RASSVET (2nd batch): Reports of my Death have been Highly Exaggerated

 

Russia is in the process of building a constellation of satellites in Low Earth Orbit for data connectivity, called RASSVET (Russian for 'dawn' or 'sunrise'): a kind of Russian Starlink. Understandibly, given the military use of Starlink-like constellations, this buildup is under scrutiny by western analysts. Some of which appear to be jumping the gun, as I will argue below.

Two batches of operational RASSVET have been launched so far: batch 1 of 16 objects on 23 March 2026; and batch 2 of 16 objects on 19 July 2026. They are designated with the RASSVET 3 name, to set them apart from the prototype tests RASSVET 1 (likely 3 satellites, part of a launch with many objects in 2023, launch 2023-091) and RASSVET 2 (also likely 3 satellites, probably 2024-092G, H and J which reached 777 km altitude).

On 2026 August 31, the Institute for the Study of War advances (ISW) published an online report claiming that the second batch of RASSVET's was 'failing' to reach orbital altitude. 

Late August, the objects from this second batch were at altitudes of 300-400 km rather than the 510 km reached by the first batch.

In a wider claim, they also point out that the orbit raise of the first batch stopped at 550 km (actually: 510 km) rather than the 870 km published for the constellation in a TASS newsbulletin from 8 years ago. They take this to indicate a failure to reach the intended operational altitude.

This claim of 'failure' was quickly picked up by multiple news media, e.g. Ars Technica.

So, is RASSVET, and specifically the second batch failing? When we look at the actual data, I'd say that, to paraphrase Mark Twain, reports of RASSVET's death have been highly exaggerated.

Below diagrams show the orbital evolution of RASSVET 3 batches 1 and 2 so far. Each colour represents a satellite in the batch, each dot is an orbit determination (plotted against time). The x-axis shows Mission Elapsed Time, the time, in days, since launch.

 

Click diagram to enlarge

Click diagram to enlarge

We can see that the RASSVET's from the first batch seem to have settled at 510 km orbital altitude. Three objects from that batch did not (yet) reach that altitude. 

One, RASSVET 3-4 (2026-061D) appears to have failed and reentered into the atmosphere on 6 June 2026. The other two might still be in the process of orbital raising. It can be seen that for batch 1, the orbital raising was at different moments for different subgroups in this batch.

At first glance, these diagrams also seem to corroborate ISW's claims about the second batch: so far they did not reach more than 400 km in orbital altitude. ISW claims there is an "apparent failure of all 16 satellites from the second launch to reach their intended operational altitudes". ISW does seem to take the 510 km orbital altitude of batch one as some sort op intended altitude.

But the assessment of batch two compared to batch one in the ISW report is seriously flawed and not borne out by a closer look at the data. There is a reason I chose to express the X-axis in the diagrams as "Mission Elapsed Time": this aids direct comparison of the timeline of both launch batches.

Below is a combination of the diagrams for both batches, in which I have cut off the diagram's x-axis at 80 days. This makes both diagrams more directly comparable. The diagram for the second batch ends with the last available orbits from yesterday, 53 days after launch of this batch.

Click diagrams to enlarge

 

It can be immediately seen that the orbital altitudes reached during the 53 days since launch are very similar for both batches: in both cases, the orbital altitudes reached after 50 days is between 300 and 400 km. If anything, batch 2 actually appears to raise their orbits a bit faster (!) than batch 1 did.

We can't even say that the lowest three objects from batch two, RASSVET 17, 23 and 26 currently at around 300 km altitude, necessarily 'failed'. In the first batch, some objects initially stayed at that altitude too during the initial 50 days, but were subsequently raised to higher altitudes.

I therefore see no indication that the RASSVET-3 batch 2 is failing to raise their orbits in comparison to batch 1, contra the ISW claim. 

Back to the initial batch, RASSVET 3-1.Of the 16 objects launched, one clearly failed, and reentered. While unfortunate, we should note that with Starlink, frequently one or a few satellites of each launch batch also fail. This is just a given: on each n objects launched, some will fail.

