Tuesday 25 September 2012

[UPDATED] The 21 September fireball: a small Aten asteroid?

-- edited/corrected 25/9 15:25 UT. I initially made a small error in the used trajectory azimuth (not properly taking into account effects of a spherical earth). That is corrected, but the conclusions do not alter. --

In my previous post I presented clear evidence that the splendid fireball seen over NW Europe on September 21st, 2012, was a meteoric fireball. I also presented a first, very preliminary idea of its trajectory.

Based on that trajectory, I can now present some very first, very cautious conclusions about the heliocentric orbit of this meteoroid.The solutions strongly favour an identification as an Aten asteroid.

The entry azimuth of the fireball from the reconstructed preliminary trajectory is around 80 95 degrees. Based on observations by Ramon van der Hilst who observed the fireball from Bussloo, the estimated entry angle for the fireball is about 5 degrees only: a very shallow, earthgrazing angle which explains the long trajectory. (I asked Ramon to estimate the angle of the fireball with respect to the horizontal at the moment Ramon was looking roughly perpendicular to the preliminary trajectory. That angle, about 5 degrees as Ramon reports, should be close to the entry angle)

I used these values and an 18-20 km speed estimate to compute a nominal heliocentric orbit: and then played around by widely varying the values for speed, entry angle, entry azimuth around these nominal values.

The interesting point is, that for all of these, I get an Aten orbit as a result. Aten asteroids are asteroids whose perihelion lies within the orbit of the earth and who's aphelion lies only just outside the orbit of the earth. They have a semi-major axis < 1.0 AU and aphelion (just) over 1 AU.

The aphelion values I get for the approximate fireball orbit, are in the range 1.0 - 1.15 1.05 AU, the semi-major axis values are in the range 0.9 to 0.6 AU. Solutions based on higher speeds (I varied between 12 km/s and 30 km/s in my calculations) favour the slightly larger aphelion values and shorter semi-major axis.

A wide variation in entry azimuth (I tried between 60 and 110 120 degrees) and entry angle (I tried for values between 5 and 45 degrees, the latter clearly a too large value by the way) does not alter this picture much: they all result in Aten orbits.

I need to alter the trajectory direction to values significantly larger than entry from a direction of  120 degrees (well past due east) to get aphelion values that start to get well beyond 1.15 AU and semi-major axis values > 1.0 AU.

For the current very preliminary nominal trajectory solution (entry azimuth ~82 ~95 degrees, entry angle ~5 degrees) I get these values when varying the assumed entry speed of the fireball:

[editted table 15:25 UT to reflect new calculations/correction of error]

Vini    q    Q     a     e     i

12.0   0.82  1.00  0.91  0.10  6.5
15.0   0.46  1.02  0.74  0.39  15.0
18.0   0.31  1.04  0.67  0.55  20.7
20.0   0.24  1.05  0.65  0.62  24.8
25.0   0.16  1.09  0.62  0.76  37.4
27.0   0.13  1.11  0.62  0.79  43.7
30.0   0.11  1.14  0.62  0.83  54.5

Vini is the initial speed (in km/s), q the perihelion distance (in AU), Q the aphelion distance (in AU), a the semi-major axis (in AU), e the eccentricity, i the inclination.

These values should be taken with caution and only as rough indications. There are (still) large uncertainties in the trajectory and entry angle, as well as the speed of the fireball. They do show however (as well as variations on the trajectory not listed here) that an Aten-orbit is the implied solution.

The Earth encountered the meteoroid close to the meteoroid's aphelion, when it was moving almost in parallel with the Earth.

NOTE / UPDATE 26/09/2012, 19:25 UT: There is some confusion on the web regarding my analysis and the "retrograde"/ "prograde" character of this object.
The "retrograde"character is only true for an earth-centered orbit (i.e., an object orbiting the earth, such as an artificial satellite). An east-west movement in that case means it is "retrograde" (against the motion of the earth's rotation).
This is not necessarily the case for a sun-centered orbit however. An east-west moving object then can be (and is, in this case!) in a normal, "prograde" orbit (=moving in the same direction around the sun as the planets). The difference is the frame of reference: earth-centric versus sun-centric.
So beware: the "retrograde" orbit refers to what the orbit would be for an earth-orbiting satellite (which this object was not). The Aten heliocentric orbit presented here, is however prograde.

