Musk loves his Starlink sat constellation – but astroboffins are less than dazzled by them The International Astronomical Union has warned against the rise of satellite constellations in Earth's night sky, such as SpaceX's Starlink system, since their brightness and noise could hamper future scientific research. “[We’re] concerned about these satellite constellations,” the union said in a statement this week. “The …
Nope, the Chinese have designs on the back side and have an active program right now. Elon is fixated on Mars. Blue Origin may beat SpaceX to doing anything substantial on the moon. If Elon was serious about the moon, he would have done well to send a package of tools and supplies there on the demo flight of the Falcon 9 Heavy.
6. Lack of beaches.
Most ground based telescopes have moved to remote islands apparently to reduce light polution, or at least that's what astronomers will tell you, but I've always assumed it was so they got paid to go on jaunts to places like Hawaii and the Canary Islands.
They're certainly always the most well-tanned people in a physics department.
If the space junk, including Musk's twelve thousand LEO comsats, is dense enough to force astronomers to move their telescopes to the lunar farside, then its very likely to be dense enough to make further launches of any other satellites, let alone support equipment and astronauts extremely dangerous or even impossible.
I notice Musk isn't saying a lot about how he intends to clean up space below his unilaterally chosen orbital altitude. With the enormous number of satellites he intends to put up there, this requires better than 99% certainty that dead satellites can be deorbited safely and on schedule and that none of them will collide with each other or with other orbiting objects.
A little Google goes a long way. These are LEO satellites, and the licence to launch them also included extensive requirements on deorbiting. And beause they are LEO, atmospheric drag will bring them down quickly upon request, and within 5 years I belevie if they are completely dead and cannot deorbit themselves.
I thought this had been mostly debunked over the last week?
Apparently modern astronomy tends to use stacked multiple short exposures, rather than one long exposure. Take one image, a satellite streak ruins it. Take 1000 images, a satellite streak ruins one, so you chuck that and stack the other 999.
Given that Starlink will eventually consist of thousands satellites, I can imagine you might have to chuck out a significant proportion of images.
Given that space telescopes tend to focus in on a tiny fraction of the sky, I think you'd probably need orders of magnitude more satellites flying around for it to cause a real problem.
I'd also expect a significant portion of satellites to be orbiting in a similar plane (close to equatorial), in the same direction as the Earth's spin (because you get that convenient boost from the Earth's rotation when launching). For that reason, trying to image things that fall in that plane would be expected to suffer more from the odd satellite getting in the way. As the orbital inclination goes up, you'd expect to get fewer issues - relatively few things are put into polar orbit unlsess strictly necessary.
If you're going from c. 200 satellites in LEO to 10,000+ when all the constellations are up, then you're getting 2 orders of magnitude more satellites.
You'd be wrong to expect them to orbit in the equatorial plane. This is only true of GEO satellites in very high orbits. LEO constellations like Starlink effectively create a net around the Earth so that they can maintain both continuous coverage of the surface with multiple satellites and inter-satellite line of sight communication.
These things will be orbiting around 20 km below Hubble, which typically points away from Earth.
Things above Hubble (and thence getting in its way) are much more likely to be in the (much) higher geostationary orbit, which is going to be equatorial by default. These are the things that are going to bugger up space telescopy.
I'm only talking about space telescopy here, since ground-based optical telescopy is pretty much redundant since Hubble went up, because of that pesky atmosphere being in the way, which is far more of an issue when trying to image things that are far away (or just small, Dougal).
Also worth noting that the 'next gen' James Webb is due to go up into the L1 Lagrange point, so nowhere near anything in Earth orbit...
...ground-based optical telescopy is pretty much redundant since Hubble went up, because of that pesky atmosphere being in the way,...
Ever hear of adaptive optics?
Or what about the 3 Extremely Large Telescopes? The 24m and 39m 'redundant' telescopes currently under construction, spending US$1.3B and US$1B respectively? Or the approved 30m at US$1.4B that's about to start construction?
That's a lot of money being spent on something that is nearly redundant.
The Hubble telescope doesn't have a lot of life left in it and there is still a lot of observing that can be done on the ground. If you are an astronomer (or a grad student) getting telescope time is a huge task. The bigger and better the telescope, the harder it is to get time and your best bet is trying to piggyback on somebody else's observations if you can. Sort of an issue when you are trying to finish off your thesis. JWT is so far behind schedule that you will only be able to count on getting anything from it is when it's finally launched and it's confirmed to work.