Two RASSVET's from batch 1 might (or might not) still be in the process of orbit raising, and 13 of the initial 16 have settled at 510 km altitude. This is indeed not the 870 km orbital altitude TASS mentioned in 2018 (eight years ago). But is that relevant? It is very well possible that the initial goal for the orbital altitude was later adjused to a lower altitude (which means better latency) for operational reasons, just as Starlink did. Starlink's initial FCC filing from 2016 was for an orbital altitude of 1150 km, while the first operational launches targetted 550 km instead. They are now lowering their operational altitude to 480 km.

I therefore see no clear reason to conclude that the 510 km altitude where the RASSVET-3 batch 1 satellites settled, points to a 'failure' to reach operational altitude. As an operational altitude, 510 km is fine, it is somewhat lower but still in the same ballpark as the initial orbital altitude of Starlink. Assuming a typical lifespan of 5 years (similar to Starlink) and periodic orbit maintenance (like Starlink), it should be no problem. It should again be noted that Starlink is lowering the orbital altitude of its constellation, initially 550 km, to 480 km now, which is lower than the batch 1 RASSVET's.

So, in my view all this talk of RASSVET "failing" has no real truth to it and jumps the gun. It rather seems wishful thinking.

The reentry of Starlink-1541 over the northern Netherlands on 24 July 2026

click map to enlarge

On 24 July 2026 near 20:26 UTC (22:26 CEST), a defunct Starlink satellite, Starlink-1541, reentered over the North Sea, northern Netherlands and Germany. 

I observed the reentry myself (along with many, many other people). I was on a sailing holliday onboard the historic sailing vessel 'De Onderneming' at that time. We had sailed from Medemblik to Oudeschild on the island of Texel that day, and were sitting on deck in the harbour of Oudeschild (53.03898 N, 4.85058 E) during twilight, when Ide, the deck-hand who was facing west, suddenly said "What is that?!".

As I turned my head in the direction she was pointing into, I saw a bright fast moving object rising from the west in the twilight sky, leaving a trail in its wake. It initially was "blinking" in a semi-regular fashion, which put me on the wrong foot for a few seconds, thinking: "aircraft?". But then, as it rose higher in the sky and passed through the zenith, fragmentation became obvious, and I realised it was a space debris reentry. It continued fragmenting while moving to the eastern horizon. It left a vapour trail that lingered for more than 10 minutes.

I did not have my phone directly at hand, as it was charging in the cabin inside our ship. So I do not have imagery of the reentry itself, but did shoot this image of the lingering vapour trail:

lingering vapour trail. Click image to enlarge

This lingering vapour trail consists of amongst others fine aluminium and carbon particles

A nice collection of photographs and video of the event was gathered by the IMO here, and in this NOS news item (which quotes me). The video below of the event is from the meteor camera of Erwin Harkink (NL000C) in Elst, the Netherlands, where it was lower in the sky:

Of course, my next question was: "what object was this?". As I was on holliday, I did not have my laptop with software with me, only my phone. Logging into Space-Track, a TIP message (22:51 UTC ± 2 hr) with a difference of (at that time) about two hours for Starlink-1541 (2020-055AB) seemed to indicate that satellite as a potential candidate, but with only my phone at hand I had no means to check whether the time of the event matched a pass over the Netherlands. Luckily, Joseph Remis on twitter quickly made the connection and established Starlink-1541 as the identity of this reentering object. 

Several hours later, the final post-reentry TIP issued by CSpOC appeared and was a clear match: 24 July 2026 20:26 ± 1m UTC near 52.7 N, 9.6 E (as mentioned before on this blog, Jonathan McDowell and I believe that these very accurate post-reentry issued ±1m TIP's are based on US DoD satellite observations of the reentry fireball). As Dutch News reports at that time where still identifying the event as a 'meteor', I had a post-midnight whatsapp interview with a journalist of the National Broadcaster NOS about the real character of the fireball in order to set the record straight.

