UARS down over the Pacific ocean

By Phil Plait | September 24, 2011 9:02 am

NASA has confirmed that the Upper Atmosphere Research Satellites, UARS, burned up over the Pacific Ocean last night, September 24, between 03:23 and 04:09 a.m. UTC (11:23 p.m. and 12:09 Eastern US time). I have no other reliable information on it, but I expect we’ll get more updates soon. There were lots of reports last night of it falling over Canada, but those were mistakes or hoaxes. Apparently some people were fooled by meteors, Chinese lanterns, and possibly even the planet Jupiter. That’s happened before.

If I find photos or such I’ll try to post them, but I’ve heard no reports of witnesses, and I’ll be away this afternoon for my TEDxBoulder talk, so if any pictures turn up I may not be able to get to them. I imagine SpaceWeather will post any if they crop up.

Thus ends that saga. If you’re curious, you can read about the history of the UARS and what we learned from its 15 year mission to investigate our planet’s atmosphere.

Related posts:

UARS update 5: new predicted re-entry tonight at 05:10 UTC +/- 2 hrs
UARS update 3: new predicted re-entry tonight at 03:16 UTC +/- 5 hrs
UARS update 2: new predicted re-entry at 00:58 UTC
Update: satellite *might* fall on Friday at 22:00 UTC +/- 9 hours
NASA satellite due to burn up some time in the next few days

MORE ABOUT: re-entry, satellite, UARS

Comments (42)

  1. Mike

    So if it didn’t fall over Canada what is this video showing?

  2. Peter

    If it came down over land, would it make enough of an impact to be detected by seismographs?

  3. What? Jupiter fell too? I’m guessing THAT would definitely have been detected by seismographs.

  4. Christopher Jablonski

    It may be anticlimactic to some people, but I’m sort of comforted by the fact that it fell in the Pacific. It fell in the place it was most likely to, and I like that. And think I read one too many Twitter freakouts last night of people convinced it was landing in their area.

  5. Tom

    One thing I was never sure of: NASA said that if UARS fell on your property, “don’t touch it, it’s ours.”

    But if the satellite would have come down on land in the middle of nowhere (I.e middle of the Sahara) would NASA have sent a team to retrieve it?

  6. Len

    Ok, what happened to NASA saying they would be able to pinpoint it’s crash 2 hours before it crashes? Now, they have no idea where it crashed, except to say some where in the Pacific? There’s just been too much dis-information and wrong information from NASA and news sites. And what, no one spotted this huge show it was supposed to be? The show NASA said would be like long last fireworks? Ok, right NASA (nasa = never a straight answer)

  7. Mike Saunders

    You deleted my message questioning that it is a hoax? Really?

    Shouldn’t being a skeptic being reserving judgement before all of the information comes in? Otherwise you get situations like the crater in Peru where something bizarre happens.

  8. Magenta Cadigan


    Yes, they would. If it comes down in a retrievable location, they will indeed retrieve it.

  9. VinceRN

    So I guess it went down just before I was outside looking up last night. So close! If it had just stayed up 10 minutes more…

    Still, beautiful night, well worth being outside looking up.

  10. Alan

    So sometime during a >0.5 hour window it entered the atmosphere over the North Pacific off the US coast and nobody really knows where it went from there.

    People often say that if aliens are visiting Earth, then where’s the evidence? Well, a massive satellite which we knew was coming made a fiery reentry across the sky, with untold millions of people waiting and watching, and half a day later, with all our technology, nobody is even sure what continent or ocean it came down on.

    Perhaps hard evidence of what is in our skies is not actually so easily obtained, as intuition suggests.

  11. Mike

    So it burned up over the area it was most likely to burn up over? How dull.

  12. flip

    I’ve been watching Jupiter over the last week as it’s been fairly clear skies and Jupiter happens to appear over my backyard in a good spot for watching. It’s pretty obvious (never actually having seen Jupiter before, and looking it up in Stellarium the first night I saw it because I thought it was unusually bright for your average star) and I definitely wouldn’t confuse it for a falling star, let alone a falling satellite. I don’t know exactly, but surely a falling satellite would move across the sky faster than Jupiter?

