Why can't observatories just stop capturing for a few seconds when Starlink satellites pass though their field of view?

Given that the positions and trajectories of the Starlink satellites are public, why can't the telescopes just ignore the photons they receive when the satellites pass through the field of view?

It seems that they could "just" stop image capture for a few seconds before and restart a few seconds after.

Is it that simple, or are there other complicating factors?

• Sep 14 at 8:53

Tl;dr, From personal experience, it’s not worth it. It’s a lot easier to throw away bad data than to try to calculate when stop. There’s not a lot of ambiguity as to whether it messes up the data.

As of right now, it’s not that big a problem either. Sure, some are particularly unlucky, as seen with a Magellanic cloud observation in early 2020. The concern is more the future: 12,000+ satellites (that are unusually bright compared to similar satellites) could cause problems in the future, for not only observational astronomers, but for the amateur observer’s night sky; just like telephone lines would ruin natural views, the fear is that too many satellites would do the same to our night skies.

There are other concerns as well, but what it boils down to is a lack of a way to deal with it. Right now, the occasional satellite coming into view messes up a frame or two of data, you consider it an unlucky break, and you move on with one less data point. There may be some state of the art observatories that consider whether a satellite will cross their observational area at night, but I have never worked at one that does. And that’s because it just isn’t worth it for how frequent a problem it is. Space is big; sometimes the area you’re looking at is really small, and between those two you’re normally OK for the most part.

To my knowledge, no (commonly used at least) archive possesses a collection of all satellites from all the different entities (government, businesses) and where they’ll be. Sure this information is accessible, but there’s quite a few entities on the Earth that have satellites in the sky; to have all of their info in the same place is a major undertaking. Now there may be some versions of this for those looking to put new satellites in space, regarding orbits etc, but there is not a tool that I’m aware of that has relevant tools for observational astronomers.

And right now we don’t need that. And honestly it’s a lot easier to toss out bad data than to try to weave intermediate stops into your tracking program for your telescope. But maybe some day we’ll have to plan observations so that we don’t waste too much valuable telescope time on bad data. Time will tell.

• It unfortunately can be difficult to do that. When you’re taking data, you’re observation is tailored to the object you’re looking at. Look at it too long, or next to something too bright in comparison to your object, the CCD (the thing collecting photons from the sky) becomes saturated, resulting in a loss of information. On top of that, depending what kind of observations you’re doing, you might be averaging several pictures into one, and some bad data points or, conversely, a lack of data points would mess up your average. The approach and difficulty can depend greatly on what youre doing Sep 14 at 7:39
• "But maybe some day we’ll have to plan observations so that we don’t waste too much valuable telescope time on bad data" - but in the meantime, I would expect there to be more satellites and therefore more of a problem, rather than less. :( Sep 14 at 15:46
• The solution is to make Musk feel so guilty he starts launching orbital telescopes for you.. If you plan this right you might even get an orbital observatory and rides to and from. Sep 14 at 23:07
• It is not stated in the answer, but it is important to understand that the telescope sensor senses all the added up photons over a certain amount of time. It doesn't sense individual photons so that you can just delete the photons you don't like from the dataset. Sep 15 at 8:09
• Note that 4 pi steradians divided by $\sqrt(12\,000)$ satellites gives a mean spacing of less than 10 mrad (about 1/2 a degree) between the satellites (when viewed from the center of the Earth)... This means, for random instance, the VISTA telescope will essentially always have one or more such satellites in its field of view (when pointed in their orbital band (That image is an orbital picture for just the first <2000 satellites.)). Sep 15 at 21:55