A Sky Full of Satellites: The changing orbital population and the impact of megaconstellations

CFA to work on spectral energy distributions of quasars which is a very successful thing and I got lots of citations out there he did as well but anyway um and he but he did most of the work and he pretty much stayed here ever since but with a little deviation to Huntsville for a year or two because he wanted to see a couple main engines being tested uh here he worked he came back to work on Einstein the old uh the very important Imaging next for Observatory and he started working on something which is he sounds incredibly boring but it's uh it's actually more important than almost anything any of us have done in astronomy and that is work on data standards so that you can have interchangeability and display radio on top of Optical on top of X-ray and all that began with the work he did when he was on Einstein but today he's not here because of his astrophysics uh so much as because of his peculiar Hobby um which is uh which goes back to before you're an undergraduate right oh yeah since I was about 13. okay where he got this strange obsession with uh tracking all human space objects literally everyone at one point you could just ask him what is this what's the name of this love that is orbiting he would tell you although it is deorbited since I believe uh I think it's got a little beyond that now right it's hard to keep up now yes but the result of this is it went from Hobby to Obsession to there's another level where he just bursts out to become an expert and in fact his database which you can see at Planet 4589 is uh named asteroid which has a lower number but nine nice uh it is actually better than the esa database or the NASA database or so far as we know the Department of Defense database is to technology which has the best public database at least in the world and he did all this on evenings and weekends and lots of vacations in Dusty archives all over the planet uh and uh it's a very strange thing to specialize in from my point of view but it's obviously done great things for Jonathan however suddenly it had nothing to do with astronomy but suddenly because of SpaceX starlink constellation suddenly astronomers are very interested in his database and or and the results that come out of it uh so his two streams of work and hobby and work have combined and he's going to talk to us today about A Sky Full of satellites and thank you very much Martin uh so yeah uh we are on the way to having A Sky Full of satellites and that has implications for uh astronomy particularly Optical astronomy from the ground and um uh when I was an undergrad uh summer intern at the Royal Greenwich observatory in the early 1980s my boss was fgs Francis Graham Smith uh who already at that time in a talk in 1980 published in 82 said the cumulative effect of an increasing number of long-lived satellites represents a very serious Hazard to astronomy and and that was when there were only 350 working satellites in orbit there are 20 times that many today so he was pretty prescient now we have a problem and I'm going to go through this uh this is just sort of my plan I'll introduce the issue I'm going to do a little digression into the demographics of the satellite population more generally to give you some context I'm going to do some simulations of how bad it's likely to get for an observer on the ground I'm going to talk about how we monitor what the space companies are actually doing I'm going to give you some good news about what SpaceX is doing to mitigate the brightness of these satellites and then I'm going to talk briefly at the end about the astronomy policy response and I want to thank this long list of collaborators notably Connie Walker who's really leading the effort here in the U.S Richard green doing a lot of work on policy Meredith rules on the observing side and a cast of uh uh some amazing characters around the world so I'm going to start off by talking about the problem more and more frequently you get these streaks on ground-based astronomy images and you know there are some are short and Summer Long and summer fat and some are Blobby like this you can imagine uh this tumbling object uh flare flare flare could easily be detected as fake sources in a source detection program so it's not a trivial thing to get rid of them even when they're sort of somewhat rare you can do some medium filtering when it gets to the level of more than one per wide field image then it starts to be a real problem so I'm going to just mention I'm only going to be talking about the effects on optical astronomy there is a big threat to radio astronomy which I'm just going to ignore uh those of you who remember uh Ashley zetterer from her time here is now at NSF uh uh working on these issues as well as Harvey list so you should get one of them here to tell you about that so in May 2019 SpaceX launched the first 60 starlink satellites on a single rocket and the next day Marco lambrook and the Netherlands took this image of 60 magnitude 2 objects crossing the sky and we kind of went oh that's brighter than we expected what happens when there are a hundred thousand of these like the companies are talking about that could be bad and so we started actually paying attention which we hadn't been before and so uh I organized a bunch of satellite hobbyists to do some early measurements of brightness once the satellites oriented themselves and raised their orbits to 550 kilometers instead of the deployment altitude near 300 they were significantly fainter and this histogram shows how bright they were this is only the satellites that reached that higher 550 kilometer orbit and they're they're mostly magnitude five or six although there is a bright tail to three or four uh even at that altitude and so they are largely naked eye objects in the summer of 2019 here and uh that this is from a paper I published in 2020 uh and when they come off the rocket they're they're this much brighter train of satellites and here's some pictures from around the world of these Starling satellite trains from different launches uh showing uh showing them in the sky uh so they've done some mitigations the satellites are a bit fainter but they're still some of the time naked eye and Dark Skies but more importantly that very first uh observation shows that it is technically possible if you are a super villain whose goal in life is to destroy ground-based astronomy and you're a billionaire you can now you do now have the technology to deploy a very bright constellation of satellites in low orbit that would outnumber the