"...might have to chuck out a significant proportion of images."
It's too bad that there's no feasible way to have, like, a database of satellites. It might need perhaps about two lines of data, data fields that could be called "elements". Maybe the database, or file, could be called, "Two-Line Elements". That seems like a good name. They need one for each satellite.
One could imagine a digital computer system that would be able to predict the passage of satellites, by making use of the Two-Line Elements data set. The computer could be interfaced to the big silly telescope to know where and when it'll be pointed. They'd want to type in the Longitude and Latitude of the big silly telescope. The rest is trivial code.
Then the system could manage the exposure to preemptively skip the ruined images, if that's actually important to anyone.
"Short" exposures may be 15 minutes of one filter, there may be three filters so that's perhaps 45 minutes per colour exposure and for the dimmest objects you may need 100 full colour frames. That's 4500 minutes of exposure ... Bear in mind that in the UK you've already got to contend with clouds, planes, birds, meteors, dew, bad seeing and light pollution and you will realise that it's much more difficult than just "using 999 out of 1000" ...
So, the obvious questions. Given 12000 objects, how many extra 'streaks' will be on an exposure of that 4500 minutes over say a 5 degree field of view? Will it depend on latitude of the observer and location being viewed or will the net be more or less pole to pole? What will visibility of the satellites be like in final orbit and for how long before/after sunset? So many questions ...
I would think that with the technical innovation that goes into these devices, combined with the highly accurate 'flight path' of the these satellites, an algorithm could ensure that the exposure is timed to avoid visual interference and either pause the current exposure or not take an exposure at that time at all. Maybe even just mask the satellite digitally across the exposure in real time?
"an algorithm could ensure that the exposure is timed to avoid visual interference "
And that completely wrecks the whole field of amateur space photography, we can't afford that level of control over our systems, often its hours outside in the freezing cold at night with just a camera, a cheap telescope and a camera remote.....
So the universe will have check some service to time its movements. The funny thing about some astronomical events is that they occur at particular times and can’t be moved to wait because some wanker has put a bird in the sky.
Comments are all concerned with visual astronomy – radio astronomy may also be impacted.
Get a grip. Also see comment about the time required for some observations.
OneWeb is even more ambitious. Still the problem, known as the tragedy of the commons, is well understood. Unfortunately, as has been demonstrated many times, we usually have to wait until resources are either depleted or polluted so much as to be nearly irrecoverable before we change our behaviour; and so far I don't see any reason why this nascent commercial exploitation of NEO will be any different.
The only thing we can hope for is that this satellite-based internet turns out to be the next great white elephant. The demand for mobile telephony and data around the world is being met almost everywhere by low-cost, low-tech solutions.
SpaceX at least have a vested interest in the launch business and hence in avoiding space junk. They've also said they are addressing light and radio pollution issues. Some of this is already required to get a license to operate.
So, we don't have to wait, and aren't waiting.
One has to ask why the satellites are so reflective? Other than the solar panels surely they can be given some less irritating finish? I believe Musk has already said the next ones won't be so bad. So why were these ones made reflective? Did he WANT them to stand out like a searchlight? Or did they just not think about other people?
In space, the only way to transfer heat is through radiation (because vacuum means there's no fluid through which to conduct or convect heat). Since incoming sunlight is already radiation, the satellite can only do one of two things: absorb it in or reflect it away. If it isn't reflected away, it gets absorbed in by default.
"Sure, they're reflective. But they're also tiny. From the ground they'll just look like more stars. Nocturnal wildlife won't give a damn."
Aside from all the various nocturnal wildlife that uses stars for navigation. They're likely to give quite a bit of a damn.
Sure, if such animals exist. Colour me skeptical. Citation needed.
Aside from humans, I've ever heard of such a thing, but I happily admit lacking detailed knowledge of the intricacies of nocturnal animal navigation.
How would such animals cope with... oh I don't know... clouds? Plus, the night sky changes signifcantly as the Earth orbits the sun and rotates on its axis. In all cases the night sky is far from constant.
I suspect any animal using stars for navigation can probably cope. Filter out faster moving stuff as noise, focus on the more constant, static stars.