    Slightly off topic, but I just spent 20 minutes watching the moon rise at dawn. I could see a star peeking out next to it too, and the dim light from the sun made for a nice silhouette against the unlit portion of the moon. Pity I can’t yet figure out how to take photos using my telescope.

  13. KC

    One of the big problems with this sort of thing is that people have expectations that are far too high. They figure that we can send a probe to one of the outer planets with pinpoint accuracy.
    They don’t get that a dead satellite with no power or fuel is a totally different situation.

    People also have this huge misconception that NASA and/or the government and/or military have an impenetrable worldwide radar detection system that can track anything from the size of an asteroid down to a sparrow at any speed in three dimensions at any distance from the ground to Pluto. This of course is complete nonsense and spawns a lot of conspiracy theories about the space program and UFOs. etc.

    “If it came down over land, would it make enough of an impact to be detected by seismographs?”

    If the entire thing had survived to hit the surface then yes maybe – but this thing must have broken up into hundreds of pieces – the largest piece NASA estimated would weigh only 300 pounds. By the time reached the surface it wasn’t moving very fast, so any signal on a nearby seismograph probably would’ve been indistinguishable from a fallen tree.

  14. flip, yes, a falling satellite would move across the sky much faster than Jupiter.

    You have demonstrated that by looking at the sky for a few nights and looking up what you see in a reputable source, you now probably know more astronomy than 99% of the population.

  15. Kim

    In the BBC article they feature the images that Thierry Legault took. And he is “only an amateur”. Cool.

    And for Mike Sanders saying Phil deleted your post because it called it a hoax? Were there any swear words or insults in there? That is normally the only time comments are deleted. Otherwise he leaves them here so the rest of us can roll our eyes and explain.

  16. mike burkhart

    I have an idea ,with big pices of space junk why not put a massive amount of explosives on it and blow it into smaller pices that will burn up in the atmospher thus no one is in danger.

  17. Giffy


    A massive satellite would still be a very small space craft. Is it possible an alien craft flow over the pacific once or twice and we missed it? Sure, possible in the same sense of anything. Is it likely that we have been repeatedly visited by aliens over land? No.


    Not a bad idea but that would add weight and more weight for that leaves less for science. Plus if the explosives failed to detonate then what you would have is a falling bomb.

  18. flip

    #14, Vagueofgodalming

    Thanks for the info! I remembered reading one of Phil’s posts and how the satellite was falling slowly, but wasn’t sure how slow was slow.

    As for the second sentence: it’s all thanks to this blog. A few years ago I wouldn’t know the difference between Jupiter and a satellite, so I can see how some people could mistake the two based on ignorance alone. My getting a Galileoscope late last year has helped me learn more about what I’m looking at, and Stellarium has helped me pinpoint what I see and give me labels. (Thank FSM for good computer programming!) Phil and others have helped me understand more about what’s up in the sky technology-wise. The stuff I’ve learned has made me wonder why in the world I didn’t know it before. (And wishing I paid more attention to science at school)

    My only problem is that I don’t know *enough* about astronomy. But then, that’s why I keep coming back here 😉

  19. Chris Crawford

    The fact that many amateurs in the Pacific Northwest failed to see it does not mean that it didn’t pass over that area. I was watching from southern Oregon and I knew exactly where and when to look. I scanned all around the region for 15 minutes, using 10×50 binoculars. I never saw anything. But if the thing were still in action at that times, I wouldn’t have seen it unless it was burning up, because it was in the earth’s shadow and so had no source of illumination.

    The video from Canada looks quite realistic to me. The fragments are in the right location, they’re moving the right direction at about the right speed, higher ones seem to overtake lower ones, they disappear after a period of time — all the basics check out. Sure, it’s shaky, but if you just track the objects themselves, it’s pretty clear that these objects are in the sky and moving.

  20. Randy Owens

    Wouldn’t it be down under the Pacific by now?

  21. Pete Jackson

    UARS smoke. Don’t breathe this!

  22. Stargazer

    #16, I believe UARS was originally intended to be retrieved by a Shuttle, and not left to reenter uncontrolled.