After returning home a week later, I used the latest orbital data available from CSpOC to run a reentry model on Starlink-1541 using our Delft University of Technology's Astrodynamics Toolkit (Tudat), in order to try to reconstruct the likely reentry trajectory. I varied the drag area in the model untill it resulted in typical reentry altitudes (< 90 km) over the northern Netherlands around 20:26 UTC. This is the resulting approximate trajectory (meaning: there is some leeway in exact altitudes and times, but in general this trajectory should be close to reality); the first map gives approximate times, the second approximate atmospheric altitudes:

trajectory and approximate positions (click map to enlarge)

trajectory and approximate atmospheric altitudes (click map to enlarge)

As can be seen, at the harbour of Oudeschild on Texel (one of Dutch Wadden islands) we were almost directly underneath the path of the reentry. Indeed, we saw it pass almost through the zenith, in agreement with this sky-track for Oudeschild harbour resulting from the Tudat reentry reconstruction (note that it was in twilight so only the brightest stars were visible in reality):

reconstructed sky trajectory for Oudeschild (click map to enlarge)

Starlink-1541 (2020-055AB) was launched from Cape Canaveral on 7 August 2020 on a Falcon 9, with 60 other Starlinks and two additional satellites. On 8 February 2026, after 5 years of service, it lowered its orbit in preparation for deorbit. On 24 July 2026, it met its end over the Dutch Wadden Sea.

Starlink-1541 orbital history (click diagram to enlarge)

This was my first ever reentry observation (I tried several times when pre-reentry forecasts suggested an object might reenter over the Netherlands, but in vain): something that was high on my wishlist and can now be struck off. 

I was lucky to see it. Initially, I was to go on holliday on the Scottish west coast, but the sailing vessel in question got a damaged motor, so that holliday was cancelled. I then went on a shorter alternative sailing holliday on the Dutch Wadden Sea with 'De Onderneming'. This brought me right under the reentry path! We were also lucky to be outside on deck when it happened. If only I would have had my phone in my pocket rather than charging inside the ship....

Another failed deorbit of a Falcon 9 upper stage

Falcon 9 upper stage. Image: SpaceX

On 2 February 2026, SpaceX launched Starlink 17-32, with 25 new Starlink satellites, from Vandenberg SFB in California. The satellites successfully reached Low Earth Orbit, but something went wrong with the upper stage, and not for the first time. SpaceX stated on their website:

During today’s launch, the second stage experienced an off-nominal condition during preparation for the deorbit burn. The vehicle then performed as designed to successfully passivate the stage. The first two MVac burns were nominal and safely deployed all 25 Starlink satellites to their intended orbit. Teams are reviewing data to determine root cause and corrective actions before returning to flight.

This is not the first time a controlled Falcon 9 upper stage deorbit failed: this in fact was the fourth time in barely 1.5 years time. 

In July 2024, during the launch of Starlink 9-3, the upper stage failed to re-light, the engine destructing instead, leaving the payloads in a too low orbit. The upper stage later made an uncontrolled deorbit.

In September 2024, the upper staged used for the Crew Dragon 9 launch failed to do a proper deorbit burn, with the stage missing the designated deorbit area. 

In February 2025, after the launch of Starlink 11-4, the upper stage failed to do a deorbit burn. The resulting uncontrolled reentry 18 days later caused a spectacle in the NW European sky, with parts surviving and impacting in Poland (see this earlier blogpost).

And now something similar happened again with the latest, 2 Feb 2026 launch. The upper stage (2026-022AB, catnr. 67675) failed its controlled deorbit. Initially left in a 110 x 241 km orbit, it had an uncontrolled reentry twelve hours after the launch. 