visible stars and and so and change the night sky for everyone and so you might think that even if you're you know in a space-faring not in the space very Nation but a nation that's still you know maybe cares about the night sky for for cultural reasons you might want to have a say in this rather than it just being uh one country's regulatory Authority that approves this happening uh and so there is nothing right now there's no Global regulation that would stop some other country licensing an extremely bright constellation of many thousands of satellites as we shall see um so I'm using the word constellation I thought I should emphasize that I'm not talking about volpecula and Ursa Major these are satellite constellations uh a whole bunch of satellites owned by the same organization carrying out the same mission in similar orbits and so what do you do to make a satellite constellation you put a bunch of satellites in an orbit let's look at this 3A 3B 3C orbit here spread them out along the orbit and then you do the same thing in another orbit with a longer tube of ascending node that's maybe 20 degrees to the east and again and again and again until you have a shell of orbit of satellites in orbits surrounding the earth at a particular height and a particular orbital inclination to the equator and that's one shell and then you can make multiple shells at different heights and different orbital inclinations to serve different customers constellations are not new we've had constellations since the Cold War but they typically back then were you had a couple of satellites or four or maybe eight in the case of this Soviet Australia 1M constellation where they had launched eight at once and they would operate and then be replaced by another eight but in those days the satellites didn't have propulsion and so they stayed in orbit and so even though this constellation is all dead all 360 satellites from it are still in orbit uh the more modern constellations before starlink were again there they sort of had up to 100 or so working satellites they tended to be left either half propulsion or be in lower orbits so that they wouldn't litter after they died the biggest one from a company called planet in San Francisco has 200 satellites Imaging the whole Earth every day uh and uh and they have about another 80 that are dead but but haven't re-entered yet so you can see there's a whole bunch already up there in this sort of of order end of order 10 to the two regime but starting with starlink we're now going into the regime of 10 to the 3 and 10 to the four it's a whole different story and so as of uh an hour ago when 49 more starlinks went into orbit so I I had two versions of the talk one one which had yesterday's numbers as much if this one actually went off which it did so we're now up to 3822 starlink satellites have been launched and 544 from the Rival British based company oneweb and there are these huge things with 10 meter long solar panels these are not tiny cubesats they are quarter ton satellites but this is just the beginning and if you look at the licensed applications to the U.S Federal Communications Commission the FCC and the international telecommunications Union the itu uh here are some of the constellations that seem to be more or less real there's other filings that I haven't included because I'm not convinced that they're they're really going to happen uh and uh many many big constellations including this ridiculous one from a company called espace Telecom Mogul Greg Weiler who's really based in the US but he put in a an application in a kind of flag of convenience way through Rwanda so Rwanda has informed the itu that they want 300 000 satellites in this constellation it is hard to imagine that this is serious uh or or how they would achieve this but it is a real application to the itu so I think we have to pay attention to it and these satellites the starlinks are currently making about five to seven the one webs are smaller and in higher orbits they're magnitude eight to nine and so those are two faces the naked eye but still nine is still pretty bright by uh you know astronomer standards so what do we know about these constellations when you apply to the FCC for example you give these tables these are four different constellations and they're defined by uh each constellation has some number of shells and each shell has a given altitude of inclination and so on so if we zoom in on the starlink generation 2 shell number three it's defined as being at 350 kilometers at 38 degrees inclination to the equator with 48 orbital planes and 110 satellites in each of those 48 orbital planes for a total of 5280 satellites in that shell which remember is one of like nine and their whole the whole starting Gen 2 is going to be thirty thousand so that's actually a Shin given that they're circular orbits to do a pretty good steady state simulation of what this constellation is going to be like shall do in a while I just want to show you one of these this is the some one webs this is a 36 one web satellites stacked on a dispenser on top of a Russian rocket stage so this is how what the launches uh look like they come off the stage they deploy their solar panels and here are from last October the 426 one web satellites in orbit and you can kind of make out the orbital planes and you also see off the coast of South Africa this uh stripe which is a bunch of uh of 36 that had just come off the rocket and are still close together and haven't raised to their operational orbits and so that's one of these trains like I showed you in the earlier photos but the story from last year was one of many new entrants in the game a whole bunch of companies from China from the US from Russia have started launching test satellites for their enormous constellations two or three but it's the harbinger of of things to come including this enormous satellite from the Texas company St space mobile uh which which has a huge communications array that in certain orientations is about mag2 so I want to step back for a minute and talk about uh the overall satellite population to put these Mega constellations in some context and show you how things are changing uh so I already showed you the uh the plot on the left which uh is the number of active satellites versus time trending up really fast in the past few years but if you add to that the total number of tracked objects in space including all the junk and debris there's about 25 000 objects that the space force are tracking down to about 10 centimeters in size and the active satellites are the colors I'm using for them are are pink