If this really is a thing.
The one with the satellite crossing the field of view - that has a date of 2018 so is nothing to do with the recent SpaceX constellation? Although it does demonstrate the kind of issue it could cause.
It's early, I've not had any tea yet, maybe I've missed something?
Now I really don't know if this is true and if I was more concerned I would try to research it, but the image from Hubble (which is just of 'a satellite' not these satellites) would be less likely to occur with Starlink due to the Hubble orbit being higher than Starlink. Therefore any views out to the universe wouldn't have this issues unless Hubble needed to capture an image at a very specific time that was near to the boundary of Earth (with hubble facing back almost towards Earth).
Would that be a reasonable assumption?
"Would that be a reasonable assumption?"
No. Hubble orbits around 540km. Starlink's outer 'shell' (I understand it has satellites at several levels) is in the 500km region, which could be a problem for Hubble. Some of the other constellations will orbit much higher and could cause much more significant problems.
And that's just for Hubble. Most astronomy is conducted not by professionals with multi-billion dollar ground and space telescopes, but by skilled and dedicated amateurs around the world. Who's going to compensate them for the loss of their dark skies or provide them with software to compensate for the satellites in their images?
I've ranted at my council when they changed the street lights, and they added baffles, not that they helped much, apparently safety at night is more important than me seeing the stars!
I suggested turning off the lights so people can learn that they can see at night without lights.... I think council workers have rarely seen a real dark night....
"Now I really don't know if this is true and if I was more concerned I would try to research it, but the image from Hubble (which is just of 'a satellite' not these satellites) would be less likely to occur with Starlink due to the Hubble orbit being higher than Starlink. Therefore any views out to the universe wouldn't have this issues unless Hubble needed to capture an image at a very specific time that was near to the boundary of Earth (with hubble facing back almost towards Earth).
Would that be a reasonable assumption?"
The problem, which also address previous comments about simply coming up with a tracking algorithm to avoid satellites, is that astronomers don't necessarily have much choice about where and when to look at things. If something interesting is happening in a specific direction at a specific time, either people can try to look at it or not. For example, there is a supernova detection network in which a whole bunch of telescopes all turn to look at an event as soon as something is detected; obviously we have no control over when and where an event will actually happen so if a satellite happens to get in the way that's just tough. Or for an example directly involving Hubble, the Deep Impact probe a while back fired an impactor into a comet with pretty much all the big name space telescopes watching. While that's slightly more controlled, you can't just sit around waiting forever until a comet happens to be in a good position for every telescope involved.
So even if all Starlink satellites are below Hubble, it's still a potential problem, even if it's not as bad as it is for ground-based telescopes. But on top of that, that's a rather big "if". In fact, all of phase 1 is planned to have the satellites in higher orbits than Hubble. The first set have been launched in a lower orbit due to concerns about debris, but it seems unlikely that will be the case for all of them, especially since most are planned to be over 1000 km, so reducing them down to the 500 km or so of these first ones would be a massive change.
On the other hand, Hubble is in such a low orbit mainly because of it being launched, and serviceable, by shuttle missions. Most space telescopes go much higher than that and, assuming it ever actually gets off the ground, it's direct successor the James Webb telescope won't even be in Earth orbit at all. So while this will remain a major issue for ground-based telescopes, once Hubble is retired it should be relatively minor for space-based ones.
I think most of the commentards up to now are missing the main point, it's not so much their visual interference, it's more about the fact that they are all broadcasting radio waves back in towards the planet.
For radio astronomy this is very bad, and cannot easily be countered.
Yes you can filter the frequencies used by the satellites, but the sheer strength of the signals means there will be a degradation of the ability of the radio telescope to pick up fainter signals from the cosmos.
Asking a genuine question as I don't know...
Is that an actual fact that they can't filter out that signal? Anywhere on earth has a lot more radio interference to deal with (man made and natural) than this and that is across all frequencies. Filtering out the specific frequencies regardless of 'strength' would appear to be trivial unless they were very close to the sort of frequencies being scanned by the telescopes. Combined with the fact that the telescopes are highly directional and the path of these satellites is known it doesn't seem that it would be too difficult to avoid.
It's not like the power of these at frequency x will smother all other frequencies. Satellite TV and Satellite telecommunications doesn't seem to have been cited as a significant issue previously, or has it?
