise at the end, before going into my thoughts on the matter.

Note: At some point between writing the initial series of tweets, the author pulled them from twitter and made their account private. It seems they were not able to handle the backlash from Musk supporters, so I won’t name the author.

@twitter thread

Ok. Starlink’s kind of blown up over the past 48 hours. Let’s get some things straight. Despite their creator insisting otherwise they WILL be naked eye visible, for at least an hour, more like two, after dark and before dawn. In higher latitudes in summer…all night….

Here’s a great photo from @DrDasB – and a video from Marcin Loboz – both demonstrating just how visible they are. They will spread out over time, obviously, but visible they WILL remain. This is just 60 sats in one plane. Eventually this one plane will be a constant stream – one every 90 seconds. However, give it a week and they will be going overhead in the middle of the night and they wont be visible. – just like the ISS has times when it’s visible or not.

A trail of the satellites captured on camera – over time they will be just as visible, and more spread out.

HOWEVER – the plan ( see FCC filing ) is for, just in this 500km shell, a further 23 planes full of 60+ sats each. That means that there will be 2 -3 planes of spacecraft overhead at any time – which means they will never be a time when at least a couple of planes are not in view during that couple of hours after sunrise/before sunset when they’ll be visible.

I mean – that’s pretty obvious right? If you’re going to have global 24/7 coverage from LEO – you need a LOT of spacecraft, over everywhere, all the time. That’s what 66 sats in 24 planes gets you. The FCC filings show there are plans for MORE shells at higher altitudes. These will be a little dimmer, but they will also remain in the sunlight even longer after sunset/before sunrise. They, too, will never not be overhead.

So – the night sky, at the time when it is looked at the most, by the most people ( in the evening, an hour or two after sunset, before bed time ) is going to be imposed upon by Starlink. There will probably be 10-20 of these visible at a time. Professional astronomers will have a REAL headache on their hands. TLEs (orbital data) will exist for them all…but as these sats will be doing active collision avoidance and station keeping using their Krypton engines, the orbital data will be inaccurate very quickly…

So – when they’re overhead, astronomers will have to plot observations around them, with pretty large error bars in their location. In some cases, observations will just not be possible. Once there are 4 shells of these, there will be 40, 50 or more in the sky at once. I mean – @DrDasB‘s picture was 1am. from Brighton. That’s FOUR HOURS after sunset…and less than four hours before sunrise. Right now – a full constellation would be naked eye visible, ALL NIGHT, from the south coast of the UK. Just one example. Someone could do a numerical analysis and produce numbers down to the nth decimal place on just how much the sky will be impacted. Fact is – with just 4% of the first of 4 shells on orbit – we already know. It’s going to be really bad.

So – the night sky is forever compromised by Starlink. If you had asked an astronomer before this launch if it would…. they could have run the numbers and told you it would. But people are listening NOW…because it’s too late and the pics and vids are coming in. And after Starlink will be the other mega-LEO-constellations that will follow – each making the problem worse and worse. So – the question is – is worth doing? I mean- who cares, right? Epic satellite based internet for all! Get the world online! Information will save lives etc etc. The problem is – stealing the sky from the world to do it….well….shouldn’t we have been asked first?

If someone wants to throw up a cell phone tower, there’s an environmental impact study, public hearings….and if there’s enough objection – it doesn’t happen. Sometimes it gets a bit NIMBY (Not in my back yard) but with a LEO mega constellation – it’s in EVERYONE’S back yard. Is there another way to get the world on line? Should a company even be allowed to profit from the night sky in this way? Technological progress is a great and powerful thing – but when it will impact the globe – should the globe not be consulted?


If you love the night sky, go and see it now before it’s too late. Elon just opened Pandora’s box, and it’s going to shit on the night sky. And if you’re a ground based astronomer…..your job just got a lot, lot harder. Because profit comes first. I am disgusted.