and red and then cyan to pull out the Starlings and so you'll see that although there's a bunch of them they're not dominant and so you might think okay this is not transformational uh and indeed when you think about the fact that below 10 centimeters pencil beam surveys suggest that there could be a million objects down to one centimeter which are objects still big enough to really hurt you if they hit you at 17 000 miles an hour um the uh uh you know this is what's another few thousand right but I'm going to argue to you that that's the wrong way to think about it because if you filter this data on Mass and on orbital height and look at the ones that are big and therefore bright and low and therefore bright uh you were and skipping over to save time lots of really interesting stuff you can ask me about later um here are the ones that are big and bright more than 100 kilograms below 600 kilometers and what I'm showing here is the evolution of the population from 2005 to now where the cyan is just the starlink constellation and so you can see that just in the past couple years it has become completely dominant in this category it's completely dominating big objects in low orbit and that only ten percent of the full constellation is is up yet so uh uh so you know Elon said when when he started launching oh yeah you know there's thousands up there already it won't make that much difference but this shows no it's really changing the environment and it's changing the frequency of close misses and all of these things it's it's uh it really is a qualitative change so I'll summarize the slides I didn't show the upper part of low earth orbit above 600 kilometers is mostly debris it hasn't that population hasn't been changing that quickly the lower part is now dominated by payloads in 2005 it was dominated by debris now it's starlinks at the big end and cubesats at the small end and there's been transformational change in the past five years uh a couple other things that come out of my database commercial satellites which used to be like one-third of activity in space are now vastly dominant over government satellites and just in past couple years the Chinese commercial space sector and this is something that doesn't get talked about enough the Chinese commercial space sector has started to exist and become important and that's going to have implications when you have many different players from different companies in China who are new on the scene a lot of our governance methods in space sort of have been really haven't been incorporating the Chinese very effectively and we're going to need to we're not going to have a choice okay so now I want to move on to the actual uh simulation work that I've done and uh and so with those uh constellation uh uh definitions that I showed earlier you can instantiate the steady state of these constellations assuming they're all in their final orbits uh and this just shows you it gives you a sense of the latitude distribution this is for starlink one web and the proposed Amazon constellation and they all look pretty much the same which is that they have very dense coverage between about 52 North and 52 South and very little coverage in comparison over the polls where there aren't as many customers so if you live in London you will have many starlinks overhead if you live in Edinburgh you won't um unfortunately the weather Roe is not great for observing as I recall from my internship there yeah well that's also true um so you could look at this and ask the turn it around and ask the question if I'm in a particular latitude and I'm looking in the sky how many of these satellites will be overhead will be above the horizon uh and it'll depend on the orbital inclinations of each of the satellite shells and for each of these colored shells that I have here from a particular uh plant constellation uh they have this one is a polar one this one's an intermediate latitude and then the purple one is a low latitude one they have these double peaks near near the latitude of the inclination of the orbit a little inside so when you add them all up what you tend to get is uh the same as we saw in the previous plot basically which is not not so bad above 50 degrees but I'm pretty flat from 50 North to 50 South with with details and this particular one had 4 000 satellites above the Horizon pretty much all the time the good news of course is that a lot of these satellites are in the earth's shadow and are not reflecting sunlight at you so we have to calculate is how many are illuminated and this is pretty straightforward geometry you basically need to calculate three angles you need to ask is the satellite above the Horizon right now you need to ask is it night is the sun above the Horizon uh where the Observer is because if not it's daytime and the optical astronomers are asleep and uh and then finally is the satellite high enough that even though it's night where the Observer is it's daytime where the satellite is and it's illuminated and so if you have the JPL ephemers you know where the sun is and you can then use your orbit models to to simulate this for all the many satellites and I took a subset of the constellations that I think are likely to happen excluding this ridiculous 300 000 one and so that ended up uh uh mostly Western ones one Chinese one at 13 000 satellites total of 67 000 satellites that we may expect to be launched in the next five to ten years and uh so other people are actually trying to work on simulations that would include how bright are these satellites gonna be and I haven't done that for two reasons one is was that most of these constellations are basically going to be annoyingly bright anyway and they're going to be magnitude 5 to 10 and even down to magtan that's still a concern and the other reason I haven't done it is it's a really hard problem to predict so uh I've just said how many let's assume that they're going to be bad how many of them are there going to be and I want to emphasize so if you're in a higher orbit then each satellite can see more customers so each Observer can see more satellites but also the satellite is a it can see the sun longer Into the Night uh because it's nice and high up and more certainly if it's higher the angular velocity is slower and so the surface brightness on your CCD of the trail is higher at other things being equal and so High orbits are bad for the effect on astronomy low orbits are bad because the satellites are really bright and so they might be naked eye which is bad for the general public but also uh Tony Tyson at Reuben uh has shown that if they're brighter