Hacker News new | past | comments | ask | show | jobs | submit login
Eu_sst: 20% probability for large satellite collision (twitter.com/eu_sst)
142 points by AlphaGeekZulu 29 days ago | hide | past | favorite | 95 comments

The two satellites involved are:

OPS 6182 (10820, 1978-042A): https://www.n2yo.com/satellite/?s=10820

SL-8 R/B (12443, 1981-041B): https://www.n2yo.com/satellite/?s=12443

They are on head-on course. With a probability of about 80 percent they will miss each other. With a probability of more than 20 percent they will impact at a speed of more than 50000 km/h in an altitude of 780 km.

The two objects are not important (an old Russian rocket piece and an old US weather satellite).

But upon impact, the objects will create a large cloud of debris, that might collide with more objects, in the worst case in a chain reaction.

The danger is not falling debris, but that they will produce more damage in their orbit, maybe hitting more essential space equipment in the future.

No need to hide in a bunker.

Did they miss? Both those links show them still there, but I suspect they are calculated not real time

They missed

Can somebody give a layman explanation why the debris would stay in the same orbit? I would expect a lot of energy to go into destroying both objects, especially in a head-on collision.

Put quite simply - the energy doesn't go into destroying objects. There's only a very small amount going into heating things up from mechanical deformation [and a bit into spinning], but fundamentally, splitting something in a million pieces just means you have a million pieces that have the same total kinetic energy in orbit.

However, they also exchange kinetic energy & impulse with each other, meaning you get a mix of slower and faster pieces. This is what the Gabbard diagram shows:


Anything that is sped up in the collision (i.e. "steals" energy from another piece) has the same perigee (orbit low point) but a higher apogee (orbit high point.) The reverse happens to anything slowed down / stolen from. That's why the graph looks like open scissors.

Some pieces with lower perigee will friction against earth atmosphere and burn up, but the ones with higher apogee become a serious problem...

(also, for anyone who can't remember to distinguish apogee & perigee: think of your periphery, it's the things close to you :D)

My mental crutch was always Apogee Software. It's more intuitive that someone calls their company "High Point Software" than "Low Point Software".

I think of "peripheral" as being "far from the center of my vision".

Fortunately, years of Kerbal Space Program means the apo/peri prefixes are pretty solidly ingrained.

Ah crap :/

…well I guess the "center of vision" is the center of the orbit ellipse… ;D

(just trying to salvage the mnemonic aid for myself)

My mental crutch is "apotheosis" (the high point of something)

A lot of the debris won't stay in the same orbit. It'll be like a mini-explosion, spraying out debris in every direction, but then add the existing velocity to that. So yes, the debris will be in another orbit. The question is, how much of a change occurs.

In the twitter thread, there is a Gabbard diagram that shows a prediction of the orbits that the pieces will end up in. The satellites currently have a reasonably circular orbit, with apogee (highest point) and perigee (lowest point) about 800km. After the collision, the apogee is likely to be higher and the perigee is likely to be lower, so the orbits won't be circular any more.

The way this happens is - if the debris piece is knocked directly backwards, then its apogee will be about the same, but its perigee will be lower. If it is knocked forwards (yes, that can happen), then the apogee will be raised, and the perigee will be about the same, and the piece could remain in orbit for a long time. If it is knocked up or down, then both the perigee will be lowered and the apogee will be raised. If the perigee is low enough, then the pieces will scrape through the upper levels of the atmosphere, and de-orbit eventually.

It looks like the prediction is for the lowest perigee to be about 200km. It may take a long time for even those pieces to de-orbit.

It doesn't. It splatters into many related orbits.


Conservation of momentum. All of the teeny little bits of the two colliding objects still have the exact same average momentum of the two original objects, now there are just hundreds of thousands of them flying off in slightly different directions. The movie Gravity actually gives a fairly good, if overly dramatized, example of what happens when objects collide in space.