End content @twitter thread

I am not an astronomer, so some of the issues talked about by this person don’t apply to me directly, but here are the main points as I see them:

  • Any additional space junk is bad for scientific observation
  • The current Starlink satellite cluster is producing visible light pollution in our night sky
  • This will only get worse as more clusters are launched
  • There isn’t any oversight for preventing this thing in the same way you would get on land
  • It’s being done by a private company, and more will follow if there isn’t sufficient backlash

My Views

I want to protect the night sky. Despite the amount of junk that’s up there right now, I think it’s important that we stop adding more to it. I don’t want private companies to litter the night sky with any kind of pollution, no matter how short-lived or naked-eye-invisible it might be.

Musk is by no means the only offender in putting up light-polluting stuff into the sky. There’s a Japanese start up that wants to sell people coloured meteor showers. A novelty which I’m sure will grow old fast. I’m not even much of a fan of fireworks, let alone man-made meteor showers.

But Musk is certainly the most high profile offender of late, and Starlink will be a much more permanent fixture in the night sky. I would also have expected him to know better.

I felt that the best way to explain myself would be to argue against some of the proponent’s views. I scoured twitter, and will attempt to debunk them:


Starlink is low-earth orbit, so won’t create space junk

Assuming that none of these satellites get hit with just the right angle of unidentified debris to knock them out of their orbit into a higher one (an unlikely, but not impossible outcome), the low orbit isn’t a concern for making more dead space junk.

But the issue here isn’t about dead space junk. It’s about junk that pollutes the currently visible night sky. When I look up, I want to see the naked sky for what it is, I don’t want to see random satellites that I never asked for and that don’t benefit me in any way littering my view.

Starlink won’t be all that visible anyway, so it doesn’t matter

Assuming it didn’t. Even the stuff that isn’t invisible could affect people, for example ground-based radio astronomers. A hunk of metal, even if it doesn’t reflect visible light, will reflect radio signals. I’m not a radio engineer, so I can’t attest to how much it would affect me.

But, evidence thus far seems to suggest that yes, it will actually be visible.

With current light pollution, it doesn’t matter

Aeroplanes, the ISS, Iridium flares, current space junk… There’s lots of stuff up there that already causes light pollution. What harm can more do?

I would quite like if we used fewer planes and urban areas produced less light (not least because of the environmental benefit).

Other satellite manufacturers, such as the iridium constellation, are moving to create satellites which won’t create visible flares. If they can do this, why won’t Starlink?

And all the current space junk that reflects stuff needs to come down anyway, so in an ideal scenario it wouldn’t be creating light pollution of any kind.

I reject the premise that I was ever fine and happy with light pollution to begin with. Adding yet more is causing insult to injury.

Stop being NIMBY, Starlink will benefit humanity

NIMBY is often applied as an argument when you’re building something like a wind farm, and locals who live nearby don’t want to see such things, so try and block it from happening. Even though it would have greater benefit for all.

This is a poor argument for many reasons:

  1. No-one was ever even given a chance to be NIMBY about this. On any ground based project you’d have a public consultation first so everyone could argue the pros and cons. With Starlink the first cluster is already up there, more are coming, and there’s nothing we can do about it.
  2. A wind farm or a power plant affects the view of a small surface of the planet, these satellite clusters affect the whole night sky across the planet.
  3. It won’t benefit humanity, it will benefit paying customers and the Starlink company. It in no way benefits me, so I don’t see why I should have to put up with it
  4. The cost (pollution of the sky) vs benefit of satellite internet access (internet access) isn’t worth it. It takes the defeatist attitude that we couldn’t possible improve coverage in a land-based manner.

Elon Musk is trying to do good, so it’s OK

No. Elon Musk is trying to make money.

But lets for a moment assume that he is completely benevolent, and that he was trying to do good and nothing more. What he has done sets a bad precedent. The lack of any kind of push-back against this makes it clear to other companies that our night sky is free for them to monetise as they see fit, no questions asked.

How long until we get clusters of low-orbit, quickly decaying, satellites that are organised to spell out an advertiser’s name? That’s a perfectly feasible next step.

How long until we get a company that puts higher-orbit satellites into use, stretching the envelope a bit further? If Starlink proves successful, that is a logical next step. And then there really will be a risk of space junk, not just light pollution.