than mag-7 they can actually cause electronic ghosts in his CCD that don't just leave a streak but but affect the whole field uh and so that has led to a recommendation to the companies that they should really try and be fader than meg7 whatever else they do um because otherwise uh Reuben is going to be in trouble so so here's a reference simulation that I'm going to be coming back to and what this shows is the number of illuminated satellites as a function of time of night and this simulation is for kind of the worst case it's the summer solstice when the sun is not far below the horizon it's for 30 degrees south to model the uh observatories in Chile uh and it's only showing how many satellites are actually not above the Horizon but above 30 degrees elevation because we're professional astronomers and we like to look at the Deep Sky RMS less than two uh and so what you see is that in the uh and so as you go into the night these vertical lines show anglement below the Horizon for the Sun from sunset at zero to end of astronomical Twilight at -18 degrees and so what you see is that during astronomical Twilight the uh the number of illuminated satellites is pretty constant about 500 in this simulation this black line is the total number I'll go into these details uh and then it drops by midnight to only 100. uh although the drop is not that fast and so you have a few hours where it's intermediate um so you might ask is 100 100 satellites above uh elevation 30 bad or good and I'll argue in the next slide that it's bad but but let's just understand this plot a bit more um so these different colors are the different constellations I'm modeling and then the black line is the sum of them right and so for example this red line is the one web constellation in a high orbit and because it's in a high orbit it completely dominates the high means 1200 kilometers in this case uh at midnight it's dominant starlink has these magenta lines generation one generation two where they're as bad as one web in the Twilight hours but they drop to pretty much zero at midnight because they're too low to be illuminated uh that late Into the Night um and so the the altitude of the constellations is the key thing in what effect they'll have in the middle of the night now uh just for illustration's sake I uh I had these two horizontal lines here and here showing the number of magnitude four stars above 30 degrees and the number of magnitude 5 stars above 30 degrees and so you see that in the Twilight hours there will be more illuminated satellites than mag 5 stars but that's remember that a bunch of these satellites are actually going to be fainted than mag-5 so it's not we're not in the regime in this simulation where the sky is dominated by satellites rather than Stars so what is a adelites in the sky in the upper part of the sky mean it corresponds to about 0.04 satellites per square degree and they're going at an angular velocity most of them have about 0.35 degrees a second so it's quite fast so it doesn't take long to cross your uh field of view in the telescope and so you can say how many streaks do I expect on a typical wide field exposure if I expose 40 minutes uh in a field of view that is D degrees wide then I get of order 1 times T times D streaks uh in my data so a hundred is bad because because expecting one streak in every one minute one degree image if you're doing wide field studies it's not great uh and it doesn't take much tweaking of the parameters to make this much greater than one uh so if you're doing short exposures in a narrow field of view you're fine if you're doing long exposures in a wide field of view you're going to have some issues so uh what if you what if you're not observing an air mass less than two what if you want to do things all the way to the horizon uh then things are much much worse the geometry says that most of the visible illuminated satellites are much closer to the horizon and so instead of 500 in Twilight you're seeing four thousand and over a thousand even at midnight uh and uh and so and that is now much more in Twilight than the number of Mag six stars above the horizon uh so uh so the the the uh it's gonna look like you're near Logan Airport when you're when you're looking toward the Horizon uh uh once these constellations are deployed so to recap big constellations at low altitudes meaning 500 kilometers or lower have the potential to be naked eye maybe design can mitigate that uh you can have many satellites visible in the Twilight hours which is what Ordinary People call night right uh astronomical Twilight is what I'm talking about the first few hours of the night so it's going to affect the experience of the night sky for people around the world because they mostly look at the sky you know in the early hours of the evening before full astronomical Twilight uh the bigger constellations uh the Bitcoin station said higher altitudes are probably going to be too faint to be naked eye but even in uh at midnight in mid-summer they're going to be illuminated in very large numbers and so wide field ground-based Imaging is going to see a lot of streaks and other people have done very similar simulations these ones are an update the ones I've shown you are an update of a paper I published in 2020 uh Olivia Hano Pat scissor case Bassa and Samantha Lawler have done similar studies that basically give pretty much the same answer I'll just mention that if e-space is serious and really launched 300 000 satellites then it just gets horribly worse um with you know 20 000 illuminated satellites in Twilight above the Horizon uh uh and uh really you know just then we really will be in the case of many more satellites visible than Stars so I want to go back to this first simulation and ask okay but I did give you the worst case right I gave you mid-summer uh and so how do these predictions vary with the time of year and with the latitude of the Observer and so on the right I have a simulation that's one month after the summer solstice it is quite a bit better at midnight instead of a hundred Stars illuminated 100 satellites illuminated rather you get only 60 satellites illuminated at midnight but the slope of this dip right isn't isn't so sleep so so you're still gonna have a significant fraction of the night when uh there are plenty of illuminated satellites making streaks in your data uh if you go all the way to the winter solstice things are much better the sun is too far below the Horizon to illuminate the satellites you lose you know an hour of your admittedly shorter night uh at either at each end of the night and versus latitude I