back to article SpaceX Starlink sat streaks now present in nearly a fifth of all astronomical images snapped by Caltech telescope

SpaceX’s Starlink satellites appear in about a fifth of all images snapped by the Zwicky Transient Facility (ZTF), a camera attached to the Samuel Oschin Telescope in California, which is used by astronomers to study supernovae, gamma ray bursts, asteroids, and suchlike. A study led by Przemek Mróz, a former postdoctoral …

  1. Flocke Kroes Silver badge

    Great potential for new science!

    When lots of images have had streaks filled in by machine learning AIs searching those images for unusual objects will start finding new types of stars and galaxies only in the repaired parts images. ML corrections are fine for making pretty pictures but less good for discovering new objects.

    On a (rare) clear night I can see the Moon, one or two planets, perhaps a dozen stars and about five aircraft. I have yet to see the ISS or a fresh batch of Starlinks. The Perseids have excellent correlation with cloud cover wherever I am. I saw the Milky way for the first time when on holiday in Tenerife. That was utterly beautiful and I thoroughly recommend looking up if you ever get over 50km from a village full of street lights.

    I would like to thank Przemek Mróz for coming up with some proper numbers for the cost of Starlink and I would like to see similar work for other types of telescopes to weigh against the benefits.

    1. graeme leggett Silver badge

      Re: Great potential for new science!

      This is a study of actual effect to date. Which is bit late to determine the cost benefit ratio other than in hindsight.

      Expect more papers will come out along the line of "Starlink messed up things this much" but they aren't going to stop Starlink.

      Perhaps its a plan to push more observatories into space - they'll be needing rockets to put them up there. Cynic, moi?

      There was a study predicting effects.

      1. ThatOne Silver badge

        Re: Great potential for new science!

        > bit late to determine the cost benefit ratio

        Which is actually irrelevant because the benefit is in your pockets, and the cost is for some bearded weirdos you absolutely don't care about. Obviously said bearded weirdos will complain a bit, but that's progress, bitch!


        > Perhaps its a plan to push more observatories into space

        That only works for non-astronomers. Space telescopes are useful, essential for some wavelengths, but over all they have terrible limitations. They are terribly expensive (and thus very rare), impossible to repair or upgrade, lack flexibility (changing instruments), and make some observations impossible: You won't make a 12-month continuous survey using Hubble, it would cost you a fortune (and other astronomers would try to kill you!). But you can do this easily and very cheaply with some university's smaller telescope.

        Space telescopes are just one (1) tool. You can't just have a Torx screwdriver in your toolbox.

      2. awavey

        Re: Great potential for new science!

        Jonathan McDowell an astrophysicist at Harvard, has been writing papers,giving talks and simulating the effects of Starlink on astronomy from Earth for some time and his conclusions match those of these Caltech papers. SpaceX have mitigated the risks of Starlink, changed the designs and engaged with the astronomical community and it wont ultimately be that big an issue at the end of the day.

        The problem comes from the other mega constellations being proposed the OneWeb, the Kuiper the SatNet the Sphere.Lots of those have higher or different orbits so will catch more reflection of sunlight for longer in astronomical observations and will be entirely managed by countries who arent that bothered about international cooperation.

        But the public have only heard of Starlink, so the press frame the story that Starlink creates a problem.

    2. Anonymous Coward
      Anonymous Coward

      Re: I have yet to see the ISS ...

      Well, then I can't help but feel that maybe you probably haven't looked very hard. For example, even though London is not exactly a bastion of dark skies, if there are no clouds in the way, and you look at the right time, the ISS is very easy to spot.

      1. werdsmith Silver badge

        Re: I have yet to see the ISS ...

        The ISS is often the brightest thing by a long way in the night sky round here in nightglow streetlight UK, it’s very hard to miss. We also get very bright Venus and easily naked eye observable Saturn, Jupiter and Mars. I’ve watched Starlink deployments and back when it was still current I pre seen the space shuttle close to the ISS. All constellations are clear and distinct.

        With a telescope or binoculars the Jupiter moons and Saturn rings are easy targets. And I’m in a town.

        It’s the nebulae that get lost in the murk, really annoying. The Milky Way is also not getting through, but it’s fun enough to do planetary observing.

      2. Gene Cash Silver badge

        Re: I have yet to see the ISS ...

        It's only very easy to spot if you know what to look for.

        The first time, I needed an app to point me to the correct spot of the sky, as no one bothers to describe that you're looking for a non-blinking fast moving dot. It moved much faster than I expected.

        1. Eclectic Man Silver badge

          Re: I have yet to see the ISS ...

          There is a NASA web site specifically for spotting the ISS:

          1. GruntyMcPugh Silver badge

            Re: I have yet to see the ISS ...

            Indeed, during the Summer months, I usually check that to see if the ISS is coming overhead, so I keep an eye our for it if we're sitting in the garden. Once you've got your eye in, it's fairly easy to spot. I live near a busy airport, but the ISs is quite distinct, once you know.

  2. spireite Silver badge

    1800 ish starlinks, but what about the rest?

