Jonathan McDowell

Been exploring the amazing GCAT space dataset - it’s been a good way to drive some dashboard feature experimentation using fun data. Still need to work on my dashboard design skills, though. GCAT: https://planet4589.org/space/gcat/ Dashboard example: https://trilogydata.dev/trilogy-studio-core/#screen=dashboard-import&a…

Jonathan McDowell of the Harvard-Smithsonian Center for Astrophysics gave a great talk at the CfA on this topic a few years back. He created GCAT, a database that attempts to track all artificial space objects. https://planet4589.org/space/gcat/web/cat/index.html I'm looking forward to Public Observatory Night resuming in person someday. :(

The launch escape system is part of the Crew Dragon spacecraft, which has not entered service yet, and would probably not be used to deliver satellites anyway because its crew-carrying capabilities add significant weight and cost. Rocket launches have gotten fairly reliable in recent years -- launch success rates have been hovering around ~95% since the 1970s. [1][2] A 5% chance of killing 7 astronauts is unacceptably high at essentially any cost, hence the launch escape system on the Crew Dragon. However, for an unmanned launch, the entire launch escape and recovery system would have to add little more than 5% extra cost on top of the satellite and launch vehicle for it to be a sound purchase. A Falcon 9 rocket costs about $62M. A very complex satellite can cost over $200M, so for easy math, let's say that the entire system costs $300M, 5% of which is $15M. Those $15M have to cover the following: - powerful launch escape engines (e.g. SuperDracos) that are not recovered if the satellite reaches orbit - hydrazine fuel for those engines (highly toxic and reactive, has to be handled with great care) - carrying all that extra weight to orbit - parachutes and their deployment systems - hardening the payload to survive the physical stresses experienced during landing - refurbishing the payload after it falls into corrosive salt water Given all this, I am skeptical that a launch escape system for unmanned payloads is economically feasible. [1]…

Not a chance. Everything larger than 10cm is cataloged and there are multiple catalogs from different sources. https://svs.gsfc.nasa.gov/5258 https://planet4589.org/space/gcat/

Isn't Starlink already basically a distributed datacenter in space? they have like ~9k+ satellites up there already at least according to: https://planet4589.org/space/con/star/stats.html. what am I missing here?

Jonathan McDowell keeps track of all the Starlink Sat orbits, including failures: https://planet4589.org/space/con/star/stats.html

Not feasible. That would entail putting shrapnel into orbit (unlike extant anti-sat weapons which are short-range suborbital), which would mean a full orbital launch for every satellite target orbit. There's hundreds[0] of Starlink orbital groups already, so that'd require hundreds of independent orbital launches in a short timescale—far beyond China's launch capabilities today. [0] https://planet4589.org/space/con/star/planes.html (On general principles, you could argue you'd need 1:1 launch vehicle parity (number, not payload) to defeat a satellite constellation this way. For each satellite launch, you'd need one corresponding anti-satellite launch into that same, newly-defined orbit).

For anyone interested in current data like this, Jonathan McDowell maintains GCAT which is a General Catalog of Artificial Space Objects (and does so fastidiously). https://www.planet4589.org/space/gcat/index.html Be warned if you planning to ingest this dataset, the dates are fun =)

“Many” seems to mean only a dozen-odd. Which seems like an awfully small number to image the entire earth at meaningful levels of resolution or frequency. Which also seems to be the article’s lesson. FWIW, it looks like Starlink alone is up to 5,942 satellites in orbit [0] these days [0] https://planet4589.org/space/con/star/stats.html

Their 66 satellites are going to support a much lower aggregate constellation throughput and spot beam count vs StarLink. They’re fundamentally constrained by lift costs and low constellation satellite count. Iridium has 48 spot beams per satellite each covering 250 miles. Each StarLink cell covers ~15 miles with a similar spot beam count, with 5,442 satellites currently operational (as of this comment). https://planet4589.org/space/con/star/stats.html (Disclosure: StarLink customer, no other affiliation)

It's unfortunate that this misinformation is being upvoted on this site. The headline is completely incorrect, as is the statistics they're using to base the article on. https://twitter.com/Marco_Langbroek/status/1705562829225410697 https://planet4589.org/space/con/star/stats.html There has only ever been 350 starlink satellites that have deorbited, and only 8 within the last two months.

Actual source: https://planet4589.org/space/con/star/sg76.jpg

Thanks. See also: https://planet4589.org/space/con/star/sg76/ It looks like unit 30038 is almost back down to its insertion elevation.