  23. Chris Crawford

    I’ve been searching high and low on the web for any additional information on that Okotoks video and all I can find are numerous statements that the video was quickly debunked. But I can’t find any actual debunking. Some people pointed out that there were no debris trails, but that doesn’t concern me: a falling object in a flat trajectory behaves in much the same manner. There is a class of meteors that just skim through the top of the atmosphere, traveling all the way across the sky, and these sometimes lack extended trails. Moreover, the objects are low on the horizon; it’s hard to say with any certainty but I’m guessing that they were only about 10º or 20º high. That would put them at a distance of about 100 to 150 miles away from the observer — far enough away that only the brightest portions would be visible.

    The best argument in favor of the hoax hypothesis is that the trajectory projection I saw suggests that UARS would have tracked a bit north of Calgary, but this video plainly shows it passing well to the south of Calgary.

    Various news outlets are claiming that the local police have not gotten any reports of something hitting the ground nearby, which most people seem to interpret as evidence that the video is a hoax, but in fact the video suggests that the impact site would be east or ESE of Calgary, at a range of maybe 100 miles, which would put the impact zone along an empty, unirrigated section of prairie between Brooks and Medicine Hat.

    The only other debunking I have seen on the Internet is a bunch of yahoos claiming that the video shows streetlights. I would very much like to think that I am not a nutcase and accept the NASA version, but so far, the evidence seems to me to favor the Okotoks hypothesis.

  24. Keith Baker

    The shaky video (post 1)seems to show me lights in the sky, sofar so good,
    Moving kind of horizontal for a long time, that is less likely to be something falling out of the sky, more like floating on the wind.
    Also they seem to move very slow and that is not like anything else I have seen falling out of the sky and burning up!

    I think these are chinese lanterns set off deliberately to create the story (not by the taker of the video perhaps, no accusations of that kind from me).

    Also if the higer lights are closer, they would seem to overtake the lower ones (further away) due to their changing perspective relative to the camera, is logical really.

    If this video was taken by day you would not have thought it strange at all I think.

    In my mind this way of looking at it is the skeptic mind reserving judgement. Not the other way around. We can believe everything we are told and filter later or do it the scientific way and quickly derive at what is right.

    There are times I can remember I nearly crashed my car thinking I saw something incredible….only to find out it was some chinese lantarns. The damn things are hard to identify at night high up in the sky. I am wondering if NASA would have taken responsibility for one of these burning somebodies house down as they are a direct result of their satelite re-entry news?

  25. HvP

    Chris Crawford,

    How about the possibility that what was seen in the video is merely a swarm of Chinese lanterns floating on the breeze?

  26. Robin

    Hoax. That’s funny.

    In the late innings, critical thought is still far behind. If it doesn’t score quickly, the game will be lost.

  27. I knew it! :-)

    Been away camping for last few days but had heard of the “falling rogue satellite and the risk it poses” on the news – their words not mine – & said to my brother and folks that it’ll probably come down over the ocean somewhere and burn up without many if any folks even seeing it. Now, sure enough, that seems to be what’s happened. 😉

    Pretty safe prediction really given that ocean is two-thirds of our planet’s surface. The view from a randomly generated spot on our globe anywhere is water on every horizon far as the eye can see and no land in sight. Seems Earth is really rather mis-named .. 😉

  28. Ian S

    Mike and Chris Crawford, this is what a falling satellite looks like: l note the large persistent trail, the sudden changes in brightness and colour as it breaks up, the increasing number of light sources and persistent trails..

    This is what chinese lanterns look like: skip to 1:30 if you don’t want to watch the whole thing.

    Now which looks more like the canadian video?

  29. Robin

    @ Messier Tidy Upper (#26): Hmmm. Interesting assertion that the “view from a randomly generated spot on our globe anywhere is water on every horizon as far as the eye can see and no land in sight.” It certainly seems to fly in the face of your statement that water covers about 2/3 of the Earth’s surface. Let’s examine your claim from a factual, objective point of view.

    The Earth’s surface is approximately 70.8% water. Now, since your statement specified any random point, I’m going to assume that you were basing your “predictions” on random numbers from a uniform distribution. That means that any point on the surface of the Earth has the same probability of being THE “randomly generated spot”. After all, there’s no logical reason to have unequal weights for different points on the globe. Immediately that tells us that the actual probability of that “randomly generated spot” being in water is 70.8%. Here I assumed that the Earth was large enough that application of the Law of Large Numbers was appropriate, thus meaning that 70.8% from the water surface area of Earth was also the probability of landing in water. Interesting.

    Of course, just being in water does not mean that you can only see water on every horizon, and since the distance to the horizon while in an ocean is about 3.4 miles, that means that randomly generated spot will have to be more than 3.4 miles from any land mass. We can immediately then exclude all the water within 3.4 miles of a coastline. So the probability of seeing only water on every horizon goes down and is significantly less than 70.8%. The probability goes down even further when coastlines aren’t perfectly straight, say when a coastline has a sharp 90° bend or summat. That’s not all: that water “exclusion zone” will have to be extended because, of course, land is not all at sea level. That means the water exclusion zone will have to be extended yet again due to things like hills and mountains. Likewise, you’ll have to extend the exclusion zone to account for large cities with tall buildings.

    Alas, it seems that you just can’t land anywhere on Earth and be guaranteed to see only water on every horizon. Of course, it’s still more likely that you’ll only see water, but more likely wasn’t your claim. Once again, statistics trump hyperbole.

    If you’d like, I can send you an executable file that will generate random coordinates that you can then check in, say, Google Earth. Please note, you’ll need to check a lot of points to satisfy the Law of Large Numbers. Checking only 10 or 20 won’t be sufficient.

  30. Randy Owens

    @Robin #27: While you’re at it, you should ask, “what’s a meta for?” Doesn’t seem like you know.

  31. The Okotoks video from Canada is suspect for a number of reasons.
    (1) The lights are too steady, show too little evolution. Real reentry events show fragments appearing, disappearing, flaring, and being clearly of different brightness.
    On the Okotoks video, nothing of that all. Steady lights, a string of them. This is more typical of someone releasing a series of Chinese lantern balloons.
    (2) While reentering satellites cause long duration fireballs, the video is abit too long (6 minutes!)
    (3) Somewhere in the video, the guys says its September 22nd – a wrong date.
    In short – there is too much just not right about it.

  32. Chris Crawford

    Given the many uncertainties about this video, I’m certainly willing to give the Chinese lantern hypothesis some credence, but the strongest argument against that hypothesis comes from the spectral signature of these things — bluish. The incandescent lights in the same video show up as strongly reddish, so we can’t pin the difference on the camera. Chinese lanterns are strongest in the sodium lines.

    But that raises an additional difficulty for the UARS hypothesis: there SHOULD be some red in those images, if only because they’re so low in the sky. Which suggested to me another hypothesis: goose bellies. There have been a number of cases in which ghostly sets of lights turned out to be flights of geese reflecting ground lights. This would put the lights much closer to the observer, which would eliminate the reddening effect of transmission through lots of atmosphere.

    I did a quickie analysis of motion of the lights and saw some big variations in speed of the objects. This, I think, is the single most informative datum we have. Falling satellite pieces would not show that much variation, and they certainly would not show any positive third time derivative in spatial position, as these do. For that matter, Chinese lanterns would also show very little third time derivative. Geese would provide the closest fit to that datum.

    I am therefore ranking the hypotheses as follows:

    1. Geese: very likely
    2. UARS: unlikely
    3. Chinese lanterns: unlikely
    4. Fake: very unlikely

  33. Messier Tidy Upper

    @29. Robin : I’m going to assume that you were basing your “predictions” on random numbers from a uniform distribution.

    Actually not so good an assumption.

    I was basing my prediction simply on the fact that two-thirds of Earth is covered by water plus looking at a globe.

    Look much area the Pacific ocean covers – that’s pretty much half the planet if you centre the globe on the strip of land starting at the Bering sea between Alaska and Russia and going south all the way to Antartica. No, the distribution of water (& land) isn’t remotely ressembling uniform. But still 70% of the planets surface is ocean versus 30% land. 😉

    Of course, it’s still more likely that you’ll only see water, but more likely wasn’t your claim. Once again, statistics trump hyperbole.

    My claim was :

    “The view from a randomly generated spot on our globe anywhere is water on every horizon far as the eye can see and no land in sight.”

    Which was non-specific as whether it was “more likely” or whether I was guaranteeing anything. What I was meaning (& maybe I could’ve phrased it better) was that if you land at any randomly chosen spot on the planet you are most likely to be landing in water and out of sight of land.

    Likewise, you’ll have to extend the exclusion zone to account for large cities with tall buildings.

    OTOH, waves can reach reasonable heights and block your view of land too – and there’s the question of fog, precipitation – and darknesss – to consider when assessing visibility of land too! I’ll leave out the question of tides and sea level rise plus whether we count the smoke from distant volcanoes, fires and dust hazes plus mud, reefs and sand flats as sighting land or not! 😉

    If you’d like, I can send you an executable file that will generate random coordinates that you can then check in, say, Google Earth. Please note, you’ll need to check a lot of points to satisfy the Law of Large Numbers. Checking only 10 or 20 won’t be sufficient.

    That’s a pretty neat & interesting idea – thanks for the offer. :-)

    Of course, another experiment is to get a map of the globe, close your eyes, trace along the UARS descent pattern and after a random number of seconds, stop and open your eyes. Variations and tighter methodological approaches to that can be addded as best suits , eg. random number generator, participation of a second person, use of varying sized globes or world maps, etc …

  34. icemith

    @28. Ian S Says:

    “… This is what chinese lanterns look like: …”

    I checked out this video, and while it was a cute idea, I concluded it was not a good idea at all.

    “Hmmm… nice idea, BUT, would you be happy being the cause of a forest fire? What goes up must eventually come down. Can you be certain that it is definitely extinguished?
    And if twenty or more guests set their lanterns off, the winds, even gentle breezes, would scatter them after a while, spreading the chance of a fire. Dried grasslands, wheat fields and suburban areas would be in jeopardy.
    So, nice idea? No, no, no. Sorry.”

    In fact I would not be surprised if it is an illegal activity in some jurisdictions, certainly with the high likelihood of fires, and especially in dry Summer months.


  35. icemith

    My other comment re the UARS re-entry, concerns the physics of a falling body through the atmosphere.

    That the Satellite has already met its demise, exactly where, well nobody has reported finding anything (yet), it still must have been subjected to the laws of physics, whereby after mostly burning-up in the upper atmosphere, due to friction, the remaining, probably insulation, some objects being shielded by something else “taking the heat”, or very light-weight objects and other non-flammable bits may have reached lower, but denser and cooler bands of that atmosphere, where they would have slowed to zero.

    At that point, the objects would be just falling to the surface, and the resistance of the air would be the major restriction to the velocity it could achieve. Not unlike the maximum descent of a free-falling parachutist, before he pulls the ripcord. If I remember correctly, that is about 120 mph, approx, 200 kph. It just will not go any faster, unless there is an external influence.

    The woman who was reported to have been hit by a piece of insulation in the Eastern USA, a few years ago, was lucky it was only a small piece. If it was larger, and more solid, it would have been a different story. Imagine being hit by a cricket ball/baseball, just dropping out of the sky into the stands, or being hit by the same at the fastest rate ever bowled or pitched, then you would definitely regret it, or maybe not, if it was your demise!

    I wonder if some Hollywood writer is not at this moment, working on the first draft of an Airliner being struck by a larger piece of satellite as it flies across the ocean. (Of course, the chances are twice as likely than flying across land!) But it could happen……


  36. Robin

    MTU, actually your proposed experiment is only valid if a person can randomly point at any position on a map and if that person does that a sufficient number of times to satisfy the Law of Large Numbers. That will be a lot. Still, it is important to make clear when one is talking fantasy and when one is talking about reality or what is probable. After all, statistically, about 30% of those random points on Earth will have those randomly positioned people not only seeing land on the horizon but also having it under their feet.

  37. db26

    With all our technology we can’t even determine when or where a useless piece of junk enters the atmosphere?

