3 ms·

Maybe I am a bit naive, but I don't understand how this is a big problem. Sure it would have been a problem in the past where long exposures were captured to fi

Maybe I am a bit naive, but I don't understand how this is a big problem. Sure it would have been a problem in the past where long exposures were captured to film, where one satellite streak may overexpose the entire frame. But in the digital era, it will always be 100% known where the satellites were at a given point in time, so why can't software be used to exclude the affected frames and pixels for analysis? Surely that would not result in a very large amount of data being lost?

petschge 5y ago
Because telescopes are so sensitive that the reflections of a bright satellite inside the optics bascially ruin all pixels in the frame. And there is so many satellite that all frames close to dusk/dawn are impacted. So you loose 1/3 of your observing time. And all of the time in which you are geometrically most likely to find object on orbits that intersect the Earth orbit.
sebzim4500 5y ago
That works when there are only a few thousand satellites, or when you have a very low FOV, but once you have 100k+ satellites or you are doing a sky survey that becomes unfeasible. This interview goes really in depth on these issues. https://www.youtube.com/watch?v=-0w0dM_e9a8 https://www.youtube.com/watch?v=-0w0dM_e9a8
LeifCarrotson 5y ago
It's mostly solved for well-funded researchers with modern professional-grade telescopes and professional-grade analysis software. That's far from pervasive. "100% known" is overstating the situation, there's no easy-to-use API for an amateur to query location, date and time, alignment, frames, and pixels to get the location of satellites. Amateurs with basic software - or researchers who don't have the funding/equipment/software triad - just pointing a DSLR at the sky have to manually stack their images, for these users it's far easier to use a long exposure. The other problem is that sensor thermal and atmospheric noise vs. photon signal ratio is mitigated by long exposures in a way that's not the same as stacking short-exposure images, each with their own changing contribution to noise.
- PeterisP 5y ago
  If we're talking about the tradeoffs between feasibility of astronomy as a whole versus the interests of (potential) users of satellite internet, that may be an interesting debate, but if it's a situation where professional astronomy at major scientific institutions would be mostly okay but just amateur astronomers would be shafted, then I'm afraid that the debate would be very one-sided. Telecommunications is arguably more important for societies than astronomy - that might be debated, but they're definitely more important than amateur astronomy. Amateur astronomers are a sufficiently small niche with limited practical influence so that any society - no matter if democratic or authoritarian - would IMHO quickly decide that sacrificing the whole hobby of amateur astronomy as such would be a trivial cost to pay for even small practical gains in rural telecommunications (which affect far, far more people than amateur astronomers), otherwise it would be a tail wagging the dog. Amateur astronomers can have their voice heard - and taken into account if it's not too difficult (e.g. the changes in Starlink satellite coloring to reduce brightness is a good example) but they do not get veto rights. Their interests matter if there are no conflicts, but when multiple interests collide, the niche community will likely have to concede. You might convince societies that certain results of astronomy are essential and should not be traded away, but if these important results can be obtained even if "researchers who don't have the funding/equipment/software" are excluded, well, then excluding them is a plausible (and even likely) option and amateur astronomers will have to just take it and tolerate satellite streaks in their pictures.
orbital-decay 5y ago
Perfect exclusion is impossible. Even if you know how much photons you're going to get from a satellite, physics dictates you will actually get a close, but not necessarily exact number; this uncertainty alone can be many times larger than the entire number of photons you get from a faint star, essentially destroying the observation. Many estimates are talking about 1/3 of the entire useful observational time being discarded at many critical instruments, for only one mega-constellation of the Starlink size. >But in the digital era, it will always be 100% known where the satellites were at a given point in time This has little to do with the digital era, and a lot more to do with the Cold War. (and it's far from 100%)
- CrazyStat 5y ago
  > Many estimates are talking about 1/3 of the entire useful observational time being discarded at many critical instruments, for only one mega-constellation of the Starlink size. Can you provide a citation for one or more of these estimates? The only public estimate I've come across (not that I've been looking) was under 4% for Starlink and OneWeb [1]. [1] https://spacenews.com/radio-telescope-faces-extremely-concerning-threat-from-satellite-constellations/ https://spacenews.com/radio-telescope-faces-extremely-concer...
Robotbeat 5y ago
It has to do with design of the vera Rubin telescope imager. Because it uses CCD instead of CMOS (which has become useful scientifically in the last few decades), it is susceptible to light bloom washing out a whole row of pixels or more. They could’ve adjusted for the possibility of satellite trails (which have been a thing since Sputnik) and taken into account the possibility that satellite constellations would increase over time (after all, there are several LEO constellations in orbit already for over a decade and mega constellations have been proposed since the 1980s and 1990s), but they chose not to. Which is fine, perhaps. Other mitigations are possible.

