Lucy in the Sky
with Debris

Speakers
Meredith Rawls (MR), Isabella Ong (IO), Seet Yun Teng (YT)

YT

Could you tell us more about Trailblazer and how Trailblazer started?

MR

Trailblazer is a small part of larger work I've been doing in the greater domain that is concerned with “satellite constellations affecting astronomy”. This particular project came out of a realisation that even astronomers weren't aware of this problem and it was sneaking up on us. It was coming out of nowhere because, prior to 2019, there weren't any constellation satellites. SpaceX launched their first set of Starlink satellites in 2019, and everybody was surprised at how bright they looked—even SpaceX were like, Oh, we didn't think it was going to be this bright. Ever since then, astronomers have been slowly noticing that our data has this new contamination source. It hasn't affected our ability to do science yet. But as the population grows, the space environment is literally changing out from under us, or above us. It's really hard to know what the sneaky long term effects are going to be on our ability to do science. So I care about it as a scientist who wants to use data taken from the ground to understand the universe. But I also care about it just as a person who lives on the planet, who would like to be able to look up and not have my children just think it's normal to see swarms of satellites instead of stars.

There's been a few different fronts where astronomers have been starting to address this. I'm part of the Centre for the Protection of the Dark and Quiet Sky from Satellite Constellation Interference (CPS), a centre that the International Astronomical Union (IAU) created for dealing with this issue . I'm leading one of the sections that centres on observations and data analysis. I had already started Trailblazer as a side project with a colleague of mine who's finishing up his doctoral degree right now. It was just two of us trying to cobble together a website where we could say, Hey fellow astronomers, if you happen to get an image that has a satellite (which is going to be an increasingly common occurrence), and you would like to share it publicly to both enable us to study what happens to our images when they have this kind of contamination and also communicate the problem a little more clearly to the public in general, here's a website where you could do that.

The goal was to eventually enable thousands of images from all across the world and from different telescopes, so that we could do a study and try to understand how the satellite population is growing and changing with time and how it is affecting different science cases. This hasn't really happened yet, partially because the project was put on hold a year or so ago because I went on maternity leave and Dino was trying to finish his PhD. So it's just been sitting stagnant for a couple years now, which is why some of the data is not all that recent. But we do have a good set of test images up there, and I'm hoping that now that we have a new software engineer who was just hired to work on different aspects with the IAU CPS, she will have some time to revisit Trailblazer, improve the website and connect it with a general observation upload service that she's building.

Sometimes what's happening is that satellite companies themselves are actually requesting that astronomers observe their satellites and measure how bright they are. They don't all want to do the right thing, but some of them want to actually try to make the satellites dimmer and see if they are meeting their goals. It’s also a new thing for satellite companies, thinking about how to design a dark satellite—you can't just paint the whole thing black because then it gets too hot. I'm not a material scientist, so I don't know all of the ins and outs of this, but it's a tricky problem. Because everything is reflective to some degree, and of course, the sun is present, how do you minimise the brightness that you see from the earth if you're going to be launching satellites? So they specifically want astronomers, as well as amateur enthusiast astronomers (backyard telescopes), or just even visual observers who record the stuff they see in the sky as a hobby (which is already a whole community in the United States).

We want a way to have these people put scientifically useful information in a database that we can then analyse in some consistent way, in order to write papers and report back with meaningful conclusions on how bright satellites really are, what kind of impact it is having on science, and what we want companies to do differently. So that's what we're focusing on now. We hope to then bring Trailblazer back into the fold, so that once we have a way to get the boring numbers version of the data, we can also have a nice visual presentation as a supplement, or maybe even as its own entity. So that's how Trailblazer has morphed over time, and why there's not a whole bunch of new data coming in at the moment. This is still a big problem showing up in a lot of people's images. We just haven't had the bandwidth to keep it top of mind recently.