Does the debris have a degrading orbit and eventually get pulled earthward and burn up on entry? Or is that chaff out there floating around forever just outside the gravitational pull?

Some of the pieces will have a lower orbit, which might be low enough for them to scrape the atmosphere and bring them down. Some of the pieces will be placed into a higher orbit (although the lowest point won't be raised).

In the case of these objects, at ~800km, the time for orbital decay is often measured in decades.

Try millennia...

Excellent explanation, thanks very much!

Missed? https://twitter.com/EU_SST/status/1380601898164744199

> UPDATE: #EUSST’s network of sensors has only detected a single object or echo at passes over three radars after close approach. Most likely, the collision between SL-8 R/B and OPS 6182 did not take place. We will continue observing the objects to confirm this assessment.

Leolabs is tracking as well. Their data suggests a slightly lower but >2% chance of collision


The Leolabs Visualizer is pretty amazing. I didn't understand what the red gates were on the map until I looked inward - they're running their own radar sites in different parts of the world to track the objects privately. That's cool.


Puts away popcorn

That's somewhat more comforting, but alas, hope alone is not a strategy.

You can watch a map of both here!


How soon after the event will we find out whether they collided?

Tangent: I went to find UTC, and they've redesigned time.gov, it looks great on my desktop.


My personal favorite time/NTP website is



time.is is pretty great! Their calendar is real handy as well, especially to see duration between two dates: https://time.is/calendar

Love time.is. You can embed JS from them in your own webpages for custom clocks.

It happened about a year ago, and I think one of the developers commented on HN:


I never realized that there was two time zones in Arizona...

For anybody interested, they missed:


(Link further in thread.)

We don't know yet. Satellites (especially defunct) don't have active live tracking like ships. Satellite sites show them based on calculated trajectories.

Edit: as of 2021-04-09T20:21:00Z EUSST thinks they missed. https://twitter.com/EU_SST/status/1380601898164744199

leolabs' next anticipated contact with SL-8 R/B 12443 is at 21:06UTC https://platform.leolabs.space/catalog/L5136#next-planned-pa...

next anticipated contact with OPS 6182 at 20:00UTC https://platform.leolabs.space/catalog/L11618#next-planned-p...

So what's that, about 90 minutes from now assuming I've got the right info?

Are these satellites visible by telescope, can someone in e.g. Ankara try to spot SL-8 R/B 12443 as it as it flies over there in a few seconds?

Is there an automated amateur run satellite tracking network? Just tracking points of light at dusk/dawn is an easy thing and wouldn't even require pointing, imagers could be fixed.

  date -d "2021-04-09T17:18:21 UTC"
result: time in your timezone (on linux)

So, in about an hour.

How likely is a massive satellite collapse is to happen in nearest future? Are there any educated guesses/estimates?

The thing to look up is "Kessler syndrome".

It appears we have already passed the threshold at some altitudes, so a satellite collapse will happen in the future, but probably not for hundreds of years.

Anyone knows above where it could happen? I see that it's almost 800km above earth, but what about nearest lat/lon?



(north-east Russia)

Do near misses (hopefully) like this happen without us knowing, or do we track almost everything in orbit?

The current tracking infrastructure detects pieces down to 10cm. All of those are catalogued, get identification numbers, and are tracked.

Active sats often do small collision avoidance maneuvers.

Yes, but more frequently between small anonymous debris pieces (due to their quantity)

In summary, a rocket body and an old weather satellite want to have a nice big hug. At high speed. Producing a cloud of 4 million fragments. 400 of which will bigger than 20cm.

Except that they'll probably miss (by about 10m). Probably.

Really interesting how much of this is statistical probabilities, and how much we can model about this. Right down to particle sizes.

And yet, I still have so many questions. Mostly involving the maths from https://what-if.xkcd.com/13/ and whether we could push either of them out of the way...

I’m curious as to how they get those sizes and quantities! Too they have cad models of both objects or are they just taking the radar returns and material and making an estimate?