Space is an important step for our economy

Is it though? Right now the only people who can really make moves into space are the already-rich. In my mind the world economy works best when there is a level playing field for anyone, no matter how big or small to compete on the capitalist market. But right now, with a few launch companies out there, and the cost of entry still very high, I don’t see how a space economy will benefit anyone other than companies that are already large and individuals that are already rich.

Perhaps this isn’t so much an argument about Starlink, but more for the expansion of our society into space. I’d be fine with that, as long as it doesn’t require the launching of hundreds of small satellites into orbit.

SpaceX is making space access cheaper

…And therefore Elon Musk should be allowed to do whatever he wants without anyone speaking out against him?

Starlink’s profits will be funnelled back into space development, and sure this may help to bring cost to access space down, but I am not in any way happy with how this would add to the general pollution of the night sky.

It is more than possible for SpaceX to continue to develop cheap access to space without resorting to launching massive waves of satellites.

If we want progress we have to put up with more satellites

I can respect this argument, in so far as what you constitute “progress”. If this means building space stations, research labs, operations to move to other planets, then yes, that’s fine. But these would be one-off machines in higher orbits, built to last, and built to service humanity for a long time.

The scale of light pollution, with entire constellations of these tiny communications satellites, covering the night sky just doesn’t seem to me to really constitute “progress”. Especially when all it brings is internet access, something that could be achieved on the ground.

It will give people in deprived areas internet access

It gives people in developing countries access to a network wholly owned by one individual company. I’ve seen enough of how other american network providers handle such monopolies to know that allowing them a nice captive market isn’t necessarily going to be better for consumers.

And like I’ve said above, if we really cared about giving internet access to these areas, we could achieve this in a more permanent fashion by providing monetary aid, or even shared engineering, to these countries to set up land based communication networks.

By offering grants or investments to local companies in developing countries, these areas could manage the network locally, instead of just funnelling cash back to a rich american. I would argue that that would provide greater benefit that Starlink’s proposal.

To Summarise

I think that for some people the sheer scale of what is being build might be missed. Our Engineer’s estimates suggest that a grid of 10-20 points of light in the sky may be visible at any one time once the full network is launched. Imagine that. Sure, it might sound a cool concept, but how quickly will that novelty wear off?

After decades with only the odd flare from an object up in space, we could finally be experiencing the end of clear night skies with these new far more permanent installations. It reminds me of a scene from the cartoon Futurama where the cast visit a remote country park to view the unpolluted night sky, and all Amy can focus on are the satellite flares. “The sky out here is amazing. Look at all those satellites.”

A grid of man made satellites, not just the intermittent points of lights from current satellites or aeroplanes, slowly rotating around the globe, ever present in our night sky? And for what? So some rich man can profit off of selling internet access to deprived areas?

I reject this idea as nothing but pure profiteering, and I think that it is incredibly unfortunate how easy it is for one individual to have the potential to make such a mark on the beautiful night sky of our planet.

Join the Conversation

  1. Thank you for such an interesting and thought-provoking article, with a very clear and logical argument throughout.

    I’m not a technology luddite at all (after all, despite its ease of use, Vivaldi users probably tend to be more technologically-minded on average anyway) – but I found myself agreeing with your objections throughout. I’m not anti-tech and anti-progress, but I firmly believe in the appropriate and beneficial use of technology, and disagree with the unnecessary use of technology for technology’s sake. This to me, is bordering on “technology for technology’s sake”, and looks very-much like an attempt to create a new market and then monopolise it. In other words, it looks suspiciously like it’s just for the money. I hope I’m proven wrong on that front. I doubt I will be, but I hope I am.

    The zombie-argument often touted for new endeavours like this (for instance, genetically-modified crops in the ’90s), is how it will benefit the 3rd world. The reality is that the firms that produce these technologies are solely looking to maximise profits and dominate the market, and don’t give a hoot about anyone not able to pay to licence their technology – so as a result, poorer countries don’t see the benefits because they simply can’t afford to.

    Plus, why on Earth do we need superfast internet all over the globe anyway? I’m sure a starving Ethiopian would rather have water and fertiliser before anyone addressed their desperate need for a 5g internet connection; I’m sure a Zimbabwean will not worry about zero lag on YouTube cat videos until they have first achieved full supermarket-shelves and a life free of the fear of government brutality. There are much bigger issues to deal with first.