I will say it's interesting that you have space debris mixed in with satellite brightness, which I think is absolutely correct because it's an extension of the same situation. But in the astronomy community, people tend to really think about it as two separate issues. They tend to think about debris being more of a consideration for security, defence, or military, to do with  debris tracking and collisions, and then they think about satellite constellations, which are bigger, rapidly increasing, and reflecting more sunlight, as a separate issue of concern. I focus a lot less on debris and a lot more on commercial constellations, but I think that they're very connected. If something crashes into something else, you suddenly get a bunch more debris, and at some point, we will hit a threshold—you can't put infinite stuff up there forever. But at the moment, we're just running a wild experiment with very little regulation. So I'm really glad that you guys are working on this and trying to present the larger issue in a new light.

IO

I know we hop between satellites and space debris, but ultimately, we see all of this as contamination and interference. Even though we understand them as two distinct problems, for us, the choice of “space debris” is very deliberate. We tend to not use the word “junk”. There’s a very fine line between a working satellite, which is not junk, and when it ceases to be functional; even though the form stays exactly the same, it becomes junk and then becomes a collision risk. So the reason why we use the word “debris” is because it describes a product of violence or destruction, or something that has been through a fragmentation or destructive event.

We feel that space junk is a passive term which indicates something that's unwanted up in space, but debris describes more of the violent nature of the object, because things are moving so fast and orbital space is not as passive and inert as what most people think. I think maybe because we come in from a visual art perspective, we can start to think about the nuances between the words we choose and what it means. So even though satellites and space debris are two problems, we feel that, ultimately, they affect how we appreciate the night sky.

MR

Yes, it’s ultimately human-generated objects in space. I like the subtle, violent connotation of debris, I think it's appropriate here. Everyone has a relationship with the night sky to some degree. Different cultures have different stories and important traditions. I recently got the chance to join a Zoom meeting with Polynesian wayfinders based in Hawaii. They were talking about how they've noticed this change, because they literally use the stars to navigate their boats. Of course, they were able to look up and figure out that they were artificial satellites, but they were just like, wow, nobody told us that they were going to do this. I think that's been an issue in Canada as well. I believe there's a law that requires First Nations people to be consulted over any change in their environment. Of course, that didn't happen, because a satellite company in the US or wherever that’s launching all these satellites didn't, you know, send a letter to all the First Nations groups in Canada seeking consent to change the sky. That's not a thing. Maybe it should be a thing, but it certainly isn't.

It's interesting, because it affects the whole planet in a way that very few other issues do. Maybe the oceans to a degree. As astronomers, we say, let's build a telescope as far away from the city and high up on a mountain, we'll put it in the most isolated and driest location possible, and there won't be too much contamination from human-made light pollution. But all of a sudden, that doesn't work. Especially for radio astronomy. I don't know if you've had a chance to talk with any radio astronomers?

IO

Yes, I made a field trip to visit the Murchison Widefield Array (MWA), a radio telescope in Western Australia. It’s in the middle of the desert and very hard to get to. MWA has been operating in that spot for the past 10 years, and they just published a paper about how they have started to get Starlink interference with their telescope readings. They spent the last 10 years building this up, and it's also going to be the infrastructure that the Square Kilometre Array will add onto as well. It's scary to think that since Starlink—like you said, it was launched in 2019—there’s been a lot of increased interference that you would not have predicted four or five years ago, and it's a lot of money, effort and resources that’s been put into building this telescope.

MR

Especially in radio astronomy, sometimes you don't get a chance to look at your data in detail for several years because you're busy amassing this full data set. You need a ten year baseline, or you just are busy, and suddenly you're finally looking at data from 2021, and then you're like, Oh no, what's all this interference in it that I wasn't expecting to see? So I get the feeling that as time goes on, this time lag is going to continue, but it's going to be even worse by the time people realise. That's going to be interesting too, as people are noticing it more and more at different amounts of delays.

IO

When I was in Western Australia as an Artist-in-Residence at Curtin Institute for Radio Astronomy, a fast radio burst (FRB) researcher showed me a radio plot. You could see the FRB signature in, but half the plot was just white lines. It was probably a passing plane or satellite, he wasn't sure, but half the graph was unreadable and unusable. I think maybe it features more in radio astronomy readings, but how bad is it for optical and visual astronomy?