Monday 24 September 2012

More on the 21 September 2012 fireball: why it definitely was a meteor

I should have done this analysis earlier but did not have the time available until now. What follows now is a quick and back-of-the-envelope kind of calculation, but in my (not so) humble opinion it is adequate to the question at hand.

It concerns, of course, the splendid slow fireball seen widely over NW Europe near 21:55 UT on 21 September 2012. I posted on it before, focussing on saying "no" to the suggestion that this could have concerned a satellite reentry. In the post that now follows, I further strengthen the conclusion that it was not a satellite reentry, but a genuine meteoric fireball.

The map above gives a quick (and not particularly accurate) back-of-the-envelope reconstruction of the fireball trajectory. It is based on trajectory descriptions from Bussloo in the Netherlands and Dublin in Ireland: by taking reported altitudes (with respect to stars) and general directions of reported start and endpoints, and an assumed altitude of 50 km, the trajectory above is what approximately results. (update 19:10 UT, 24 Sep: an updated version of the map is at the bottom of this post).

The resulting trajectory is some 1000-1200 km long. In what now follows, I have taken 1100 km as the distance travelled by this fireball.

Observers near the western and eastern ends of the trajectory would probably not see the complete trajectory. Observers approximately mid-way, in mid-Britain, would potentially see most if not all of the trajectory (from experience I know you can see bright fireballs from distances of 500 km).

Observers report durations between 20-60 seconds: most video's on the web suggest a 40+ seconds duration.

It would take a reentering satellite travelling at 8 km/s (the orbital speed at decay altitudes) about 138 seconds or roughly 2.25 minutes to travel this distance. While the reported fireball durations are long, none of the reports nor videos comes even remotely close to that value.

A meteoric fireball travelling at the lowest speed possible for such an object, 11.8 km/s, would take 93 seconds to travel that distance. This is still longer than almost all of the reports suggest, but clearly getting closer.

If we take an estimated duration of 60 seconds, the 1100 km trajectory length results in a speed of  approximately 18 km/s.

18 km/s is a very reasonable speed for a slow, asteroidal origin fireball.

(it is, let me repeat, also way too fast for a satellite reentry).

Meteorite dropping fireballs typically have speeds between 11.8 and 27 km/s. A speed near 18 km/s sits squarely in the middle of that speed interval.

(update: diagram added 14:45 UT, 24 Sep)
(click diagram to enlarge)

The 60 seconds probably represents the upper boundary value for the duration of the fireball. If we take a shorter duration of 40 seconds, the speed already increases to 27.5 km/s.

This quick back-of-the-envelope reconstruction therefore shows that this must have been a meteoric fireball, quite likely of asteroidal origin, and we definitely can exclude a satellite reentry.

The fragmentation described and filmed is not unusual for meteorite dropping fireballs (see the video's of the Peekskill meteorite fall in my previous post). The object probably entered the atmosphere under a very shallow angle, which together with the slow speed explains the unusually long duration of the event.

Meteors of this kind are rare, but they have been seen before. Think of the Peekskill meteorite fall, but also the famous 1972 daylight fireball over the Grand Tetons (that had a duration of over 100 seconds) and the Cyrilid Meteor Procession from 1913 (that lasted minutes).

Note: a previous post gives a number of other lines of evidence which likewise suggest this fireball was not man-made space debris.

UPDATE: a further update is given in a new post: a very cautious orbital solution suggests an Aten orbit.

Note 2: on how I made this quick and (emphasis) rough trajectory reconstruction. I took observations that contain clear sky locations: e.g. a sighting from Dublin stating it went "through the pan of the Big Dipper"; the description from Bussloo observatory in the Netherlands; and later adding a.o. a photo from Halifax, UK, showing it just above the tail of Ursa Major. These descriptions can be turned into directions and elevations. Next, I drew lines from these sighting points towards the indicated directions, marking distances roughly corresponding to 30, 50 and 80 km altitude as indicated by the observed elevation [ distance = altitude / tan(elevation) ]. Near the start of the trajectory I marked 50 and 80 km, for Britain and Ireland I marked 30 and 50 km. These points then provide you with a rough trajectory.
From Dublin the object passed through North towards west. From Bussloo the object started NE (azimuth 60 degrees): these are important points of information too as it shows that the object started at least as far east as the Dutch-German border (and more likely over Sleswig-Holstein in N-Germany) and had its endpoint at least as far west as the northern part of Ireland.

Above: Updated map version, 24 Sep 19:10 GMT , also showing the principle of how it was reconstructed for three sighting locations. With thanks to Ramon van der Hilst for providing more detailed information on sky trajectory as seen from Bussloo (NL) on request.

Imaging a near-zenith pass of the new NOSS 3-6 pair (NROL-36)

As I wrote earlier, the NRO launched a new pair of NOSS satellites, NOSS 3-6,  on September 13th, as NROL-36. Two days after the launch, I got my first look at these objects, albeit under very poor conditions (see my post here).

Sunday morning, 10 days after the launch, the cloud cover broke and I finally got a renewed and much better view of the two NOSS objects and the Centaur r/b from the launch, during a near-zenith pass. Some very thin cirrus clouds (dispersed aircraft contrails in part) were in the sky. These eerie pictures, taken shortly after 5 am local time on the 23rd, are the result (click them to appreciate them in full glory):

(click images to enlarge)
the NOSS 3-6 duo

the NOSS 3-6 Centaur r/b

The top image shows the payloads, 2012-048A and 2012-048P, traversing Perseus (alpha Persei star association in top). The P-object is leading over the A-object: movement is from lower right to upper left. I could see both payloads naked-eye, at about mag. +4.

Currently, the two satellites are still notably further apart than operational NOSS-es are, as they are still in the process of active manoeuvering. Ted Molczan believes that eventually, the A-object will probably overtake the P-object and become the leading object once the final operational configuration is reached.

The second image shows the Centaur r/b (2012-048N) traversing the Cassiopeia-Perseus border (h and chi Persei in top, stars of Cassiopeia near the bottom). It was very bright, initially +1 just after shadow exit, then +2. I could see no clear periodic brightness variation: the slow tumbling that was apparent in the days right after lauch and which might have been due to remnant fuel outgassing according to Ted Molczan, apparently has subdued.

The USA 237 r/b

On the 16th, I imaged the geostationary satellite USA 237, which is perhaps a 6th Mentor (see the bottom part of my previous post here).

On the 19th, I used the 37-cm Rigel Cassegrain of Winer Observatory (MPC 857) in Arizona to image the USA 237 r/b of this launch (2012-034B) :

(click image to enlarge)

In addition to the USA 237 r/b, I also did one of my periodic observations on Prowler (90-097E) that same night using the same telescope.

Saturday 22 September 2012

Fireball over N-Europe on 21 September 2012, 21:45 GMT was likely NOT a reentry

UPDATE (24/9/2012): more and definite arguments that this was not a reentering satellite, can be read here in my follow-up post from Sep 24th. This includes a first rough trajectory reconstruction for this fireball.

Reports are pouring in from The Netherlands, Britain, Ireland and other N-European countries about a very bright, extremely slow fragmenting fireball appearing around 21:45 - 21:55 GMT (23:45 -23:55 CEST) on the evening of 21 September 2012.

Various video's have been posted on Youtube, notably by observers from Britain (large parts of the Netherlands were clouded out, including the all-sky stations):

Because of the unusually long duration and slow movement, some people have suggested the possibility of a satellite reentry. For various reasons, this is however very unlikely.

Multiple reports make clear the object was moving from east to west. A report of observers from Bussloo Observatory, the Netherlands, for examples states that the fireball appeared in the north, moving from Perseus  to Bootes, almost horizontally from east to west. Similar reports (e.g. here and here) come from Ireland.

