Fri, 08/30/2019 - 12:51
In the future, probably after my retirement, I would like to contribute to citizen science with some CCD photometry work.
However I believe that my plans could change, I found this worrying news:
https://en.wikipedia.org/wiki/Starlink_(satellite_constellation)
So my question for the experts is: what impact will such a vast satellite constellation have on amateur CCD photometry?
and also in general on ground-based astronomy and even on associations like AAVSO?
Regards,
Alessandro
Of course, if their satellite system even happens, the effect on photometry will not be welcome. But some estimates of effect should be possible. We know their projected count, altitude, and thus velocity, so we should be able to compute Poisson probabilities of a satellite's sweeping any given image. Rule of thumb, low-orbit satellites like the ISS sweep about a degree of sky a second at zenith. But first, does anyone know how big these satellites are (to estimate their magnitude at ground)?
Some thoughts:
Really resurrecting an old thread here...
When it comes to photometry you're likely going to be okay. When their orbits are raised and they are in their final orbit locations the impact theoretically on visual astronomy will be negligible. Consider how much stuff is up there already that's been launched since the 60's.
Radio astronomy is another animal, but I don't know a whole lot about what freqs are used for radio astronomy and how close they are to what would be used for internet.
I know that meteorologists are concerned about it though...
About the magnitude here they speak from 4th to 7th:
https://www.space.com/spacex-starlink-astronomy-observations.html
About the number, the satellites of SpaceX alone could be from 12000 to 42000, but other companies such as Boeing are planning a similar constellation, so the number could rise.
A newbie question:
what happens if in a frame the track of a satellite covers the variable or a comparison star? do you have to break the photometric sequence in two (before and after this frame, as at the meridian passage with a german equatorial mount) or do you just have to eliminate the frame from the sequence?
In the meantime I have searched better on the net, and apart from many generic press articles I think the most interesting sources are two articles from these authoritative sites, the first one an IAU statement regarding the problems of optical astronomy:
https://www.iau.org/news/announcements/detail/ann19035/
and concerns about Kessler syndrome:
https://www.scientificamerican.com/article/the-risky-rush-for-mega-constellations/
Regarding radio astronomy, I listened in a postcast of a broadcast on the Italian public radio to an interview with the head of the Radio Astronomical Station of Medicine (Bologna, Italy), who explained that although the satellites will transmit in a highly directive way, with the beamforming technique, however, the problem of secondary lobes remains sufficient to blind radio telescopes.
Maybe a way out could come from the law:
https://www.scientificamerican.com/article/the-fccs-approval-of-spacexs-starlink-mega-constellation-may-have-been-unlawful/
Hi,
i use Muniwin. So I can only speak for this software. If a long sattelite trail appears over the whole frame, and it is not near the variable, compare or check star, the measurement is ok.
If the sattelite covers a selected star, the following is possible:
1) Muniwin cannot match the frame, so it will not be taken into account.
2) The Photometry ADU will be other than the rest of the frames. Maybe only the error bar is rising, if the sattelite trail is close to a star, if the trail is direct over a star, i think you can sort out the frame...
As the previous poster stated, the question will be how many frames of e.g. 100 will be affected...
wbea
The satellites are powerful enough to cause major problems for radio telescopes, but there are practical technical solutions to the problem. It's a concern, but it's a manageable one.
If and when the total number of sats in the Starlink and competitors' constellation reaches 40k and more, the math gets kind of easy: take any square degree of the sky at any given time and you are more likely than not to have one of the sats in your field, right?
Now, whether the sat is illuminated at that time is another question. They will be in relatively low orbits to achieve low latency in communication, which makes them bright when illuminated but less likely to be illuminated as they will spend more time in Earth's shadow.
Better still, photometry is only affected if the trail passed thru any of the target star or comparison/check stars vicinity. Typically these areas of the field that are actually used for photometry are a super-tiny fraction of the field.
I don't think anyone should get discouraged from engaging in amateur variable star photometry. At worst, we'll need to insert a better quality assurance step into our workflows to reject a small fraction of contaminated images, with a lot of potential for automation.
I would like to believe that there might also be a bright side to it for astronomers in rural areas or people connecting to remote telescopes .... however, I'm afraid, reading initial reports, that the main business case for these networks is high frequency stock market trading and military applications (as light travels actually considerably slower thru fiber cables compared to in vacuum, taking what appears to be a detour via satellites is actually speeding up communication over long distances on Earth).
HB
I made rough calculation for LSST. When maximum number of satellites would be in orbit and they would be distributed uniformly around the Earth, 12 of them would be in the field of view of LSST at any timeMOMENT. Now take typical exposure time of LSST and movement rate of ~1 deg/sec... The picture would be pretty sad.
Tõnis
But these are in very low orbits, hence in Earth's shadow most of the night. So we will have dark objects briefly occulting stars and comps at random. It might affect photometric accuracy slightly in those cases, and difficult to track down which star was occulted at what time exactly.
I agree, most of them occupy very low orbits and hopefully can't be seen during midnight. I wonder what about when Moon is up and they reflect moonlight. 8-meter very fast telescope can detect pretty faint objects...
That occultation possibility even did not came into my mind, but it is indeed highly probable.
Short analysis on fingers: when such satellite moves 1 degree per second and it's projected size in the sky is 1 arcsecond, then star is occulted for about 1/3600 seconds. Now if exposure time is 1 seconds, total collected flux drops correspondingly, so its ~0.3 mmag. In reality, those satellites are pretty small (reference: 1 meter @500 km is ~0.4 arcsec).
LSST standard exposure time is 15 sec, then one such dip would be in the order of 20 umag. In reality even smaller (because satellites are smaller than 1m).
Best wishes,
Tõnis
Agreed, occultation of target stars by satellites won't be a problem.
But the orbits merely being low doesn't solve the problem. During the summer at 39N latitude, I've seen ISS passes almost all night long.
Why is Starlink not considered pollution like any other?
Even itf it would be considered pollution,how would that change anything? If making tons of money and providing a service to the military is balanced against increased pollution, who would win? See...
HB
Now there's a definitive argument.
Not to forget occultations by birds and bats and swarms of insects and whatnot ...
interesting articles:
https://www.astropublishing.com/3FAM2020/46/
https://www.aanda.org/component/article?access=doi&doi=10.1051/0004-6361/202037501