Where it deorbitted is not exactly known. The US Space Force/CSpOC published a TIP for 3 February 2:29 ± 1h UTC. Our Tudat model suggests reentry one orbital revolution later, at 3:47 ± 1.5 h UTC. The maps below give the trajectory over the uncertainty window of respectively the CSpOC TIP, and our Tudat model. The uncertainties overlap each other.

trajectory over CsPOC TIP uncertainty window. Click to enlarge

 

trajectory over the Tudat uncertainty window. Click to enlarge

 

The controlled deorbit aimed for should have happened around 17:35 UTC, just after completion of the first orbital revolution (see map below of what the original plan was). Instead, the upper stage continued to orbit for some 7 more revolutions (or 10 extra hours).

original deorbit plan. Click to enlarge

The reentry of a Falcon 9 upper stage over NW Europe on 19 February [UPDATED]

In the early morning of February 19, 2025, near 3:45 UTC (4:45 CET) scores of people in the Netherlands, UK, Germany and Denmark saw a slow, fragmenting object move through the sky. The event garnered much attention. 

Above is a video of the event obtained by the camera of @dutchspace.bsky.social in the Netherlands.  Another stunning video, by the meteor camera of André Knoeffel in Brandenburg, Germany, is here. The video below is by Nick James from the UK, and shows the start of the reentry near 03:43 UTC and start of fragmentation.

The event was caused by the reentry of a SpaceX Falcon 9 upper stage, object 2025-022Y (cat nr 62878) . This upper stage, from the Starlink group 11-4 launch from Vandenberg, California, on 1 February 2025 23:02 UTC, 18 days prior to the reentry event, failed to deorbit in a controlled way after launch and payload insertion. It was supposed to do a deorbit burn after releasing the payloads, and deorbit over a specified area in the Pacific Ocean around 00:45 UTC on Feb 2, near the end of the first revolution (see map below). With the deorbit failed, it however stayed on orbit for almost 3 weeks, and then had an uncontrolled deorbit over NW Europe on February 19.

 

click map to enlarge

The general character of the Feb 19 event (slow, fragmenting) as well as time and location fit well with the Falcon 9 upper stage reentry. The CSpOC final TIP for this reentry (which we suspect is based on a satellite observation of the reentry fireball) is 19 February 3:43 UTC ± 1 min, near 52.8 N, 5.6 W. The reentry trajectory runs over the British Isles, the northern Netherlands, northern Germany and Poland.

Below is the ground-track for the last orbit up to reentry, and a detail of the approximate final reentry track over Europe, with the Netherlands indicated in red:

click map to enlarge

 

click map to enlarge

A possible piece of debris has been found at Komorniki near Poznan in Poland, media reports. The object on the photograph looks like a Composite Overwrap Pressure Vessel, a component of the Falcon 9 upper stage that has survived reentry before, e.g. in March 2021 when a Falcon 9 upper stage reentered over Washington. The location of the find matches well with the reentry trajectory, as can be seen in the map above. [update: reportedly another one of these composite-wrapped pressure tanks was subsequently found in Poland]

The event caused a lot of media attention: I did several media interviews on the reentry, amongst others with two Dutch national tv channels.

 

This week: rainshowers with a slight chance of falling satellites

The spectacular image above is a frame from video footage obtained by cameras at Añasco, Puerto Rico, part of a cameranetwork from the Sociedad de Astronomia del Caribe (SAC). It was shot on 7 February 2022 near 6:40 UT (2:40 local time) and shows what clearly is a satellite reentry, the reentering satellite spectacularly breaking up into many fragments.

Below is the actual video. It shows two objects appearing about 1 minute apart, both reentering and fragmenting. Especially the second object is spectacular. The two objects could belong to one object that has broken up earlier; or be two separate objects close together in the same orbital plane.

 

The reentering object(s) can be identified as belonging to a batch of 45 49 Starlink satellites launched on 3 February 2022, two-and a half days before the reentry sighting from Puerto Rico.

A day after the launch, when most of the deployed satellites yet had to raise their orbits, a Coronal Mass Ejection (CME) from the Sun arrived at earth, creating a geomagnetic storm. 

During a geomagnetic storm, the upper atmosphere warms up and expands somewhat, causing an increase in drag in low orbital altitudes. SpaceX therefore put their just launched satellites, still in a very low orbit, into safe mode and a least-drag attitude. 

This however prevented them from raising their orbits, and it did so long enough to make it impossible for some 40 of them to be rescued. Over the past days a number of them (at least three, possibly more, including the object over Puerto Rico, at time of writing (9 February 16:00 UT)) have already reentered and the other objects, some 40 in total according to SpaceX, will reenter over the coming week.