Almost all non-polar satellites move prograde,  from west to east (or north-south and v.v. for a polar orbit). An east to west movement would necessitate the object to have a retrograde orbit (meaning that it moves counter to the earth's direction of rotation). Such objects are extremely rare: they literally amount to only a handful of objects (including the US FIA Radar satellites, and the Israeli Ofeq/Shavit satellites/rb). For this reason, it is extremely unlikely that this fireball was a reentering satellite.

Update 24 Sep: in the comments to this blog post, the issue was raised of the potential reentry of a classified object. However, the larger classified pieces are tracked by us amateurs. We have no likely decay candidates among the retrograde objects that we track. We can account for and hence exclude the FIA's for example (the rocket bodies of that launch were deliberately de-orbitted right after launch so are no candidates either). The Israeli Ofeq/Shavit are no candidates as their orbital inclinations never take them over the Netherlands and the British Isles. And there are simply no other suitable retrograde objects -- end of update.

There are moreover no unclassified reentry candidates for this date listed by USSTRATCOM on their space-track portal. Given the brightness of the fireball, this should have been a seizable chunk of space debris, that really would have been tracked (and predicted). Again, this makes it very unlikely that this fireball was a satellite reentry.

While the duration of the fireball is unusual, it is not unprecedented. In many ways, the descriptions and video are reminiscent of the Peekskill fireball that dropped meteorites near Peekskill in 1992:

(below: two video's of Peekskill fireball, 1992)

It is therefore my opinion that the 21 September fireball was most likely of meteoric origin: a chunk of asteroid. Alas, any surviving remains appear to have splashed down in sea (update: or possibly Scotland - N. Ireland).

The duration of the event, though not unprecedented, is certainly unusual and for this reason, I am saying "most likely not" rather than "certainly not".

UPDATE (12:45 GMT, 22 Sep):  another bright fireball was widely seen from the US and Canada that same night near 20:30 GMT. There was at least one hour inbetween the two events, so they do not appear to be related (i.e. they do not concern the same fireball).

UPDATE 2 (13:30 GMT, 22 Sep): Suggestions that the fireball might be related to Chinese CZ-4 space debris, catalogue #26213, are plainly incorrect. That object (and any fragments of it) are in a 98 degree polar orbit. This is completely incompatible with the reported movement of the fireball. As seen from Bussloo in the Netherlands and Dublin in Ireland, the fireball moved perpendicular, not parallel, to the orbital plane of this Chinese space debris (and that of any related fragments).

 IMPORTANT UPDATE 3 (24/9/2012): more and definite arguments that this was not a reentering satellite, can be read here.

Thursday 20 September 2012

Europe's new weather satellite MetOp-B

On September 17, a Soyuz rocket launched from Baikonur brought Europe's newest weather satellite in an 805 x 810 km polar orbit: MetOp-B (2012-049A). She joins her sister-ship MetOp-A launched in 2006. In total, three of these polar-orbiting satellites are planned.

(click image to enlarge)

Yesterday evening, two days after launch, it was briefly clear in Leiden. With scattered clouds in the sky (see the streaks in the upper right corner of the image) I shot the above image of Metop-B passing through Lyra (brightest star is Vega).

A very faint second trail can be seen near the right end of the brighter MetOp-B trail, going obliquely upwards. This is an old Russian r/b (1982-096B) from the 1982 launch of Kosmos 1410.

MetOp-B's older sister ship, MetOp-A, is known among amateurs for producing bright flares (up to mag. -2) when the solar panels reflect sunlight to the observer (see my video here). Hopefully MetOp-B will show this spectacular behaviour as well.

In addition to MetOp-B, the short observation window amidst the clouds also permitted to observe FIA Radar 2 (2012-014A).