show here on the top left at the equator it's sort of like the winter is it's not a big problem at the equator I already showed you the 30 North the top right if you go further north to 46 so I showed you 30 South but 30 North is the same right um uh 46 North it's worse because of course the sun is not so far below the Horizon so you don't get the dip in the middle of the night so if you're living at temperate latitudes like Boston you're you'll you'll see a lot of these satellites all through all through the night in the mid-summer uh but then if you go up to the Arctic where you'd think things would be even worse because the maximum inclination of most of the satellites is capped at 50 something degrees you run out of satellites and so it's actually better it's going to be better at the South Pole if you have an observatory uh there there won't be so many satellites uh visible good luck with the weather though um what if you go into orbit well one web is at 1200 kilometers that's above most of our Leo telescopes an HST is currently it's it it used to be a 600 but now it's only at 540 thanks to atmospheric Decay and that's 10 kilometers below the main Starling constellation so the starlings are whizzing just above it as is this Chinese rocket stage which passed 35 kilometers in front of HST uh uh a couple years ago right through the center of the field of view blocking this cluster of galaxies it was looking at so that's what that's what a observing a satellite 35 kilometers for you from you with HST looks like um and uh so Sandra Crook and colleagues have looked at the archive and showed that even before starlink uh there was a serious impact of satellite Trails on HST observations at the several percent level eight percent of composite images already affected and then they model that it's going to get a lot worse but it's still not going to be as severe as for the ground-based observatories because um yeah HST does long exposures right but it has a very narrow field of view uh and so it won't catch as many satellites and also its pixels are very small and so that cuts the surface brightness but I want your wide field telescope to low earth orbit as the Chinese are planning to do at the end of this year I think you're going to run into some real problems so that's the simulations but I made some assumptions in those simulations that turn out not to be entirely true and so I want to talk to you about how we look at what the the um they're based on what the companies said they would do but what are they actually doing so we can keep track of them a couple ways from space force tracking data with Radars which is public and which is passive so it doesn't require the satellite to be alive and you get about one orbital data set a day the downside is you don't get them for the first two weeks of the satellites Mission because when they launch 60 satellites it takes a while for the space force to uh to to disambiguate them uh so some of the companies notably SpaceX provide higher accuracy orbital data from their onboard GPS on the satellites and make that public and TS Kelso at the Celeste track website provides standard orbital fits to these and so you can use those and they're available the day after launch so I should have ones for today's batch tomorrow I hope and that's great but they're only available if the satellite is still working of course and in fact that's one of the ways you twig that the satellite has stopped working uh if that data stop stops flowing and so this allows us to start study the orbital height versus time the orbital plane versus time all kinds of deployment strategies that the companies never talk about so you can see a lot from this open source analysis and I I'm summarizing a lot of what we find in uh in this sketch here of the normal life cycle of a constellation satellite because I think it's it and bear with me because this has implications um so this is based on starlink but the others are similar you deploy it around 250 kilometers so the x-axis is time uh in varying units uh you raise the orbit toward the operational orbit but you pause for a month or two at 350 kilometers before getting up to 550. and then you operate and you operate ideally the the nominal lifetime of these things is five years but at least 10 percent of them operate for much shorter times and then are retired because they develop problems of some kind so you lower the orbit again you pause for reasons that aren't entirely clear sometimes for many months and then you lower the orbit again until you get down to again about 300 or 250 kilometers and then you switch the satellite off while you do that because you have these little feeble Krypton powered ion engines and they can't struggle against the denser at lower atmosphere and so you can't keep the satellite pointed well enough to fire your engine so you just switch it off and let atmospheric drag bring the satellite in for the last couple of weeks and it re-enters at some random position uh over the Earth um and so and you see you can sort of see where this happens because the height versus time is very linear when it's under control and then it gets curved when it gets into the uncontrolled decay so here's real data here are 60 satellites launched in February 2020. and so you see that they go up they stay at this operational orbit some of them are still there a bunch of them come down retired early and note in particular here the the satellite is being lowered in mid-2021 but actually stays a full year in this intermediate orbit so I guess they were debugging it hope to save it or wanted to study it some more and then finally let it re-enter in late 2022 and so the point here is that this is not uncommon some of them come straight down but even you know this one waited a while this one waited a while and so that adds to the amount of time that these things are in orbit but not in their nominal Heights now this orbit raising thing you see the three stripes here of orbit raising is also kind of fun uh and so if we zoom in on this here are the first few months and launch in February the satellite split into the 60 satellites split into three groups of 20. 