I urge anyone interested in the topic to read T.S. Kelso's (maintainer of Celestrak) pieces on "Computers & Satellites" [1] and generally browse Celestrak - a true goldmine. J.C. McDowell [2] is also undertaking similarly impressive work. [1] https://celestrak.org/columns/ [2] https://planet4589.org/index.html

>absorbent or light diffusing material should be mandatory Absorbency is bad because it causes overheating and IR astronomy pollution. Diffusion is bad because it reflects light toward Earth. Specular surfaces (mirrors) can reflect light entirely into space and away from observers on the ground. This explains why SpaceX created their own radio-transparent, space-rated mirror. They are offering it—at cost—for use by other satellite manufacturers. https://www.teslarati.com/spacex-shares-how-its-making-starlink-satell… https://planet4589.org/astro/starsim/2022-05-07_SpaceX_Sustainability_…

Here’s the full list of Minuteman test launches: https://planet4589.org/space/gcat/data/launch/Minuteman.html

Analysis of what, exactly? The physics of the orbits they are in are pretty well understood. You can track the orbital altitudes (and failures/disposals) here: https://planet4589.org/space/stats/star/starstats.html

Apparently it takes about 100 days for the instruments to cool down far enough. https://planet4589.org/space/misc/webb/time.html

Here are most of the big ones, with timing: https://planet4589.org/space/misc/webb/time.html

Jonathan's Space Report (http://www.planet4589.org/jsr.html) is an exhaustive accounting of human activity in space (manned or unmanned), with a focus on recent launches. Released every few weeks. Back issues run to 1989 and continue to the present day. Emily Lakdawalla writes some very good explanations of space science for the Planetary Society - very accessible, but with more detail and intelligence (IMHO) than you get from other media outlets. http://www.planetary.org/blogs/emily-lakdawalla/

A dramatic factor within itself, not a dramatic quantity compared to everything else. I'm not sure exactly how much aviation fuel China burned in 2025, nor mass to orbit, but a quick search suggests that in 2024 they burned 38.2 million tonnes of aviation fuel[0], and launched 174.4 tonnes to orbit[1]. For the sake of a Fermi approximation, round that up to 40 MT aviation emissions and 200 tonnes to orbit, and assume a 20x factor of propellant CO2 emission mass to payload mass: 200 * 20 = 4e3 tonnes of CO2 from space launches 40e6 / 4e3 = 10,000 times more comes from aviation So, space could go up by a factor of 100x (compared to 2024) and still be a literal rounding error compared to just aviation. As per article, graph shows the nation as a whole is just over 12,000 million tonnes "from fossil fuels and cement". 12e9 / 4e3 = 3,000,000 If I download the image to get the full resolution, there's 586 pixels between the lines for 10,000 million and 12,000 million tonnes, so each pixel on that graph represents 2e9 / 586 ~= 3.4 million tonnes, and 3.4e6 / 4e3 = 850, so even if space launch grew by a factor of a thousand it would only shift that graph by one pixel. One and a bit, but closer to one. [0] Search for "38.2": https://www.greenairnews.com/?p=7427 [1] Page 13: https://planet4589.org/space/papers/space24.1.0.pdf

The article says "Instead, the spaceman (and only a man was considered) would have been launched on a parabolic trajectory some 300,000 metres above the Earth." I think that's a mistake. The V2 had a range of about 300 km ("For the V-2 rocket, just reaching space but with a range of about 330 km, the maximum speed was 1.6 km/s" - https://en.wikipedia.org/wiki/Sub-orbital_spaceflight#Speed.2C_range.2… ). The BBC article goes on to quote “Megaroc was essentially the Mercury-Redstone.” From https://en.wikipedia.org/wiki/Mercury-Redstone_Launch_Vehicle we see that Mercury-Redstone didn't quite make it to 200 km altitude. According to http://www.planet4589.org/space/jsr/back/news.526 , there have been no suborbital flights above 192 km. EDIT: Nope, I'm wrong. Megaroc was supposed to get to 304 km (one million feet). http://www.bis-space.com/what-we-do/projects/megaroc . Indeed, it appears that the name is short for "Mega[foot]roc[et]."

Starlink isn't at the top of my list for "likely Kessler Syndrome starters" because its a rather low orbit. They're at 550km up. From https://www.spaceacademy.net.au/watch/debris/orblife.htm puts it at likely a 10 year lifespan with some variation. The published lifespan of a satellite is 5 years. Its the stuff in the 700km to 900km range that's a problem. http://stuffin.space https://www.satview.org/spacejunk.php - and you'll see things like STARLINK-1204 already getting deorbtied ( https://www.n2yo.com/satellite/?s=45203 - note its perigee is 204.0 km - its already well below the other orbits) https://in-the-sky.org/spacecraft.php?id=45203 for more on that satellite. Starlink-61 appears to be one of the oldest ones still in orbit - https://in-the-sky.org/spacecraft.php?id=44249 though its on its way down and you can see the type of impact and the duration of the "ok, this is going down". Even with the maximum starlink, that's 1,700 satellites that will last about 10 years. The 2007 China anti-satellite test created approximately 150,000 debris at 865km. Those are in an orbit that is on the order of 500 years. Starlink isn't going to be the cause of Kessler Syndrome. --- In hunting this up, I also found https://planet4589.org/astro/starsim/index.html which is relevant to the main topic - the impact on astronomy.