    Finally, not every technology is the best or most appropriate technology for a particular situation. If I go to a bushmoot or out to some wild place, I take a basic 2g phone with me that is waterproof and has a month of battery life. I know where I have to go to get a signal, should I need one – but I rarely ever “need” a signal. If I am in a base-camp type of situation, I take a small wind-up or solar radio with me, for the occasions where I want to be updated with what’s going on in the outside world. In situations such as these, you don’t expect, want or even need an internet connection. You’re too busy with other things. An internet connection here is more of a distraction or toy that you can’t really spare much time to investigate. It would be a waste. Minimal communication technology is completely adequate. Should superfast internet be available even in complete wilderness, it will be a bit of a waste, as it just won’t get used. Why not spend the money on vaccines and farming tools instead?

    I’ve not mentioned the impact on astronomy or the beauty of the night sky, because I don’t think I can expand any further on your brilliant points – but just a quick note to say one of the first Christmas presents I asked for as a tiny kid was a telescope, and it always fascinated me and struck me with awe and wonder. I’d hate future generations to be deprived that excitement. Can you see footprints or a flag on the moon? Who lives on far-off worlds? One of the most powerful things in the world is a child’s imagination. Unfortunately it’s not as powerful as a large wallet.

    I never fully appreciated just how polluted our sky already is, until I was fortunate enough to go out walking, far-enough away from “civilisation”, a few years ago, to one of the few areas I can get to with low light pollution. I think I was about 15 miles from the nearest electric light, but it could have been further. I was completely bowled-over by seeing the sky as our pre-industrial ancestors must have seen it. I could see the milky-way, and finally saw why it was called that. I could also see the occasional satellite, but fortunately they weren’t too numerous (yet). The people who will lose the most night-sky from this project will unfortunately be in those locations where they will need it and benefit from it the least.

    I really hope the impact of this isn’t going to be as bad as your post implies.

  2. Example of the impact on ground based astronomy: The planned Large Synoptic Survey Telescope, which among other things searches for asteroids on an impact course, will have a very large viewing angle. This means that on every image of it’s 30 Gpx sensor there will be 1 to 2 satellites overshadowing important data at every time. This will result in a data loss of 3 to 10 Mpx in every single frame, which covers a significant area of space. 0.1 to 0.2% might not seem much, but space is huge and objects are tiny in viewing angle compared to that of the satellites. <- video taken with a Canon camera 50mm f1.8 lens – now imagine a 30m mirror of a telescope.

    This means (slightly exaggerated) the survival of mankind could literally be threatened by the planned 12000 Satellites of Starlink, the 684 satellites by One Web (6 of them are deployed) and the planned 3236 satellites of the Amazon's Project Kuiper. Samsung and others plan their own network.

    Additionally the full blown Starlink network adds 12000(!) objects to an orbit between 500 and 1000 (1500?) km, which is more trackable objects than are in space right now – but – can they guarantee that the none of the control mechanisms will fail?

    Of course not.

    Even if it doesn't fail because of a technical defect, there is always the chance of a 1 mm object hitting the vital part of a satellite at 30 km/s (e.g. the space shuttle got 10-50 hits on an average mission). An object comparatively that fragile like a satellite can't stand many hits in vital parts (thinking of the solar panels or in the vicinity of the steering nozzles ), and if it doesn't get hit, an object in that orbit is doomed to come down when not actively hold in position anyway (gravitational effects, solar wind). This might interfere with other satellites, I am curious how they will solve that problem. (click on the Iridium 33 orbits top left to see what that means)

    Furthermore affordable internet for all the people is (still) a pipe dream – just look at the countries with censorship (against western media), be it because ideological or religious reasons, it is easy for them to control deployment of the necessary transceivers.

    /stop text flood. I could go on for a long time 😀

    Well, In the meantime I will be sitting in my backyard waiting for the first piece of a satellite which I will nail on my wall as trophy … 😉

    1. Good points. I was aware of some of the other internet satellite projects, but didn’t know how many objects they were going to put up. For them, even on the order of 100s or 1000s seems excessive, and Starlink is going an order of magnitude further. Oof.


Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.