Not to mention radio telescopes tend to get specifically sited in places with minimal background radiation. It's well-known, for example, that radio silence is enforced near the Green Bank Telescope in West Virginia: to the point you'd be better off Amish or Mennonite living near that thing. It even has a name: The US National Radio Quiet Zone.
That's not really background radiation. Those zones are used to avoid and block sources of RFI and EMI by being an valleys and not allowing certain devices in the area (microwave ovens, radio transmitters, mobile phones, wifi etc).
The background radiation as well as other RFI still exists, it can't be blocked so easily. However this is filtered by the radio telescopes which work in a protected bandwidth. Satellites should only interfere if they are badly made (it happens) and leak into this range.
"Is that an actual fact that they can't filter out that signal? "
Could be. Not my area of expertise, but we're basically talking about orbiting cell-phone towers. They need power enough in the downlinks to talk to cellphones with an antenna the size of your thumb that may be inside a building. Including phones that are being held wrong. Gonna need some signal power to handle that. The satellites are presumably talking to maybe a hundred cell phones simultaneously. And there are probably a dozen satellites visible at any given time (A guess, it'd take me hours to work out the exact number assuming that there is enough public data to do so).
So, maybe an order of magnitude or two increase in the radioastronomy background "noise"?
Maybe, and I emphasize maybe, there could be a real problem here.
Surely this was all addressed in the mission planning and the documents filed with the FCC?
"They need power enough in the downlinks to talk to cellphones with an antenna the size of your thumb that may be inside a building."
I wonder if I should be investing in tinfoil futures about now?
I could imagine it's possible for a satellite to broadcast down to a mobile phone on earth, but how is the phone going to talk to the satellite? Battery life measured in minutes?
Radio telescopes are highly directional and are located in valleys to shield them from terrestrial interference. Satellites are in their line of sight and broadcasting straight down the pipe and filtering can attenuate the signal, but it doesn't eliminate it. That makes it very hard to pick up something like the Voyager satellites transmitting single digit watts from beyond our solar system. It could also interfere with direct broadcasts from from Mars rovers.
My understanding is that modern radio telescopy doesn't use single dishes any more, but multiple dishes spread out over a large area (hundreds to thousands of miles). Presumably these satellites will be reasonably well directional in their broadcasts, since the power requirements scale with the square of the angle subtended, and nobody is going to design something to waste energy in an environment where every milliwatt counts.
With this in mind, will they really cause a genuine problem? Presumably telescope arrays already have mechanisms to filter out interference that only affects one or a small number of dishes, and such interference would only be an issue if it occurs at the wavelengths you're trying to pick up. Can anyone enlighten me with the frequencies these things will actually operate on and how it compares to those used in astronomy?
There are lots of radio observatories using single dishes for projects. It allows more researchers to run observations than if a site or multiple sites all had to be used together.
Radio telescopes are used in astronomy, but they are also used for communications with spacecraft and landers on other planets/comets/moons.
The multiple dish/receiver thing is interferometry - it's a technique that enables you to use two or more telescopes a large distance apart to get better angular resolution in your 'image' than you'd get with either of the telescopes used alone. So if one of the scopes involved in an interferometry setup gets borked by a sattelite signal, yes, it does bork the whole interferometer.
However, there's so many possible variables involved (size of scopes, wavelengths involved, how the scopes are being used (solo or as part of a multi-scope instrument), etc. etc that there is no entirely correct simple answer. But as a general thing - thousands of satellites in low earth orbit is a negative for astronomy in general. Personally I think these planned constellations of satellites are a solution to a problem that really ought to be tackled in some other way. But lets face it, humanity hasn't exactly got a great track record of doing things sensibly, now has it? (And I am not gifted with the knowledge of what best to do in all circumstances, either! 8-} )
"...transmit signals at a frequency that is read as noise by radio telescopes..."
If only there was some sort of technology where differing radio frequencies could somehow be distinguished, one frequency from another.
It's too bad that in our society, we must all share one single big fat radio channel, just like in a certain famously-bad Bruce Willis movie.
The trick here is that radio signal quality (for whatever purpose you need) is physically dependent on the frequency. Lower frequencies carry more easily but aren't as dense (they can't hold as much information). Raise the frequency and you trade off the former for the latter. For a satellite mesh like this, there's probably a "sweet spot" frequency that allows it to carry the necessary distance and transfer at a sufficient rate. If that "sweet spot" happens to be the same "sweet spot" radio telescopes use, you're going to end up with a lot of hand-wringing, as radio quiet zones can only be enforced terrestrially: on the ground and in the air, not necessarily in space. And it will be difficult for satellites like this to be able to avoid terrestrial quiet zones in their orbits.
Evidently you are unaware that transmitters have a spectrum that isn't limited to the desired transmit band. Nonlinearities, noise and so on get amplified in the final get out too - and they might be rather far from the desired center frequency and therefore in the desired "quiet" bands.
And when you're looking at a few photons per hour with a radio telescope, even a minute fraction of the many zillions per millisecond coming off a transmitter in the sky can cause issues.
While not as bright as the sun, someone calculated around 10^17 photons per second/m^2 from that source if they were all 1 eV (reality is bigger than that). So, reduce that by a few thousands, and then assume maybe a millionth of them are out of band noise. The number is still huge. One needs a sense of perspective and the world seems very short of fairy cake. 10^17/ 10^9 (say) is still a rather large number....
http://www.coultersmithing.com/forums/viewtopic.php?f=11&t=335&p=1945&hilit=photons+per+second#p1945
DCF proposed that I was "...unaware that transmitters have a spectrum that isn't limited to the desired transmit band."
Oooh. You mean numbers, like unintentional emissions in the radio astronomy bands at perhaps -90dBc. which is a whole bunch of divide by tens.
PS: High power satellites already exist.
There comes a point where the so-called interference is fairly and legally on the RECEIVER side to deal with. Otherwise it's chaos.
Not when it comes to radio telescopes, which by necessity (picking up radio emissions from light-years away can be a very hit-or-miss affair) are extremely sensitive. Think a radio version of a quiet room used to analyze someone's hearing. There's simply no other way to do it: ANY noise will interfere with the task at hand, just as ANY radio interference will disrupt the operation of that sensitive radio telescope. Which is why they often operate in radio-quiet zones enforced by law (terrestrially, at least). So, legally, interfering with equipment that requires a quiet zone falls almost-universally to the TRANSMITTER side to deal with, as the receiver has called dibs on radio silence AND holds the legal trump card of an enforceable mandate.
Chas 9 posted a series of misconceptions.
Where do I begin?
A well design RF front-end can provide both sensitivity and dynamic range. Communications engineers know how to do this. With sufficient dynamic range, interference remains in band and can be filtered.
Their local "enforceable mandate" doesn't include space, so is irrelevant to this discussion.
Signals from satellites are weak by any reasonable measure. Compared to nearby local emissions. Many orders of magnitude due to inverse square law
Radio astronomy already have their protected RF bands. They're simply not entitled to whine about emissions in other RF bands. Seriously.
If the satellites emit in those protected bands, then that's a specific issue to be addressed. Seems unlikely.
Radio Astronomers are apparently not always very smart about RF, based on the old story of then being confused and excited by pulses of microwave. Turned out to be somebody opening a microwave oven before stopping it, very short burst of 2.4 GHz. Radio Astronomers not informed enough to recognize "2.4 GHz" ? Seriously?
Thinks forgotton. These sats deorbit within 2 years maximum if they fail.
If SpaceX gets this up and running, it will be the making of SparShip (BFR) (Starlink is the proposed long term funding source for SpaceX). At which point it will make Heavy lifting so economical that we should move our future telescopes into space. A 100t to geo and a 9-meter faring make a very large telescope. You could pack the James web in without having to fold the mirror! A huge weight and faring makes designing a large space teliscope vastly cheaper.
Next up who doesnt want fast internet accesable gloally?
Forget the other consitlations appart from Jeffs. The only way to complete is to have a cheaper launch capacity, which is impossible unless you own the rocket manufacturing and can make rockets reusable and ten use only rockets somebody else has already paid for with previous flights. The economics of SpaceX really shine on this one.
"it will make Heavy lifting so economical that we should move our future telescopes into space"
It's never going to be economical enough to replace my little $2000 Celestron scope, even if the imaging capabilities are many times what my scope can do. That's my scope and I can use it whenever I want while out camping without booking in advance and I can point it at the sun or hit it with a hammer if I want to.