    Hang on a bit.

    They are coimplaining about Starlink sats screwing the obsrvations.

    According to NASA,, there are more than 27,000 pieces of space junk up there.

    If they can handle filtering the 27,000 (and I know some or the junk is small), then removing the 1800ish Starlinks should be a doddle!

    1. Filippo Silver badge

      Re: 1800 ish starlinks, but what about the rest?

      "some of the junk is small" -> not, some; more like nearly all. Also, I understand that Starlink satellites are in a particularly low orbit. I figure that makes them more annoying than the same thing a further 100 km away.

    2. thondwe

      Re: 1800 ish starlinks, but what about the rest?

      Starlink's boxes are particularly shiny with regular reflective surfaces - most space junk isn't and is also not in low earth orbit - so generally invisible to astronomers - more of a issue as hazards to other space craft...

    3. Kernel

      Re: 1800 ish starlinks, but what about the rest?

      Having a seen Starlink group pass overhead fairly recently, I can best describe it as being like seeing a commuter train passing across the sky with all the internal lighting turned up to max - too high, long and fast to be a plane and it changes from horizontal movement when overhead to vertical as it approaches the horizon.

      Worth seeing on a clear night.

  3. Chris G

    The only real answer is international, enforcable regulation.

    It needs to be based on benefits vs drawbacks resulting from proper studies and licences issued accordingly, of course that will never happen, so look forward to a sky full of billionaire's willy waving projects.

    1. Jonathon Green

      That and cheap, compact, really, really, really powerful lasers.

      Since some observatories already use lasers to measure and correct for atmospheric effects adding a small satellite killer to the package shouldn’t be too hard… :-)

      1. Tom 7

        I have been somewhat nervous of the idea of another Carrington event but apparently that would raise the atmosphere enough to knock a lot of Musk turds out of the sky so I'm not so sure now.

      2. Cuddles

        You may think you're joking, but by far the best way to get rid of space junk is ground-based lasers. You don't need as much power as you might expect to vapourise small amounts of an object, at which point you've essentially created an ion engine that happens to have the power source on the ground. And you don't need much thrust to alter an orbit enough to eventually not be an orbit any more - it's much easier to make the orbit eccentric enough that it touches the upper atmosphere than it is to raise or lower a circular orbit.

        There's no new technology needed; powerful enough lasers are available practically off the shelf, and the tracking is already done anyway. The only reason orbital debris is still a problem is entirely political - such a system would be the perfect anti-satellite weapon that could deorbit anything in LEO completely undetectably (dispersion becomes an issue at larger distances, so taking something out of geosynchronous orbit isn't simply a matter of taking more time or turning up the power).

        This might seem to raise the obvious question that if it's so easy, why doesn't anyone already have these set up as anti-satellite weapons. To which the obvious answer is - how would you know?

  4. Anonymous Coward
    Anonymous Coward

    One day I'll put a nanosatellite with a super bright RGB led into orbit and make it blink out subversive messages in morse code while flying over China just to piss off the CCP.

    1. werdsmith Silver badge

      Put a scanning visible laser on it and have it write messages on the cloud layers.

    2. jdiebdhidbsusbvwbsidnsoskebid Silver badge

      Somebody beat you to it. Fitsat1 was launched in 2012 from the ISS. Designed by a Japanese artist, it was a 1u cubesat with a few hundred watts of white LEDs and it would flash Morse code across the sky. Lasted about a year before deorbiting.

      1. sungazer

        There is nothing new under the sun ;_;

        Mine would have a lora antenna and an android app that could change the colour as it flew over tho, and a parabola and very good pointing to track big cities so it's EVEN BRIGHTER than an isotropic led :D

  5. Anonymous Coward
    Anonymous Coward

    Pretty pictures...

    Surely if you fill them in with AI then you don't have science anymore and it becomes art.


    Stack multiple images and avoid areas with streaks

    Long exposures with telescopes that move slowly to track a distant target get the streaks.

    Stack multiple image frames and avoid areas with streaks. The area affected per frame would be smaller when you increase the number of frames per final image.

  7. NanoMeter

    Well, thank you


  8. Martin Gregorie

    Satellite spotting hints

    Getting to know Heavens Above:

    and using it as your primary reference for finding satellites etc. would be a good start. This website has a LOT of information to help you find satellites, astronomical objects, eclipses, etc. It also tells you where and when the better known and/or bigger satellites will be visible once you've told the website where you are on the Earth. The satellites it provides data for include the ISS, Tiangong, Hubble and the X-37b (when its flying).

    Obviously you still have to work out if the skies are clear enough to let you see what you're looking for, but using:


    Zoom Earth:

    or simply looking out the window helps with that.

  9. Anonymous Coward
    Anonymous Coward

    Another AI brain fart.

    What's needed is a separate band (another image layer) to identify certain pixels as invalid, and alter the processing accordingly. Filling them in as "black" could create unwanted artifices in the conclusions (e.g., a black hole has been found in low earth orbit). Only AI would think of something as stupid as that.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like