Fourty (40!) satellites reentering in only a week or so, is unique. The coming week, the chances of seeing a satellite reentry are therefore larger than usual for anyone between 53 N and 53 S. So keep an eye on the sky!

screenshot of Feb 8 announcement on the SpaceX website

Though it remains to be seen which of the 40 satellites it actually was, the Feb 7, 6:40 UT, Puerto Rico sighting can be possitively linked to this deluge of decaying Starlink satellites. 

One clue is that the orbital plane of this launch was over Puerto Rico near the time of the event, and the direction of movement (SW-NE) matches it.

 

To get even more certainty, I did some astrometry on the footage and fitted a rough circular orbit to the measured positions.The rough orbital fit I get - I measured three fragments-  yield orbital inclinations in the range of 54-56 degrees: Starlink satellites are in 53.2 degree inclined orbits, so this is close enough (given the error margin) to conclude that the reentering object fits with the Starlink orbital plane. The RAAN values also match to a degree or so. So there is very little doubt that this was a Starlink satellite reentering.

One reason why I checked this, is that someone suggested another candidate, a Falcon 9 rocket stage from a 2017 launch (2017-014B), which was also expected to reenter around this date (in fact it had already reentered a day earlier) and had its orbital plane passing over Puerto Rico at the time of the event. This rocket stage however had an orbital inclination of 22 degrees, which is clearly much lower than what I get for the reentering object in the footage.

Starlink satellites are not very big and do not have big rocket engines, so there is very little chance that anything remains and reaches Earth surface from these reentries: it will all burn up in the atmosphere. 

Note: I thank Eddie Irizarry for alerting me to the Puerto Rico event

 

UPDATE 17:30 UT (9 Feb 2022):
For some 21 of the 45 49 Starlink satellites in question, orbital elements have now been released by CSpOC. A quick assessment with SatEvo suggests reentries happening over the coming week, up to mid-February.

A weird Navigational Warning for a mass deorbit on August 9-10? [updated]

click map to enlarge

 

A weird Navigational Warning (NAVAREA XII 384/21) for "Space Debris" has appeared defining nine areas, some of them overlapping, in the Pacific for August 9, 16:27 to 17:29 UT and August 10, 17:16 to 18:17 UT.

I have mapped them in the map above. Below is the text of the Navigational Warning:

060929Z AUG 21
NAVAREA XII 384/21(GEN).
EASTERN NORTH PACIFIC.
1. HAZARDOUS OPERATIONS, SPACE DEBRIS
   091627Z TO 091729Z AUG, ALTERNATE
   101716Z TO 101817Z AUG
   IN AREAS BOUND BY:
   A. 22-52-40N 137-34-57W, 20-12-47N 134-02-08W,
      04-25-05N 146-28-48W, 06-54-48N 149-55-52W.
   B. 51-11-05N 141-36-54W, 49-40-18N 142-13-53W,
      50-44-15N 170-19-30W, 52-17-11N 170-39-50W.
   C. 12-58-15N 130-00-21W, 10-52-28N 127-06-04W,
      05-17-31S 138-47-34W, 03-13-54S 141-40-25W.
   D. 48-12-47N 135-38-42W, 46-20-17N 136-55-43W,
      50-55-14N 165-28-28W, 52-59-09N 165-19-24W.
   E. 13-53-47N 126-52-33W, 11-46-05N 123-56-09W,
      04-19-41S 135-37-56W, 02-14-45S 138-32-32W.
   F. 49-27-33N 135-51-45W, 47-43-47N 136-53-00W,
      50-56-51N 168-09-57W, 52-48-04N 168-20-28W.
   G. 14-27-06N 127-19-28W, 12-18-52N 124-23-30W,
      03-36-29S 136-03-34W, 01-31-24S 138-57-30W.
   H. 49-46-04N 136-40-41W, 48-05-08N 137-37-30W,
      50-55-01N 168-54-51W, 52-42-19N 169-08-13W.
   I. 31-49-12N 124-20-42W, 30-20-18N 122-34-43W,
      22-47-14N 130-25-52W, 24-10-15N 132-10-44W.
2. CANCEL THIS MSG 101917Z AUG 21.