  38. Ian S

    Chris Crawford,
    Not sure how you think you can calculate the speed of the objects when you don’t know how far away they are or how large they are, but hey at least you have recognised that its not the UARS sat coming down..

    Not my wedding, not my idea of a good idea, just posted it for comparison…

  39. Chris S

    Ah, Phil. About your reference to NASA’s report … say what?

    From the NASA release … “NASA’s decommissioned Upper Atmosphere Research Satellite (UARS) fell back to Earth between 11:23 p.m. EDT Friday, Sept. 23 and 1:09 a.m. Sept. 24”

    Isn’t this a 106 MINUTE window? For an item with an orbit that was already down to 87 minutes last time I checked?

    Now – if you’ve misquoted NASA, then there is a very wide range of areas covered by that time slot. However, if you have some other NASA source, and NASA has produced a typo, then 12:09AM EDT is about the time that the orbit would have crossed the edge of North America — and if it wasn’t found at that point, then it would have been deemed ‘down’ no later than that.

    But I think NASA is being clear about these numbers. The release at

    and the final orbit map at

    are both consistent with the 106 minute windows.

    So – NASA has confirmed that it DID go down between 23:37 and 1:09, and they have confirmed that it is LIKELY it went down in the Pacific.

    In the absence of more evidence (which will itself have to be subject to verification!) that’s the most certainty I think we should admit here.

  40. Eric Tolle

    Based on my calculations, the satellite came down more our less intact on a small island to the southwest of Hawaii. There seven castaways will use it in a wild attempt to escape back to civilization.

  41. icemith

    @38. Ian S.

    Phew, Ian, I was hoping not to offend anybody, but I understand your reasoning. I did not really understand exactly what a Chinese lantern was, only recalling the adornments as shown in Movies concerning outdoor Oriental festivities etc, and these were definitely attached to wires, posts and trees, making quite colorful decorations.

    Here, in Australia, we do not see them in the fashion as depicted in the wedding video. They, like firecrackers, are totally banned, and the only time they are seen is at community displays on a city wide festival such as New Year’s Eve, where the skyrockets are fired from barges in the Harbour, big bridges and tall buildings. And by specialist Pyrotechnicians.

    So I have not seen those lanterns ever, but I can understand the wonder and unusual display that would have amused the ancients. They were their “balloons”, that are now a feature of every modern kid’s experience. And the grown-up versions are very interesting too, as Phil found out.

    In my childhood home district, on the sad occasion of burying my mother in 1995, days later a Hot-air Balloon was observed passing overhead, as we drove through the streets, when it descended and realising that it had landed near houses, could have been in serious trouble. As it happened it did land … in the front yard of the pleasant little motel style retirement village complex, and right in front of the unit that my late mother had occupied, until a few days before.

    Now I personally don’t take that as a “sign” or “omen”, just a co-incidence, but it sure was memorable, and a touching memory of a loving Mother. The balloon did have an incident with a faulty burner/gas tank, no damage was done, nobody was hurt, and after repairs, lifted off again, into the wide blue yonder.

    Ivan. (I’m tempted to say, “Thanks Mum”.)

  42. Chris Crawford

    Ian, you wrote:

    “Not sure how you think you can calculate the speed of the objects when you don’t know how far away they are or how large they are,”

    I was referring to the angular velocity, which is fairly similar for all satellites. UARS was lower and faster, but it was also at a lower angular altitude, so we’d expect its angular velocity to be in the same ballpark. I’m using very approximate numbers here.

    and then:

    ” but hey at least you have recognised that its not the UARS sat coming down..”

    Whoa! The result isn’t what matters; what matters is the methodology. I’d instantly drop the goose belly hypothesis if somebody came up with compelling evidence for something else. Right now, the UARS hypothesis to explain that video is still viable. The key factor that a lot people don’t seem to realize is that, in special circumstances, a falling satellite really can appear to leave no trail during part of its descent. It all depends on the density of the air it hits, and the extreme upper reaches of the atmosphere can be highly variable in this regard. That’s why they couldn’t predict its fall with any accuracy.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar