Last updated 2015 Nov 13
The following is all in the public domain (except for certain links provided at the end of this page), though a credit to Bill Gray at Project Pluto would be appreciated. I suggest checking back every now and then; we're still finding out things about this object, and I'm updating this page regularly. I can be reached at pôç.ötulpťcéjôřp@otúlm, after you prove you aren't a spambot by removing diacritical marks.
What The... uh... Fun is up with the name? (Probably the Most Frequently Asked Question!) Since I wrote this, Eric Christensen has provided commentary on the naming scheme for this object.
The name is due to sheer luck. The name was assigned automatically, by software lacking humor or propriety. We're just lucky it wasn't WTF1190.
The object was found by the Catalina Sky Survey in Arizona during their usual efforts to find asteroids. When their software spots a new object, it gives that object a unique seven-character ID. The first character gives the year. Objects found from 2000 to 2009 start with '0' to '9'. Those found in 2010 start with 'R', 2011 is 'S', and so on, up to 2015='W'.
The second letter tells you in which month the object was found, and in which half of the month. This follows a standard scheme (scroll down to the table in the "New-Style Provisional Designations" bit). You'll see that 'T' is used for observations made from October 1 to 15.
The remaining five characters are hexadecimal digits. The third character tells you which CSS telescope got the object. After that, it's just a steadily-advancing counter, running from 0000 to FFFF, of how many objects have been found in that half a month.
Previously, the object was observed in February 2013 (when it got the designation UDA34A3) and late November 2013 (UW8551D). Before that, it was observed in 2009 and 2010, as 9U01FF6.
I'd linked the 2013 observations long ago, showing that UW8551D and UDA34A3 were actually the same object. That linkage was not especially hard. Getting the link to figure out that the 2013 data and the 2015 observations were of the same object, and then linking in the 2009 data, was quite a bit more difficult. You can click here for some technical details of how the linkages were made, and click here for some details on the linkage to the 2009 object, 9U01FF6.
What is this object? Still puzzling that one out. I don't think we'll get a really solid identification for it as "it's this bit of hardware from this mission". Though for reasons I'll describe, that still could happen; we have some more recent news that may give us some new clues to work with.
We know it's been orbiting the earth at least since September 2009, and possibly for some years before that. We can say that because we have actual images of it in space from back that far, plus a possible image from June 2009. It could be quite old -- it could even be from the first lunar mission, Lunik 1, launched in 1957 -- but it probably isn't, for reasons described below. In particular, I doubt it's old enough to be Apollo-era hardware.
I had hoped we could keep on digging out still more old observations, gradually working our way backward in time until we could say: "it passed the moon or earth at thus-and-such a time, close enough to when and where a spacecraft was there that we can say it's not a coincidence." The problem is that, at 2009, the trail has gone pretty cold.
After we linked the 2015 observations back to the 2013 observations, Eric Christensen and Alex Gibbs at the Catalina Sky Survey started looking through their old images, hoping to find additional observations that had been overlooked at the time, while Marco Micheli looked through old images from the PanSTARRS survey in Hawaii. (The images were scanned by computers for moving objects back when they were taken, and the software does a pretty good job. But the human eyeball still picks up things the software didn't. Also, the humans have the advantage that we now know roughly where to look on those images.)
CSS came up empty. Marco Micheli, however, got quite a few observations from PanSTARRS, going all the way back to December 2012.
And that was as far as we got with that method. The problem was that the object went past the moon in May 2012. It's a bit like tracking a pinball: you can see where the pinball is going now, work your way backward, and see which bumper it came off. And you might come up with a decent idea as to the direction it was going when it hit that bumper, and take a guess as to the bumper it hit before that. But pretty soon, you're just guessing.
However, I sent Eric, Alex, and Marco an e-mail with the subject line "Somewhat long shot possibility for identifying WT1190F". I pointed out that Peter Birtwhistle, an amateur astronomer in the UK, had suggested that 9U01FF6 and the 2013 objects UW8551D = UDA34A3 might be one and the same. At the time, I'd thought that Peter was quite probably correct, but I had no way of linking the two objects; there was too much time between the observations. But I suggested that if we couldn't get observations extending the 2013 object backward, perhaps we could get some extending 9U01FF6 forward, until the observations would link up.
I hesitated to suggest this. I wasn't too sure these really were the same object, and thought I might be proposing a wild goose chase. However, they dug into the archives again and started finding more observations.
The next day, Eric found two observations made on 2011 January 5, about ten arcminutes off prediction. With that, I computed new ephemerides for PanSTARRS and CSS, and realized we had a bit of a problem; 9U01FF6 took a lunar flyby in February 2011, and the ephemerides became quite inaccurate for dates after that.
Fortunately, Marco kept on looking in PanSTARRS, and got six detections. Three of them were in April 2011, getting us past the lunar flyby. That was key; I was able to get 9U01FF6 with a lunar flyby on 2012 May 24, and WT1190F with a lunar flyby on the same date, with suspiciously similar elements. It still took some tweaking (don't try this with Find_Orb at home; I had to add in some code to do the searching), but we had an orbit linking the 2013 and 2009 observations, proving they were the same object.
So we now know the object has been orbiting the earth since (at least) September 2009. Marco has found a "maybe" detection from June 2009. We may be able to run things back before that, but I'm not especially hopeful. We've gone back to the point where PanSTARRS was first observing, and having some problems, so that source may not do us quite as much good as we would like.
So, in theory, it could be any of many unaccounted-for bits of junk sent into high earth orbit between 1957 and 2009. However, I've two reasons for thinking it wasn't sent up too much before 2012.
First, the asteroid survey guys have found other objects in high-earth orbits, such as the boosters for Chang'e 3 and Chang'e 2; the Chang'e 2 spacecraft itself; and if you take a look through this list of spacecraft found by asteroid surveys, you'll see some other similarly high-flying objects. Most of the time, the objects were found not too long after they went up. Sometimes, they were lost track of (remember, tracking high-orbiting space junk is a priority for nobody), and then found again. But the idea that this object has been circling the earth for decades and only got observed in 2013 is not very likely.
Second, I've run some simulated slight variations of WT1190F backward in time. This sort of orbit isn't entirely stable. An object in it is likely to run into the earth or moon, or get enough energy to escape into orbit around the sun. But I wondered just how long this would take. It turned out that most of the variations I tried took about three to ten years, usually escaping the earth-moon system, with a few hitting the earth.
So this object could have been up there since 1957, ducking and dodging its way around the earth-moon system, never running into either and never getting thrown out. But it would be somewhat of a long shot. A more recent object is quite a bit more likely.
Another possibility is that it's a bit of hardware that had enough energy to escape the earth-moon system, go into orbit around the Sun, and eventually returned to orbit us not long before 2009. So far, we have at least one example of this: J002E3, the Apollo 12 stage that briefly orbited us in 2003. It doesn't seem too likely -- if it came from heliocentric orbit, it had to lose a lot of energy, perhaps through a lunar flyby -- but it could happen, and something like it has happened at least once.
How do we know how big it is? Well... we don't, really. Not all that precisely, anyway.
The object is too small and distant to show up as anything but a point of light in a telescope. We can measure how much light that telescope gets, though, and say that it has to be at least a meter or so across. If it were totally, 100% reflective, it could be that small and still bounce off enough sunlight to be as bright as it is. (Which isn't very. You do need a decent telescope to image it.)
Most likely, it isn't totally reflective. If it reflects, say, 25% of the light that hits it, it could be twice as big: about two meters across. If it reflects only 4% of the light that hits it, it could be five meters across. But it's probably about one or two meters across.
How do we know it's hollow? We don't (I wish the ESA folks hadn't used that word). We know that it's very light for its size, light enough to be space junk rather than a rock. But it may be, for example, a flat panel. (Though the ESA is probably right that it's hollow, for reasons I'll get to.)
This is not the first bit of space junk asteroid hunters have tracked. When one is found, we start solving for its orbit, usually on the basis that it's a rock. With space junk, that assumption can actually work for days or weeks. But then you start to notice that your observations seem a little bit off. The reason: space junk is a lot lighter than a rock. It's so light that even sunlight can exert a very gentle push on it.
After you get enough tracking data on a piece of space junk, you can determine just how light it is. More accurately, you can determine the ratio of its area to its mass: its "area/mass ratio", or "AMR". A high AMR indicates a big object (that catches and reflects a lot of sunlight) with low mass.
With rare exceptions, asteroids have AMRs so low we can't even measure them. Space junk tends to be light. It costs a lot to send things into space, and engineers work hard to send as little as possible and use as lightweight materials as they can get away with. The area/mass ratio tells us this is a "fluffy" sort of object.
I said before that the ESA is probably right about it being hollow. I say that because measurements of the brightness of this object don't show a lot of variation. If it were flat, you'd expect it to get faint when seen edge-on, and bright when facing us. But the "light curve" (graph of brightness vs. time) is nearly flat.
Aren't these objects kept track of? Sort of. Not very much. Mostly, believe it or not, by amateurs.
I think that, after WT1190F, there will be more interest in keeping track of high-orbiting satellites. But they've been generally ignored until now.
Stuff in low orbit around the earth is kept track of very nicely, almost entirely via radar. Objects in higher orbits, such as geosynchronous satellites, are also kept track of, with both radar and telescopes. (The effectiveness of radar drops off very rapidly with distance.) At these altitudes, there are concerns about hitting bits and pieces of junk at several kilometers a second, resulting in the loss of very expensive satellites. Even a loose paint chip can be a danger if you hit it ats low-earth orbit speeds.
Objects in still higher orbits are not kept track of as well, unless they are active satellites. There is no real collision hazard at these altitudes (except with the earth or moon). The asteroid hunters at Catalina, PanSTARRS, and elsewhere sometimes stumble across such objects, briefly think "this might be an interesting asteroid", realize it's actually just a bit of junk, and move on. From their viewpoint, space junk is just a distraction from finding rocks. So very few people have been at all interested in these objects. To give you an idea, I may be the only person computing orbits for such objects on even a semi-regular basis, which I do in my spare time as a hobby. The only reason we had enough data for the 2009 appearance of this object, back when it was known as 9U01FF6, is because Peter Birtwhistle imaged it from his amateur observatory in the UK. (Without Peter's data, we would have had a little less than two days worth of observations in 2009. With it, it's nearly five months.)
I've kept track of a variety of high-flying satellites, both out of curiosity and because it's useful to the asteroid community: when an astronomer finds a possibly interesting asteroid, I can usually give advice as to whether it's a known piece of junk, or the sort of rock an astronomer would want to keep track of. These objects also make excellent test data for my orbit determination software (which allows somewhat motivated people to do this sort of analysis on their own computers). I've also had hopes some junk would hit the moon, giving us a free LCROSS experiment. But again, high-flying junk of this sort has generally been considered to be not worth spending much time on.
What does the orbit look like? Click here, or on the small plot at left, for the full plot of the last three orbits of WT1190F. The small red circle is the earth. The big green circle is the orbit of the moon, just to give some scale to the chart. Right now, the object's orbit is mostly in the same plane as the moon, but be aware that we're doing a little "flattening" of a three-dimensional situation.
Also, T. S. Kelso has provided a nice plot of the "side view", what the orbit would look like as seen from a slight angle to the moon's orbit.
Also, somebody (don't know who, wish I could give credit) put together an animation showing the trajectory of WT1190F as it comes in for re-entry.
The marks along the orbit line indicate one day each of motion. As you'll see, the object marches right along when close to the earth, slows down a bit further out than the moon, then comes back down for another pass. But each time, the orbit is a little bit different: the sun pulls on it a little bit (and sunlight "pushes" it a little bit), and the moon pulls on it a little bit, and so do a lot of smaller forces. Those forces have gradually nudged the "low point" of the orbit (the perigee) over the years, sometimes up, sometimes down. It's had a few consecutive downward pushes, to the point where this time, the lowest point would be 600 kilometers underground, resulting in an impact.
Where will it hit?
Click on either of the maps for a more detailed version. The magenta blob off the south coast of Sri Lanka is the nominally predicted re-entry area.
By "nominally predicted re-entry area", I mean "the area where we'd expect to be we'd expect it to be able to see it". This is somewhat inexact. We have a very exact trajectory for this object now, right up to the point where it hits the atmosphere. In putting the magenta dot on this map, I used the ephemeris for this object in its last minutes and picked the point where it started decelerating strongly. But we don't really know as much as we might about when it will become visible, or how quickly it will disintegrate. (Which is part of the reason this object is of interest: if we already knew exactly what it was going to do, there wouldn't be much point in observing it.)
More recent news
As this object came closer to re-entry, I did some computations to see who would be able to get images of it while still in space. To do so, you (obviously) need to have the object above your horizon, and it has to be night where you are. That narrowed the field a bit. I soon found that Europe would get the last look at WT1190F, and in particular, the UK would be a good place to be. One rarely associates the UK with good astronomical observing conditions, but the object was above the horizon with twilight barely beginning a mere twenty minutes before impact.
As it happens, Peter Birtwhistle in the UK was able to get a lot of images and data about WT1190F in its last hours. The positions measured from the images were quite helpful in refining the estimated impact point (the final trajectory was probably good to a hundred meters or so, with a re-entry time good to about a tenth of a second), but this image really got my attention. It's a 20-second exposure of WT1190F, during which the object moves most of the way across the image. During that time, you see the brightness vary as the object tumbles. There are about 27 "flareups", alternating between "kinda bright" and "a little brighter". I think this means the object is tumbling once every 2*20/27= about 1.5 seconds, and we get a kinda bright flash from one end and a slightly brighter flash from the other end.
Previously, people had attempted to measure the rotation rate of WT1190F. There were two things stopping them. The object was farther away, and therefore fainter. And the object wasn't moving at a good rate, so they would measure the collected light over (say) a five-second exposure. You need to be able to see what's going on at a subsecond rate, which you can do in Peter's image.
So why is knowing the rotation rate such a big deal? Mostly because some space junk is left spinning and some isn't. In particular, Jonathan McDowell and Paul Chodas have speculated that WT1190F might be the Trans Lunar Injection (TLI) stage that took the Lunar Prospector mission from low-earth orbit up to the moon in 1998. It looks as if that TLI stage was left spinning once a second. If it slowed down a bit over the last 17 years (as might happen due to effects from the earth's magnetic field), it might be spinning about once every 1.5 seconds. (Jonathan points out that while this suggests a match is possible, it's not conclusive. There are other bits of space junk up there that also could be a match. So the quest to figure this out goes on.)
Links to images, videos, etc.
Note that everything up to this point is in the public domain (it's all text and images created by me, so I'm free to say that). Use it as you wish, though a credit to Bill Gray at Project Pluto would be appreciated. The following images and video are the property of their respective creators; you'd have to ask them about re-use.