MR

It's not as bad yet. Partly because that’s the way that radio works: the wavelength is so long that you get all these weird side lobes and interference patterns, and you can't ever just point it at one tiny spot in the sky. You're always getting a little bit of bonus signal from nearby. But with an optical image, you can take a picture of just a little piece of sky. So that's helpful, because in principle, you could avoid some of the brightest satellites. You could just take another picture layer when you realise your first picture had a streak in it and manage to get one that has no streaks, to a degree.

So one reason I got into all of this in the first place is because I work for the Vera Rubin Observatory, which is a big telescope that’s under construction in Chile, and it's supposed to turn on in around a year’s time. Currently, there’s all this activity to get ready, working with pretend data sets, running software pipelines and trying to prepare for every contingency. And this (satellite interference) is a big unknown. We don't have a lot of extra resources lying around to understand it thoroughly. People will ask me, Can't you just tell us what the impact on science is going to be? I'm like, that's three dissertations worth of work to fully study and outline all the different possibilities. For one thing, we don't know how the satellite population is going to change in the next decade. We just don't know—it's going to go up for sure, unless every satellite company magically goes bankrupt. But how is it going to increase? Are constellations launched by companies in other countries going to care about brightness? Are they going to listen to what astronomers are asking? Because right now nothing forces them to. Are they going to launch just 5000 satellites, or are they going to launch 50,000 satellites? I don't know. And so I have to try to communicate this uncertainty to scientists, who are used to uncertainty, thankfully. But they don't like it. They want to know, how much more money do I need to still do my science? How many more images do I need to take? Can I just avoid one region of sky once and then everything else will be fine? I'm like, I'm sorry, it's not that simple.

So we've come up with several different ways to handle it. But it's interesting, because this has become a large part of what I do now for Rubin Observatory, even though originally I was just working on the data pipelines in general. I do that a little bit, still, but more and more, I'm observing areas which will be affected by the fact that satellites are everywhere. Because if you have an image and your algorithm is assuming that everything in the image is a star or a galaxy, it's going to make some wrong choices when there's three streaks in it, or a glinting object that is leaving a trail of bright blobs across it. It's going to behave differently— maybe it'll assume, Oh, there's five bright stars in the line. But if we have another algorithm that's like—Hey, I just found five bright stars in a line, maybe that's not real—then that will help us not mess up or draw false conclusions.

I spend a lot of time trying to figure out where in our software pipeline it makes sense to anticipate these kinds of problems. The impact is going to increase, but I think it's going to be sneaky. I don't think there's ever going to be one moment at which we're like, “Oh, can't do science from the ground anymore, forget it, close our telescopes, we're done.” That's not going to be a thing that will happen, but it's going to get subtly harder and harder. It’s going to be really tricky to justify when you're trying to write a grant proposal to get money to hire someone to work on the problem, you're going to be like, well, we need two years of work from this person, because half of their time is going to be writing additional software to deal with the satellites that are in the data before they can measure their variable star property or whatever it is they really want to study. So it's just tricky and has a lot of unknowns. I do all the work that I can to try to make it less worse for everyone. But you know, I can't do it all.

IO

I think it's strange, because you previously mentioned that astronomers are not paying that much attention yet, and with radio astronomers, it feels like it is becoming a problem where it’s just glitched or bad data—it’s like 50% of the plot is unusable.

MR

Maybe radio astronomers have been aware of this for a little longer and are a little more panicked about it. I’m not a radio astronomer. I have colleagues who are, and you're probably right, they probably have been freaking out for longer. I have a colleague that goes to every international panel for spectrum allocation negotiation and raises this issue, he's been doing this for decades. So you know, ever since radio frequency interference became a thing, he's been making noise about it and so have others, but now it’s just way worse than it used to be because of the satellites and the debris. Well, maybe not so much the debris for radio, because at least the debris doesn't actively transmit. But it can reflect.

IO

I would imagine people would start paying more attention if it's actually affecting your work, if 50% of your time is spent correcting and scrubbing out interference. But I guess it's not across the board.

MR

Well, astronomy is a small field. People don't care about astronomers. The funding that we get for our work is very small compared to all the other sciences, and even that is really small compared to other government budget allocations. So we already feel like a small player in a bigger game to begin with. Even if every single astronomer was mad as heck about satellites, I don't know how much we could really change about it. I'd like to think that we had some power to change it. But you know, the companies want to launch as many things as they can into space before there are any regulations prohibiting that. They want to win this space race, they want to get their hardware up first, so that people buy services from them, and they make all the money. So it's just a weird set of incentives.

YT

Yeah. I think, based on our conversations, it seems like you’d really need a commercial incentive for people to act. So one thing would be risk: if it's really starting to pose a big risk to their constellations, they need to start being careful about decommissioning.

MR

That is one nice thing—that we have a shared incentive there. They also don't want the debris to be unsustainable, because they need to be able to actually operate their satellites in space. So at least that is aligned, even if nothing else is.

YT

I wanted to ask a question about your work chairing the SATCON2 Observation Working Group that gives recommendations and strategies for both satellite operators and observatories. Could you tell us more about it, how it works, and what problems you think challenge the implementation of these strategies?

MR

That working group report that you mentioned was a first stab at indicating we need a coordinated approach to even begin to address this, and pose some basic starting points and solutions. Out of that, the SatHub at the CPS formed, and now we have an actual organisation to undertake this work. Of course, we don't have enough people or funding, but that's always the case. There's an updated draft set of recommendations that we've been working on, on the CPS website. It is a collaborative effort by astronomers and policymakers at trying to write down what we would like done differently or what kind of things we'd like people to measure and implement to try to minimise the problem. From that draft, we recently published a 30 page position paper that proposes a range of measures and mitigation strategies for different stakeholders, including the astronomy community itself, but also industry and regulators.

Oh, another recent news item on the CPS website is the unintended radio emissions from Starlink satellites. The main conclusion from that paper was that they found that even though Starlink is authorised to transmit at specific wavelengths, they are actually also transmitting in other wavelengths. That's why when they're finally looking at some of their radio data, they're like, Oh no! On one hand, that's just how physics works, it's almost impossible to not have some unintended emission. But I don't think anybody was expecting it to be in these particular bands which are often used for radio astronomy, although not technically officially protected, because we can't control where the universe has radio emission. We'd like to be able to see all of it.

IO

How proactive are operators and companies in rectifying or correcting some of this interference? I know Starlink have been looking into making their outer shield a little less reflective, and in the past starwatchers used to watch out for Iridium flares, but now it's a thing of the past because the new generation of Iridium satellites are less reflective and they don’t give off that signature Iridium flare. How proactive are operators and companies and correcting some of this?

MR

That is a great question and it varies. So far, I think we've talked with SpaceX, Amazon Kuiper, and OneWeb—those are the three big ones that actually have hardware on orbit now. Kuiper only has two satellites so far and OneWeb has several hundred, and of course, SpaceX Starlink has the most. For the most part, when we've reached out to these and other satellite operators, they have at least been willing to meet with us. They don't dismiss us outright, but we're not their top priority. As you can imagine, it’s one more stakeholder asking them, “Hello, have you considered how bright your thing is?”, and they're just like, “Fine, we'll have a meeting.” In many cases, they have their own proprietary data sets and observation techniques to figure out where their satellites are and how they're performing, but they're not willing to share any of this, because it's all industry secrets. Which, as an academic, is annoying because I like everything being open source, but I know that's not realistic.

We have several folks that we're in contact with pretty regularly through the CPS, to touch base with different industry companies and maintain contact, to check with them what questions they have for astronomers and what feedback they need. So we really do want to keep those lines of communication open. But I also worry, as more and more operators come on the stage, that this doesn't scale very well, because it's like six of us trying to do this. Suddenly, there's going to be dozens and hundreds of companies from countries around the world. There are dialogues and we do our best to keep the communication channels open, but we're not anybody's top priority.

IO

How close are we to having a guideline which would enforce satellite brightness and sustainable practices?