For well known objects like these, the mass and approximate size is known. You can Google them to find values even. The positional uncertainty and lack of data about attitude lead to modelling the objects as spheres as not so bad a choice.

One algorithm that is sometimes used: https://link.springer.com/article/10.1007/BF03546397

It would probably take longer than the notice you might get of a potential collision (5-7 days)

With a bit of luck, Sandra Bullock won't be involved in a satellite collision this time.

"Houston, I have a bad feeling about this mission."

Brace, brace, brace ...

Here's the text without having to execute arbitrary code from twitter.


Update: according to #EUSST's latest estimates using data from its surveillance radars, the close approach between #space objects SL-8 R/B & OPS 6182 expected by tomorrow will have a miss distance under 10m and a Scaled Probability of Collision over 20%! +updates to follow soon

See the below plot for the conjunction plane at Time of Close Approach (TCA) #EUSST #spacedebris #space


Update: #EUSST simulations indicate that the potential collision between the two #space objects would generate more than 4 million fragments. This plot shows the Delta-V distribution of the whole cloud of fragments #spacedebris


More than 400 of the fragments generated by the potential collision would be larger than 20cm. Gabbard diagram shows the extent of orbital regimes that these simulated fragments would reach. #EUSST continues monitoring the probability of collision. Stay tuned for more updates.


Latest update: according to #EUSST the close approach between #space objects SL-8 R/B and OPS 6182 remains stable in geometry and in Scaled Probability of Collision. Miss distance would be ~21m and Scaled PoC over 20%. This should be the last estimate until TCA.


Please omit flamebait from your comments here. You started a totally off-topic argument about Javascript which got upvoted to the top of the thread, gaining mass and choking out interesting discussion. Not cool.


While I agree with hiding the parent commit for the flamebait, that also hid my on-topic layman's explanation of what's happening :(

[Er… commit = comment… TGIF, leaving it in for fun ;D]

Good point. I've detached https://news.ycombinator.com/item?id=26752335, which was the return to ontopicness.

Replace `twitter.com` with `nitter.eu`:


I was pretty excited when I found out about nitter. After some months I realized that every nitter mirror only stays up for a few weeks before it becomes so popular it's throttled most of the time (like nitter.net).

Also, it's 17:41 UTC now... the events of this post should be occuring now. Or not occuring.

It's really easy to run a private instance of your own: https://github.com/zedeus/nitter

Mine isn't publicly accessible so never gets throttled.

I have one for invidious, probably should spin one for nitter as well :p

> Here's the text without having to execute arbitrary code from twitter

Ok, I'll bite, what's this about? What "arbitrary" code is being executed when you visit Twitter?

You're running Javascript. It's sandboxed, which is why I have it enabled by default, but often the javascript in a website is made to spy on you and your behaviour. It's not on your side but on the side of the people who made the website.

Personally I'm ok when something like discord breaks as it contains real time updates. But twitter just displays a bunch of text and pictues in a non real time manner. Even video can be played back without any js.

That's not arbitrary code in the security sense though. It's the very narrow range of things which are permitted to run in a javascript sandbox (more or less "things javascript can do in a web browser context, minus the extremely small chance someone is trying to drop a JS 0day on you using this link")

> That's not arbitrary code in the security sense though

True, you're right.

> It's the very narrow range of things which are permitted to run in a javascript sandbox

Well, yes and no. Narrow scope in terms of what other programming languages offer, yes. But in terms of limiting the amount of information that can leak from the user (you and your browser), it's not that narrow. And since est31 is making the privacy-argument, I'm guessing they are referring to that.

Cover Your Track by EFF showcases how much information is actually leaking and what Twitter et al could do with it: https://coveryourtracks.eff.org/

> That's not arbitrary code in the security sense though

Yes it is. It's not system level, but it is indeed arbitrary code - users have no lasting visibility into what it's doing, and little ability to control it.

> It's the very narrow range of things which are permitted to run in a javascript sandbox

Browsers have had a pretty naive threat model, up until recent Firefox with its new resist fingerprinting feature. Seemingly innocuous things like reporting your window size are vulnerabilities that are being used to track you, because the groups who designed the various web APIs had no security mindset besides sandboxing. One of the chief browser makers is even an unabashed surveillance company! It's going to take a long time to close or mitigate all those holes.

You can put a video on a page without js, but a lot of your customers would have a poor experience.

Apparently ≈2.24% of global users are using a browser that doesn't fully support HTML5 video: https://caniuse.com/video

That's one hell of a lot of people.

Are you able to translate this to layman's terms?

Should I find a bunker?

I assume the fear of this sort of thing is the potential for causing Kessler Syndrome.

You don't need to find a bunker, but there's a risk that we'll soon have a lot more 'space debris' in orbit. If a collision happens, it'll basically turn the colliding satellites into scattershot, leading to a higher likelyhood of other smaller collisions with other satellites (less spectacular but could still disable even more satellite). Not sure how likely the latter scenario is, but it could get very costly if it did happen.

No, there's not a risk of space debris.

>The Wayback Machine requires your browser to support JavaScript, please email info@archive.org if you have any questions about this.

did they hit or miss?

You're off by a day. We just passed 24 hours to minimum distance.

No, you are off by a day. This event happened, now we're waiting for confirmation.

Shit, you're right. I actually checked the date on my computer before posting. I swear it told me it was the 8th. What is happening.

Glitch in the matrix

it's 2021 now. They have presumably been tracking the orbits of there large objects for years now. Yes this is literally rocket science, but this collision is only a few hours away. Why can't they do better then a vague 20%?

You have random pieces of objects which may have already been impacted, tumbling, being pushed on by the solar wind, faint traces of Earth's atmosphere, and being tugged by electromagnetic and tidal effects.

These pieces are in orbit at many thousands of miles per hour, and have a random radar cross section, depending again, on whatever random impacts have already done to them, their orientation, composition.

The radar itself to track them must pass through the ionosphere, which itself causes variable time delays and phase shifts. Smaller objects return signals just above the noise after correlation in the receiver, most of them are just at the noise floor.

Observations only happen when the objects are in site of the tracking ground stations, which only cover a small portion of the sky, not including the expected collision volume.

All of that is just what I could recall might effect predictions off the top of my head... I'm amazed they can make predictions with any level of certainty.

The objects are about 500 miles up in the sky, potentially thousands of miles from the sensors looking at them, and traveling at 15,000 miles per hour with respect to us. Think about how precise your measurements would need to be to accurately predict the objects' positions in 24 hours to the nearest foot!

And that's the difficulty with the simplifying (and wrong) assumption that that outside forces don't alter the orbit- even 2021, we have a lot of room for improvement when it comes to predicting many of the other factors that come into play, like solar activity and the Earth's climate.

My guess would be because sats are not perfect spheres but - esp in the case of a rocket stage like here - complex objects which have non-trivial rotation patterns because the center of gravity might be off the visible center. So I doubt they model each and every piece of space junk out there to get a super detailed prediction - after all, what can you do if you know 95% it's going to be a hit between do dead sats vs if you know 80%.

Radar resolution of civil orgs might also be a limit.

Friction is difficult to model and is the dominant source of uncertainty.

Because the difference between collision and safely pass is less than 0.01% on the scale they model it.


Snarky replies aren't really helpful as an answer to a question - even if it was in loaded form.

It wasn't snark. This is how peer review works.

very funny. My sums should take me just a couple of hours...


Is it? Or are the current satellite view websites just showing predicted paths?


edit: nevermind

Worst case: Our planet ends up with Kessler syndrome and nobody will launch a satellite for quite a while... But in the bright side, I'm fairly sure your stocks wouldn't be alone in tanking if that happened.

All of those corporations are in the business of development and integration of satellites as well as satellite operations for commercial, public, and military customers.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact