Why is Starlink polluting the night sky a big concern if we have space telescopes?

• There's a lot of concern in the Astronomy community over the deployment of Starlink satellites. For a good discussion, see the related question How will Starlink affect observational astronomy?

But why is there so much concern over this problem, given that there are numerous space telescopes? Presumably a lot more will be launched in the upcoming years thanks to satellite launches becoming cheaper. Aren't space-based observations superior in the first place thanks to the lack of an atmosphere?

I understand this sucks for amateur astronomers but is it also a big problem for professional researchers?

Look at your link and see how few of those space telescopes remain operational. Now look again at the remaining operational space telescopes and strike the duplicates. Finally, look at how many of the remaining few are small in size. The number of large operational space telescopes is very small. And except for Hubble, once they're done, they're done.

You might want to read the answers to this Q. Space telescopes are extremely expensive and adaptive optics has gotten extremely good, is the short answer. Also everyone and their dog wants observing time, but a space telescope can only be pointed in one direction at a time https://astronomy.stackexchange.com/questions/26610/why-are-we-building-larger-land-based-telescopes-instead-of-launching-larger-one

It we could time-travel to pre-industrial society, we probably wouldn't survive very long, but we would be blown away by the night skies.

@EricDuminil you can also just go to the right part of the countryside, it's not THAT hard to get away from light pollution (though central Europe is one of the harder places to do so, but still easier than time travel), see https://www.lightpollutionmap.info/

@llama: That link is depressing. I'd need to travel 2000km to see a really dark sky.

@EricDuminil yeah, maybe even easier to go out to sea. But the dark blue regions will still be striking, so rural France or up in the alps

@llama: I've had great views from the atlantic ocean last year, but it's hard to use a telescope on the deck of a small boat. :-/ The worst part is that it keeps getting brighter and brighter. Last time I went to the Alps, they had LEDs with motion detector everywhere in the small village we were staying at. No electrical grid, no water, but bright LEDs everywhere.

There aren't "numerous space telescopes". In the infrared-to-ultraviolet range, there are *two* general-purpose telescopes: Hubble and WISE. Everything else either operates in wavelengths that can't be studied from the ground (eg. X-rays), or is designed for a specific purpose (eg. monitoring the Sun).

Ask yourself who is causing the problem (SpaceX), and who stands to gain from your proposed solution of more space-based telescopy (also SpaceX)...

Not only is going to space getting cheaper, building any telescope is getting cheaper. If you want to keep getting the most bang for your buck, a land-based telescope is still the way to go in many, many situations.

Basically, because we don’t have enough space telescopes!

Some things in life are trade offs between cost and benefit. If we all went back to the stone age, there would be less light pollution and it would be easier to do land based astronomy in more places. Thats obviously not a solution in the modern age, so we make a trade off with light pollution. Im sure a similar dynamic is at play with StarLink.

A telescope can only do one thing at a time. This one took Hubble months do to - https://en.wikipedia.org/wiki/Hubble_Ultra-Deep_Field - and could not have been done from Earth, so unless you want to wait ground telescopes are the way to go

Is it fair to say the sudden presence of 'moving stars' across the night sky is a concern for all who like to gaze up in wonder? Instead of wondering, we will all be mildly annoyed at the reminder of the extent of our interference. It's a bit like visiting the Serengeti expecting to see wild animals, but instead you see AI-powered cardboard cut-outs. Sorry.

• It's a problem because there are still lots and lots and lots of ground-based telescopes.

Ground-based telescopes are still (by far) the biggest optical telescopes, and the cost of space telescopes is prohibitive for many research projects. It will be a long time before a telescope anywhere close in size to the VLT can be launched.

Most space telescopes are specialist devices, observing in a particular part of the spectrum that is blocked by the atmosphere (so there are infra-red, ultraviolet and X-ray telescopes in space) Or doing a specific task (looking for exoplanets, or mapping the positions of stars)

Space is getting cheaper, but it will be a long time before it is as cost-effective as ground telescopes in the optical range. And so it will be a long time before all professional telescopes are in space.

So it is a problem for professional astronomers.

There is also the issue of space debris with more launches making the problem worse which in turn creates hazards when putting space telescopes in orbit and making keeping them safe non-trivial.

Not just lots of ground-bases telescopes. There are lots and lots and lots of them.

There is also the less-talked about issue of radio pollution - all those RF downlinks from the Starlink satellites will have an impact on radio astronomy. There is also the hard to quantify effect on planetary protection - virtually all of the current NEO discoveries (until/if NEOSM flies) are ground-based so blanking out sky surveys with sat trails make life on Earth (slightly) more dangerous. Big report came out recently via UN Dark and Quiet Skies for Science and Society following a big workshop on this which has more info

Yeah, and all the RF uplinks may also interfere with space based radio telescopes.

Uplinks are pretty much focused at the satelite in question, so they are rather not a problem. But the other issues still stand.

With that said, other than aesthetics, machine vision algorithms are wildly more advanced today than in the past - strategies to remove passing satellites don't really seem to be anything other than trivial. It's annoying because it requires land based observatories to correct for the new noise, but it doesn't seem anything fundamentally more disruptive than that - a mild annoyance.

@J... Not even remotely applicable. Light pollution is exactly that - a pollution. The telescopes are not ideal and a stray light lowers the signal to noise ratio. And the interesting things are always at the limit of the capabilities.

@fraxinus Yes, but satellites are not in your field of view all the time. Naturally you lose the data you would have gotten while the satellite passes through your field of view, but I don't see how it affects SNR otherwise. If you're imaging at a very high sensitivity and the satellite blows everything out while it's in frame, you still have all the data before the satellite passed by and all the data from after it passed by. Annoying to have to slice the data up in space and time, I agree, but it *seems* a mostly solvable problem? Maybe "trivial" was overstating it...

@J... Is it bad? Yes. Will it make all ground-based telescopes completely useless? No.

@gerrit In either case, Starlink isn't going anywhere, so it's either the end of ground-based astronomy *or* we innovate solutions around it. I feel science has bested harder challenges.

@J... Starlink isn't going anywhere, but it could try to reflect less light.

@gerrit By coming back to Earth for a retrofit? What you're suggesting is that they throw it away and start over. That's not going to happen.

@J... Starlink satellites have a lifetime of ~4 years. To keep the constellation, they must continuously launch new satellites. If they from now on paint all the new ones black, provide a shield, or similar, the problem will be much reduced 4 years from now. See this April 2020 Tech Crunch article. So in a way, yes, they are throwing it away.

@gerrit Which is already planned. At worst this is a temporary construction detour in the sky.

• To expand on the "space telescopes are expensive" aspect:

• Space telescopes cannot be maintained or repaired. This applies not
just to things like optics and
instruments, but also to space-specific equipment like gyroscopes and
thrusters (the James Webb Space Telescope has an estimated lifetime
of $$\sim 10$$ years, set by the supply of fuel for the thrusters it
needs to maintain its orbit).

The sole exception to this, of course, is the Hubble Space Telescope
-- but note that the repair/upgrade missions a) were incredibly expensive; b) were only possible because of its low Earth orbit; and c) are no
longer possible now that the Space Shuttle is retired.

• Space telescopes cannot use the most modern electronics, because they
require radiation-hardened chips to protect against the effects of
the much stronger radiation environment in space. Such chips are
generally several generations behind the current state of the art.

• Space telescopes do not allow for testing new instruments, which is a very important use of smaller ground-based telescopes.

An example from my own experience: in the early 1990s I worked as an assistant to some infrared astronomers at the Aerospace Corporation, who designed and built their own infrared spectrographs. To use one of these, we put it in a van, drove the van to an observatory, mounted the spectrograph on a telescope (e.g., one of the telescopes at Mt. Lemmon Observatory, or one of the telescopes at Lick Observatory, or even the Kuiper Airborne Observatory), and started taking data the same night. This is a kind of flexibility that's simply not possible with space-based telescopes.

So it's not just that we can only build the largest telescopes on the ground (which is true!), it's also that the existing stock of smaller ground-based telescopes allow for experimentation, development, and science that cannot be done with space telescopes.

At some point we might be able to build larger telescopes on the Moon than we can build on Earth. Probably not in my lifetime. And probably by then Elon Musk III will insist on providing Starlink to everyone on the Moon too. ;-)

• One thing I always like to add is that ground based telescopes benefit from being able to take huge amounts of data. The Vera Rubin Observatory will have a 3.5 Gigapixel camera. There are proposals to sometimes run it in a mode with 1 second exposures. So we're talking data rates of gigabytes per second. If you have dedicated fiber lines you can deal with that. But it's really hard to move that volume of data down from orbit, it would require multiple dedicated ground stations all over the world. Keep in mind, for science images you can't use lossy compression (no turning images into JPEGS!).

While I agree that starlink is a problem for some of the other reasons stated, the very same project requires building hundreds of ground stations around the world, capable of transferring data from and to orbit at what amounts to a total rate somewhere in the terabits/s range. So they could easily solve that problem by donating some of that capacity to science.

Sure, that would be cool. This is kind of explaining the current state of affairs. Since we don't have a big enough ground station network, there are no very large field of view space telescopes. Starlink is in very low earth orbit and the antennas point down. So you might need new hardware or extra satellites to make it work. There's potential, but I'm not sure it's all the way to "easy".

• Space telescope Astronomy research published in Science and Nature:
Ground-based telescopes (31.1%), spacecrafts (27.0%), space telescopes (22.8%). (ref)

Number of professional telescopes affected: >1050. (ref) representing tens of thousands of academic physicists, cosmologists, astronomers and other scientists.

Number of amateur astronomers affected: 200,000 to 500,000 scientifically valuable people

Cost of equipment affected: 30-50 billion of equipment.

Number of Starlinks visible in the night sky to everyday people: more than 100 starlink satellites will be visible to us when we look at the sky after sunset and before morning.

(ref)

You know how distracting moving lights are, If you look at a sleeping city with thousands of little lights, and 100 planes are flying over the city, it's confusing, you have to look away. The dawn and dusk will soon consist of looking at hundreds of low flying satellites when the first stars start showing up.
1-2 hours after sunset there will still be 20-50 obvious lights still visible, you'll have to look towards the dark side of the sky to only see higher satellites.

Hale bopp, shoemeker-levy, various exoplanets, galaxies and solar system events were discovered by hobby astronomers, and just two terrestrial telescopes on their own have found 5700 asteroids, comets, variable stars and other stellar object (Pan Starrs).

The research states 0.01 starlink lights per square degree, 1-3% of photos for narrow field telescopes will messed up at dawn/dusk, and maximum 30-40% of photos for a wide field 1 billion telescope being currently built.

You won't be able to use Starlink on any mobile phones without 5kgs of lithium batteries and a dish ( >90W Starlink / <1W 4g ), it costs 1500 euros in Europe for the first year (83 per month + 400 equipment) compared to faster fiber at 300 euros. 99.9% of developed populations already have 4g internet at home, while poor nations can't afford Starlink... no monthly prices have been announced for India or African countries.

Is it necessary? Perhaps it will be a 10 billion dollar flop because 40 million yearly customers are necessary at prices exceeding ADSL/4g/fiber and pre-orders are currently low (50k out of 40 million?).

Comments are not for extended discussion; this conversation has been moved to chat.

What is the Y axis on that chart? I'm guessing the top line is 100%?

It was a graph showing that 99% of wealthy people have 4g at home, 99% of Indians have 3g at 2% of the price of starlink, and 70% of Africans had 3g in 2020.

“while poor nations can't afford Starlink” – Starlink will be exactly as affordable as (some) people can afford, in every country. That's basically the definition of price: there's some ≈fixed supply of bandwidth for every country, and SpaceX will charge for that as much as they can, but no more than people are able to spend because unused bandwidth is wasted income.