20 go straight up to 550 kilometers 20 weight for a month and then raise and the final 20 wait for several months and then raise so why are they doing this this means that the orbit raising takes over four months before their all the satellites are operational and so this is because of differential plane procession for those of you who are not uh orbital mechanics fans um what I'm showing here in the blue are these starlink orbits at 53 degrees projected onto the equatorial plane so they look like ellipses when projected and these so here's the the line of nodes and because of the oblateness of the Earth the first order procession effect means that the orbital plane rotates around the equator at a rate of about 4.5 degrees per day it's really quite significant uh whereas if you're only at 350 kilometers the red orbits you your orbital plane actually rotates faster rotates five degrees a day so half a degree a day faster so for every month that you dilly dally in the intermediate orbit you are changing your orbit plane relative to the operational system which is rigidly rotating at five degrees a day by five degrees and for every month you're changing by 15 degrees and so this way by splitting into these three groups with one launch you fill three different orbital planes of your shell for you know no extra expense so very clever use of differential orbital procession I think but the consequence is that these satellites spend a significant fraction of their lifetimes not at the orbit height I was using in the simulations so I I maintain this page giving all the statistics the bottom line is that as of this afternoon 3822 satellites launched 35 30 of them are still in orbit 300 of them have actually re-entered already a few percent fail in the higher orbits and are re-entering I meant to call out uh in this plot the red lines are the failures in high orbit where the curvature is is due to atmospheric drag so they look very different versus time than the actively controlled ones and so there's a few that are just gonna re-enter in four or five years um and so currently only 3 100 are at the 550 kilometer licensed orbit in the application about at any one time oh yeah is that about three percent failed early so they come off the rocket something's not right either they die immediately or the commission control goes this is not gonna you know be this is in potential problem and so we're just going to let it re-enter and we're not going to raise its orbit and so atmospheric drag will bring it down in a very short time because they deploy into these low orbits and so so you know a total of like uh uh four or five percent fewer now than in the early days they really have improved the reliability um our failures but then also a bunch are just spending a lot of time in orbit raising or in retirement uh and so at any one time 10 to 15 of the system is that these lower orbits which are then brighter so bad for us and passing other systems altitudes which is bad for traffic so you also want to monitor how bright they are I haven't been doing that work but a bunch of people have notably Jeremy uh triglo and Reed in Chile Aaron boli in Canada uh and uh and so this extensive work they all basically again get the same answer which is that the starlinks are about magnitude five to six uh the one Webster magazine seven to eight the various mitigations that SpaceX has done to make their starlings less bright gained like a magnitude and then they changed the design again and it lost about half a magnitude and so it's it's it's uh it's kind of been a wash uh at the one magnitude level but SpaceX are actually doing I'm here to tell you a significant amount of work to help us um so let me first uh review the problem this is what a starlink satellite looks like it mostly is in what they call the shark fin consideration it's it's uh it's like a flat panel TV about three meters long with a 10 meter solar scale sticking up out of it uh that solar cell can rotate to be in plane with the flat panel to reduce drag for when it's doing Maneuvers but when it's operational it's vertical like that and so it's as it orbits towards the Terminator with the solar array radially outward from the Earth the sunlight will scatter off the solar array down to observance near the Terminator and we'll scatter off the bottom of the spacecraft bus to observe us a little further from The Terminator uh so what are you going to do for that there's a whole bunch of mitigations they've tried they've changed the attitude control during orbit raising they tried painting the antennas but that made them too hot they tried visors shielding the antennas from Reflections uh that caused extra drag and got in the way of their new fancy new laser cross links so they've abandoned the lows what they have done is uh change the backing of the solar cell material to the darker color and that has made the solar arrays less reflective at the cost of making the solar rays less efficient and they've accepted that hit but now they're about to deploy the generation 2 satellites they are enormous they are five times the mass a ton and a half instead of 300 kilograms five times bigger bus cross section A six meter bus 20 meter solar array just huge but they've added on various aspects of it the the this fancy new dielectric brag Mirror film that they've developed don't understand the physics personally but uh the claim is it reduces the brightness as I'll show you uh more like dark paint in various places and more attitude control strategies as I'll explain and the claim is that despite these anomalously bigger satellites they will actually be fainter than the first generation we shall see monitoring is important so this is uh satellite Engineers love the the brdf the by directional reflectance distribution function which is basically like with your asteroid Observer the the phase function uh the the bi-directional is angle of the sunlight hitting the satellite angle of the Observer looking at the satellite so this is a one-dimensional cut through this two-dimensional function uh and the purple line is how bright the satellites are now and the blue line on this log scale is how bright the generation two satellites are claimed to be uh and it's two orders of magnitude fader at many angles so that'd be really nice although I will point out that some you know the range over the full range of angles uh is seven orders of magnitude and and so uh they can be quite shiny still at the wrong angle but if this pans out most of the time these satellites should be quite faint we'll see another thing they're going to do if you're orbiting you're trying to get sunlight you have your solar array looking at the sun so one thing you can and but that reflects light back down toward the Earth so one thing you can do is lean back and that will reflect your light out into space at the cost of reducing your energy Generation by cos Theta uh and they say that they're gonna do that in this lower diagram and accept the fact they're going to have to have a bigger solar array for the same