MR

Really far away. There was a little positive movement—there was a news item just a week or two ago that the UN agreed to put on their 2025 agenda a bullet point about satellite constellation interference affecting astronomy. Which is actually kind of a big deal, because they've never previously agreed to even talk about it. So the fact that they're going to talk about it in a year’s time is a big deal. But again, the times are so different, right? Industry is just so fast. Academia is kind of dragging our feet and slow because, you know, publications take a long time, as do peer review and funding cycles. And then over here in another universe is the UN or any kind of regulatory body, and it's just a completely mismatched timeline, which is a big challenge. But we have made progress, a lot of people have worked really hard.

IO

On that note, you've been looking at this problem for a while now. What we're trying to do is to communicate this problem to a different audience, and we’re interested in the gap of imagination that takes place because of the scale and thus invisibility of outer space. How do we start to communicate and visualise the environmental problem and its repercussions? One way, for us, is to consider analogous sites on Earth, such as the Great Pacific Garbage Patch, to reflect on our attitudes towards waste.

MR

I use that analogy sometimes myself when I'm trying to explain this to people. I’ll say it's a little bit like the ocean, where, sure, you can dump some amount of garbage into it, sometimes. But you can't just do that forever, and you can't do it in huge amounts, it's not an infinite well that can just hold everything.

IO

Exactly. A few people that we spoke to use this analogy as well, because of the scale. You have your big chunks of plastic floating in the sea, but you have your microplastics, which you can't really see, which are very much like the smaller pieces of debris in space that can’t be tracked.

MR

And unknown repercussions too; we don't really know what the long term impacts are, but we're doing it anyway. For satellites, I sometimes also use a bug-on-the-windshield analogy. You're trying to look through a clear window or windshield and there's all this stuff in the way, another layer, if you will.

IO

For the installation, we’re creating collision mobiles that are a way for us to speak to the disruption of the night sky by satellites and space junk. LED lights and punched metal fragments rotate in unpredictable movements.

MR

I like how it lights up to indicate a collision. One thing that jumped out at me is that thankfully, for the most part, all of the streaks that we see in astronomical images are due to reflected sunlight on satellites. None of them are emitting billboards or lasers at the Earth yet. That may come to pass—I hope not. So it's all incidental reflection, and none of them are actually emitting light themselves. I liked that you had reflective pieces. It makes me wonder if you had a bright light that was off camera, or at the edge of the room causing weird glints depending on the angle that you're standing at, that would be another pretty realistic way (unfortunately) to represent how the streaks that you see in images come to pass, because it's a matter of geometry. When it's edge on, you don't see it, because the light is reflecting in some other direction, but then when it glints at you, it's like, oh, it’s in the picture.

IO

Yes. When we were filming the installation, the metal pieces not only reflected the external light but also the light from the LED that flickers on and off.

MR

Which is almost like a debris cascade, a representation of the possibility of Kessler Syndrome, where if a few collisions start happening, it creates a runaway effect. There are many analogies that you could possibly insert here.

YT

We’re also including one of Trailblazer’s images in the catalogue for the exhibition. It’s part of a visual essay on holes, playing with different types of holes in a visually associative manner.

MR

They are holes in the data or parts of the image where there's no information available, because it's just a streak. It's interesting that in this image, it's actually inverted from the actual photograph, but it's perfectly legitimate. I think Trailblazer just shows them like this because it's easier for your eye to see a dark stripe. But in reality, the way the picture shows up is that the stars are bright, the background is dark, and then the streaks are white and bright. There's a lot that goes into deciding exactly how to stretch and display astronomical data because there are only 256 options for colour or brightness on a monitor. But in reality, any pixel in your data can go from like one to 70,000. There's an app called SAOImageDS9 known for displaying astronomical images. If you downloaded the raw image from Trailblazer, you could open it with this software and play around with the stretch and scale of the image. You can even change the colours and how much the streaks stand out or fade into the background, if you want the stars bright or dark, or use different colour maps. It's also a tool that astronomers actually use to look at their images, and make decisions on how to compress and display them.