petschge 5y ago

Because telescopes are so sensitive that the reflections of a bright satellite inside the optics bascially ruin all pixels in the frame. And there is so many satellite that all frames close to dusk/dawn are impacted. So you loose 1/3 of your observing time. And all of the time in which you are geometrically most likely to find object on orbits that intersect the Earth orbit.

sebzim4500 5y ago

That works when there are only a few thousand satellites, or when you have a very low FOV, but once you have 100k+ satellites or you are doing a sky survey that becomes unfeasible. This interview goes really in depth on these issues. https://www.youtube.com/watch?v=-0w0dM_e9a8 https://www.youtube.com/watch?v=-0w0dM_e9a8

LeifCarrotson 5y ago

It's mostly solved for well-funded researchers with modern professional-grade telescopes and professional-grade analysis software. That's far from pervasive. "100% known" is overstating the situation, there's no easy-to-use API for an amateur to query location, date and time, alignment, frames, and pixels to get the location of satellites. Amateurs with basic software - or researchers who don't have the funding/equipment/software triad - just pointing a DSLR at the sky have to manually stack their images, for these users it's far easier to use a long exposure. The other problem is that sensor thermal and atmospheric noise vs. photon signal ratio is mitigated by long exposures in a way that's not the same as stacking short-exposure images, each with their own changing contribution to noise.

PeterisP 5y ago
If we're talking about the tradeoffs between feasibility of astronomy as a whole versus the interests of (potential) users of satellite internet, that may be an interesting debate, but if it's a situation where professional astronomy at major scientific institutions would be mostly okay but just amateur astronomers would be shafted, then I'm afraid that the debate would be very one-sided. Telecommunications is arguably more important for societies than astronomy - that might be debated, but they're definitely more important than amateur astronomy. Amateur astronomers are a sufficiently small niche with limited practical influence so that any society - no matter if democratic or authoritarian - would IMHO quickly decide that sacrificing the whole hobby of amateur astronomy as such would be a trivial cost to pay for even small practical gains in rural telecommunications (which affect far, far more people than amateur astronomers), otherwise it would be a tail wagging the dog. Amateur astronomers can have their voice heard - and taken into account if it's not too difficult (e.g. the changes in Starlink satellite coloring to reduce brightness is a good example) but they do not get veto rights. Their interests matter if there are no conflicts, but when multiple interests collide, the niche community will likely have to concede. You might convince societies that certain results of astronomy are essential and should not be traded away, but if these important results can be obtained even if "researchers who don't have the funding/equipment/software" are excluded, well, then excluding them is a plausible (and even likely) option and amateur astronomers will have to just take it and tolerate satellite streaks in their pictures.

orbital-decay 5y ago

Perfect exclusion is impossible. Even if you know how much photons you're going to get from a satellite, physics dictates you will actually get a close, but not necessarily exact number; this uncertainty alone can be many times larger than the entire number of photons you get from a faint star, essentially destroying the observation. Many estimates are talking about 1/3 of the entire useful observational time being discarded at many critical instruments, for only one mega-constellation of the Starlink size. >But in the digital era, it will always be 100% known where the satellites were at a given point in time This has little to do with the digital era, and a lot more to do with the Cold War. (and it's far from 100%)

CrazyStat 5y ago
> Many estimates are talking about 1/3 of the entire useful observational time being discarded at many critical instruments, for only one mega-constellation of the Starlink size. Can you provide a citation for one or more of these estimates? The only public estimate I've come across (not that I've been looking) was under 4% for Starlink and OneWeb [1]. [1] https://spacenews.com/radio-telescope-faces-extremely-concerning-threat-from-satellite-constellations/ https://spacenews.com/radio-telescope-faces-extremely-concer...

Robotbeat 5y ago

It has to do with design of the vera Rubin telescope imager. Because it uses CCD instead of CMOS (which has become useful scientifically in the last few decades), it is susceptible to light bloom washing out a whole row of pixels or more. They could’ve adjusted for the possibility of satellite trails (which have been a thing since Sputnik) and taken into account the possibility that satellite constellations would increase over time (after all, there are several LEO constellations in orbit already for over a decade and mega constellations have been proposed since the 1980s and 1990s), but they chose not to. Which is fine, perhaps. Other mitigations are possible.