Monday 17 September 2012

Tracking the new NOSS 3-6 (NROL-36) launched September 13th, and imaging USA 237 (NROL-15) and SDS 2-2

NROL-36 (NOSS 3-6)

At 21:39 UTC on September 13th, 2012, the NRO launched NROL-36 from Vandenberg AFB. The launch vehicle was an Atlas V containing a classified NRO payload in addition to a number of cubesats. Several analysts already suspected the classified payload of this launch to concern a new duo of US Navy NOSS satellites.

That suspicion appears to be born out by the first observations of the launched objects, which show two satellites in close formation, typical of a NOSS formation.

The Centaur last stage was the first object from this launch to be observed, by Björn Gimle in Sweden, who observed it 5 hours after launch, and then by BWGS chair Bram Dorreman in Belgium 1.5 hours later, who remarked it was variable in brightness. Alas I was clouded out in Leiden at that moment. Eleven hours after launch, Kevin Fetter in Canada observed it next. Over the following day Russell Eberst in Scotland and Scott Tilley in Canada added more observations.

The payloads were first observed by Kevin Fetter in Canada near 9:43 UTC (Sep 14), 12 hours after launch. Scott Tilley, also in Canada, next observed them on the 15th near 12:34 UT.

On the morning of the 16th, 2 days and 5 hours after the launch, it was clear in Leiden allowing me to join the chase. I (photographically) observed the Centaur r/b pass at 2:49 UTC (see image below) and then the payloads at 2:55 UTC (Sep 16).

The NROL-36 Centaur booster
(click image to enlarge)

The observations were done under difficult conditions: it was somewhat hazy, the passes were at very low altitude due east (only 14 degrees elevation for the Centaur and 21 degrees elevation for the payloads), and the phase angles for the objects were unfavourable, resulting in dim magnitudes (around +7 for the payloads). The resulting images (one of them below) are therefore not quite pretty: even with use of the 1.4/85mm SamYang lens the trails were so faint that they were marginally visible at best. As reported earlier by Scott Tilley, the leading object indeed seemed to be somewhat variable (on some images it wasn't visible at all).

The NROL-36 payloads (NOSS 3-6)
(click image to enlarge)

Over the next days, I'll hopefully be able to obtain much better imagery during more favourable passes (this morning I was alas clouded out again, however....)

USA 237 (Mentor 6?)

That same morning, following the NROL-36 related observations, I took a few images of the area near the bright geostationary Mentor 4 (USA 202, 2009-001A). These images not only recorded this 3-year-old Mentor, but also the geostationary satellite USA 237 (2012-034A), a classified NRO payload launched from Cape Canaveral as NROL-15 on 29 June 2012, employing an Atlas IV-Heavy as launch vehicle.

These were my first observations of this object: tracking of this object over the summer was done mostly by Greg Roberts in South Africa (who was the first to discover the object) and Peter Wakelin in Britain. Over the summer it had slowly been drifting westwards: as of mid-September 2012 it appears to have stabilized at 47.8 E. For me in Leiden it is currently located at 20 degrees elevation in azimuth 129 degrees (low SE), about 6 degrees separated from Mentor 4 in the sky.

USA 237 (NROL-15)
(click image to enlarge

The identity of this object is still provisional. There have been some thoughts that NROL-15 launched a stealth Misty satellite in LEO/MEO and then progressed to put a decoy (the object designated as USA 237) in geostationary orbit. USA 237 is very bright however (about mag. +8), ranking it among the brightest geostationary objects in the sky, the Mentor SIGINTs. As Ted Molczan showed, they seem very similar in terms of absolute brightness. The likeliness can be well seen in the photographic comparison below, which shows Mentor 4 and USA 237 (the crops come from one and the same image):

(click image to enlarge)

This gives reason to think that NROL-15/USA 237 is perhaps a sixth Mentor SIGINT (i.e., it is Mentor 6).

SDS 2-2

As by-catch of the NROL-36 payload and Centaur observations, I also obtained my first observations of the geostationary SDS satellite SDS 2-2 (USA 67, 1990-097B). It is located at an elevation of only 16.5 degrees, very low in the east for me. To my surprise, it was rather bright and easily visible in the 5-second exposures, of which the one below is one:

 (click image to enlarge)