output uh they think that that will be uh net a win for us and so they really are you know uh SpaceX are putting in a lot of time money real money real time real effort into making things less bad for astronomy that's nice but there are plenty of other companies as I've shown you in many countries we can't rely that all of them will try and help us and so we have to work on the broader policy effort so we're organizing to protect the night sky uh we had a working group at the in the double A S light pollution committee uh we uh had Noir lab and double as organized the second one and second two conferences uh we organized with the IU in the UN the dark and quiet Skies conferences we presented a report to the UN we had Splinter sessions in Seattle just this month and last year we'd had a lot of fun briefing all the various Washington agencies uh uh the the White House the state department the FCC the dod and others uh to show them the results of our SAT Contour analysis and so the folks in Washington and at the UN are aware of astronomers concerns now uh and they have taken them on board to a certain extent of course they're also aware that these companies want to make money and so who's going to win is is not clear uh so the iau established its center for the protection of dark and quiet Skies last year cps.iou.org uh it's still getting organized but that will be eventually the online Center for all the relevant info we're going to continue working with the various satellite companies on mitigation I had a good chat with a SpaceX guy at the double as um we're going to continue however public pressure and building coalitions outside of astronomy there's a lot of people care about the night sky and we're going to carry on working with the UN on the idea that maybe the night sky is part of Humanity's Heritage uh and the environment so I'm going to end this policy section before I wrap up by just leaving with some some questions to think about is the night sky part of the environment and if so does that mean it's subject to environmental regulations uh the outer space treaty says that countries should give due regard to the space activities of other nations is astronomy a space activity it hasn't historically been thought of as one by the space Community but maybe it is in which case maybe the due regard cause applies can we maybe get the UN to say that the night sky is a good thing we should protect it can we get the US which already lies issues licenses to companies only if they do a debris assessment to say how much space debris they're going to make can we get them to do a brightness assessment as well and maybe even regulate along the lines of you can't have more than x satellites brighter than y right so so those are some of the potential policy directions and by the way I am your local representative on the Wes committee on light pollution radio interference and space debris which is desperately looking for a better acronym a name but for now that's what it's called so feel free to reach out to me um and I'll leave you with this conclusion the mega constellations are going to be a significant change to the lowest orbit environment and to the night sky but the impact on astronomy depends sensitively on the constellation architecture particularly the altitude of the satellites and so again the low orbit satellites potentially will be naked eye objects but maybe you can have used these clever uh strategies to make them less bright uh they're illuminated near the horizon and not a not at midnight so they might be a threat to some astronomy uh if you're looking at wide field toward the Horizon but most astronomy is not going to be so badly affected by them it's the higher constellation ones but although they're not a threat to the naked eye skywatcher are a threat to professional astronomy because they're illuminated all night long in the summer and yeah it's fine in the winter but for the theorists in the audience remember some objects are some are objects that you can only see in the summer and and so I think the uh the LMC might be one and and so uh so it does matter uh so I have a web page with lots of references and resources here uh and feel free to reach out to me if you have further questions thank you very much thank you Jonathan in case you're wondering why Planet 4589 that is Jonathan's asteroid it's 4589 McDowell yeah right anyway uh questions yeah yeah you are well first of all to confirm from my experience your HST statement in 20 orbits that we got from in the last couple of months we got one image affected by satellite okay which was surprising to us but anyway surprised that it was so few yeah yeah yeah yeah my question is that you completely missed in your talk with geostationary said I did essentially affecting a lot because the Deep surveys are concentrated around the equator and for instance from Arizona you cannot observe really the I think it's UDS so there's a field at minus four degrees where all those things project and once we had a mask spectroscopic mask with a vertically allocated slits and all of them were just crashed by by the stuff because you know you have magnitude 10 magnitude 11 and it just crosses the one Arch second wide slit and it's enough to kill the origin seven Galaxy that's very interesting yeah yeah I mean I will say the thing about the geostationary satellites is that situation hasn't been changing rapidly right uh that's been the case for a long time uh and where say all of the change we're seeing in the satellite population has been in the lower stuff but is it you're absolutely right it's important to worry about that too other questions so will of the locations of these objects be known with any Precision so they want to actually potentially schedule around them right that that's the challenge so so for the ones in operational orbit um using the SpaceX ephemeris it's probably accurate enough to get it within at least an arc minute right so how how accurately do you need right uh you don't probably need to know where it is in your field but you need to know that it is going to be in your field so I think we can as long as we get predictions of Maneuvers and things like that right uh and so so we have been working that's one of the main discussions we've been having with the companies is please please give us this kind of ephemeris data if we just depend on the two line elements from the space force they're probably not good enough to do a really good job of of that avoidance stuff oh yes you have an appetite for people yelling at us on the internet um I think it's fair to say that is fair a lot of