And even if you did replace my little Celestron with a space telescope a thousand times better at the same cost, It's not the same experience as using a little backyard telescope. There's no eyepieces to mess around with. No focus dial. There's something special about knowing that the photons entering your eyeball have been travelling for a million years. I know this as a fact - I have remote access to a couple of big telescopes but I prefer to use my own little 8" one in about 90% of cases.
"Next up who doesnt want fast internet accesable gloally?"
Well... I don't not want it, but I'm not super enthusiastic about it either.
But perhaps I just don't see the huge and obvious benefits to all mankind that being able to watch cat videos literally anywhere will bring. Could you list the advantages for me? What is it going to give me (or anyone, for that matter) that I don't already have?
I hear all this rhetoric about how we need to bring the Intarwebs to undeveloped nations and how it improves quality of life, but last I checked you can't download antobiotics or clean drinking water. Perhaps these people have more pressing concerns than how many likes their selfie next to the empty village well gets?
In the developed world, where Internet access is actually useful and important, it's pretty much available everywhere anyway. In my lifetime so far I can think of maybe half a dozen times totalling maybe 15 minutes when I haven't had internet access and something like starlink would have come in handy, and on zero of those occasions was it actually something important where a starlink-type system would have made a real tangible difference to my life. We already have things like EPIRBs and satellite Internet (which, yes, is admittedly horrifically slow and expensive) to cover those situations, anyway.
Now I'm not saying globally available broadband is a bad idea, but I fail to see a benefit which outweighs the potential loss to amateur/hobbyist astronomy, let alone professional and radio astronomy.
"I hear all this rhetoric about how we need to bring the Intarwebs to undeveloped nations and how it improves quality of life, but last I checked you can't download antobiotics or clean drinking water."
But ubiquitous communication systems would allow you to make arrangements for the above, especially in situations where time may be crucial. Remember, we're talking areas of the world where telephones (even cellular ones) aren't guaranteed, where you may be lucky to use some kind of semaphore system. It's either this or the (VERY expensive) satellite phone.
Put it this way. One thing that kept remote villages...well, remote...was the lack of communication between them. A lag time of even a day dovetailed into this isolation, which tended to complicate logistics for things that cannot be locally sourced: like water and medicines for some people.
ubiquitous communication systems would allow you to make arrangements for the above, especially in situations where time may be crucial
It sure would. So perhaps rather than spending millions launching a fleet of thousands of satellites into orbit which will interfere with astronomy, we should look into other options, like supplying all these remote villages with satphones. Or CB Radios. Or EPIRBs. These solutions are cheaper, easier to maintain/repair, and more durable than an android tablet. And they probably need less power, too. If these villages are so remote as to not have phones, then power is also likely to be an issue. Yeah, satphones are expensive, but we're talking about irregular calls that are for emergencies, meaning the calls will be short and not that expensive.
Or, perhaps, maybe if we're going to buy them all android tablets (if it's for general communication too rather than just important stuff then you'd have to buy one for every villager, otherwise you're going to create conflict), perhaps we should think about using that money to buy them water pumps instead? Or maybe a few km of pipe so they can build a water pipeline? you know, actually address the root problem rather than applying band-aids? I know, I know, you want the band-aid because it's totally trendy and shiny and also enables cat videos and a new untapped market for advertisers (what do these insane advertisers think dirt poor people in remote villages with no phones are going to buy?), but I tend to think that actually addressing the real issue is probably going to be more effective and more cost-effective.
Assuming, of course, that we're actually talking about helping people and not advertising to them, or spying on them, or exploiting them as cheap callcenter labour, etc etc.
Nobody has yet solved the space junk problem, old "scientific" satellites cause just the same problem and all the other junk left up there over the past 40 to 50 years. This does seem a little like our junk is fine but not anyone else's junk please. A proper solution is needed to regulate space and what is in orbit where and what failsafe solutions are in place should a certain collision be detected. Not one group of space users telling the other to back off.
And that already exists. This launch was licenced, and that licence included a large section on the deorbit requriements, as do most other launches nowdays.
Asd it happens, even dead, these satellites will deorbit within 5 years due to atmospheric drag. If they are not dead they can be deorbited much more quickly as they have on board SEP thrusters.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