very opinionated people talk about this issue and what you know and one of the common things I often get is people are like why don't you just filter out like the time when the satellite is tracking like in time or something like that so what's your response when people surely ask you right right yeah and so that's yeah the one I get more more often is why don't you just move all ground-based astronomy into space because that'll be easy I think everybody knows but yeah we don't have to go we won't have to explain to this audience why that's not really feasible right now um but but yeah but why don't you uh just have a shutter that cuts out the data for that one second that the satellite goes over that'd be great uh uh if you can do it in software uh uh I don't think Igor can correct me that most uh imagers have that capability right now you can close for for optical Optical imager easily the shutter but you have to time it like there is capability but there's no manpower to implement it that's that that's the thing yeah I mean you could do that it's an unfunded mandate right so it's good it's not going to be cheap to do that on all telescopes it's not going to be uh and and it's not clear that these orbital data are accurate enough with the timing right if you're a second off you know it doesn't work right and so I think we will be forced to that but it's it's it's not trivial to implement oh that's another question yeah what what if any requirements are there to bring these down when they're done right so the license requires you to explain your your get rid of this strategy and historically it's been like yeah even if you go bust the company that inherits your debts you know inherits the responsibility to get these things out of space um uh and so SpaceX have said that their starlinks are designed to be what they call fully demisable they burn up completely on re-entry uh and they're in these low orbits and so they'll come down within a solar cycle or so um one way about this higher altitude will be up for centuries if they fail so they have to be much more reliable uh but they have the idea still that they will have enough propellant to lower their orbits unless they fail yes oh hi John this is maybe more wishful thinking but you know the Iridium project in the 90s failed particular lack of Interest really finances so just looking at the numbers satellites being launched but surely not not every one of them is going to have business enough to keep them operational do we have any indication yet as to whether the SpaceX analyze for instance or one of the other constellations is actually getting the response that they're expecting right are they actually going to make a profit yeah well you know I mean profitability is so lost Century I I think uh iridium went bankrupt they got bought they're still in business they launched a new generation constellation uh so they're still going despite having you know not made money for a while uh and and so yeah I I think it's too early to tell uh I think it was slower than projections uh the uptake of Starling and they've had problems rolling out the service uh the Practical parts of the service like getting uh antennas to the users right I think the success in Ukraine of uh getting Communications to people in Ukraine has sort of changed the dynamic of that a bit and they they're probably going to have a certain anchor customer uh in uh near DC that will will uh ensure that they'll make a profit that's my guess um but not all of these companies will survive that's for sure some of these systems will go bust and will they successfully retire iridium said when they were first launched oh yeah even if we go back we'll we'll we'll deal with all our satellites and they had like 80 and 30 of them just straight out failed and couldn't be deorbited if they're still none of them are still up there so uh SpaceX are doing better how these other companies will do in terms of reliability remains to be seen one more question how does this relate in height to where the space station is or where are other scientific satellites are I mean great question yeah do we have to drive out of the way well SpaceX says trust us we'll Dodge you so on the way up they go through the orbits of the Chinese and international space stations and the China complained because a couple of the starlings actually move one on the way up one on the way down moved more close than they were comfortable to the to their to their uh station and the space they said well we were fine you know we didn't go that close to you uh and uh and the problem is this that the US then went oh you see it's SpaceX are right China's wrong there was no threat our models show that it was an easy Miss but these are not regular satellites that just orbit ballistically right they're running their electric propulsion systems all the time and so that means your ability to predict where they're going to be tomorrow is substantially poorer than if you know that they're not actually firing their engines yeah yeah when you take that into account I think China were right to complain because again if I'm a national on the on the Chinese space station and SpaceX tell me yeah we'll miss you by by you know 100 meters it's not a problem I'm maybe not entirely comfortable with that one last question thank you very much great talk very nice question if two different companies want to launch satellites on the same Pockets how is it deciding who can do what that's a great question so if they're from the same country that's the responsibility of the licensing uh of the FCC for example if they're from different countries there's nothing to stop that right now geostationary satellites of slots that are allocated by the international telecommunications Union more for stopping radio interference than for stopping collisions but low orbit satellites are not regulated in that way and so there's absolutely nothing to stop China going yeah we're going to use the same orbit as SpaceX and we'll just dodge best we can um so and the problem is you know they they all the companies all go oh yo yo our automated avoidance systems are really good and there's not a problem I'm just not confident about the systematic Tales of their error distribution you know so I feel we may we may learn some hard lessons that way okay but I think we have to wrap it up here if you would like to join us for dinner just come and talk to me uh after the talk so let's thank Jonathan again thank you foreign

Channel: CfA Colloquium

Views: 264

Keywords:

Id: 0xXbyx2pIPQ

Length: 62min 40sec (3760 seconds)

Published: Wed May 24 2023