The nine areas A to I cluster in basically three regions (which I have colour-coded in the map above).

The directions of the areas point to a series of deorbits from a 51-53 degree inclined Low Earth orbit. As I have indicated in the map in top of this post, two of the three defined regions with warning boxes line up with the ISS groundtrack during the two time windows given, but I think this is coincidence (and the series of boxes south of Alaska do definitely not line up with the ISS during these two time windows. In fact, this points to deorbits from at least two different orbital planes).

Rather, my suspicion is a mass deorbit of Starlink satellites, who move in ~53 degree inclined orbits [but see update below].

UPDATE: 

After some discussion, Jan Hindrik Knot rightfully questioned whether Starlink satellites, with their ion thruster propulsion, are capable of a controlled deorbit in a designated area at all. That is a good point, which I overlooked initially.

So it appears we have no idea what will be deorbitted on August 9-10.

The combination of the areas in the mid-Pacific and those south or Alaska, to me point to deorbits from at least two different orbital planes (both inclined 51-53 degrees).

Note that, from the position of the areas, the fact that their shapes clearly point to deorbits from Low Earth Orbit, and that the NavWarning mentions time windows on two successive dates, it is clearly not related to this deorbit  (the Spectr-R rocket booster) from Deep Space either.

UPDATE 2:

The plot thickens: the on-line KML version of the Navigational Warning has appeared and mentions: 

"Authority: NASA 300917Z JUL 21"

(the versions sent to subscribers to the service doesn't mention the authorities issuing the warnings).

So it appears to be something NASA-related (HT to @john_moe on Twitter).

One possibility could be that these are emergency landing zones for Starliner (which was to be launched on July 30, the date mentioned in the "Authority:" line: but was scrubbed). Still open questions though: why August 9 and 10? Why where these same zones not published before the July 30 launch date? Questions, questions...

UPDATE 3:

I like the suggestion by Bob Christy that these are warnings for the reentry of the Starliner service module (that is jetissoned from the Starliner capsule before landing of the latter). That makes sense.

The front group of Starlink V1.0 - L26

On May 15, yet another batch of SpaceX Starlink satellites was launched, and they are currently making evening and early night passes over Europe. Since launch, the resulting 'train' of satellites from the launch has split into two distinct groups passing a few minutes after each other.

The video above is from yesterday night (19 May 2021 ~23:17 UT) and shows the leading group, which is still quite tight. The trailing group (also captured but not shown in this video clip) is more dispersed. The video was shot with my DSLR: Canon EOS 80D + Samyang 1.4/85 mm lens.

A night earlier, I captured both the leading and trailing group with the WATEC 902H2 Supreme + FD 1.8/50 mm lens (see video below). They were naked eye during that pass.

[UPDATED] Reentry predictions for the Falcon 9 RB 2021-017BN

click diagram to enlarge

In my previous post, I discussed 2021-017BN, the Falcon 9 upper stage from the March 4 Starlink launch that should have been deorbitted after 1.5 revolutions on March 4th, but didn't.

It is still on orbit. At the moment of writing, 23 March 2021 at 11:00 UT, it is in a 217 x 200 km orbit according to the latest available elements from CSpOC, and it will stay on orbit for a couple of days to come. But the end is near: the orbital altitude of the rocket stage is quickly decaying, as can be seen in the diagram below:

click diagram to enlarge

My current reentry prediction (see diagram in top of post and table below) is that it will come down in the early hours of March 26 (2021). My prediction, based on modelling in GMAT R2020a using the MSISE90 model atmosphere, appears to be well in line with the TIP from CSpOC so far.

[UPDATE: my final post-cast predicted reentry at 26 Mar 04:34 UT, which is some 35 minutes too late. It is based on a 2/3rd maximum drag surface value. Interstingly, using the maximum drag surface leads to a reenrty at 3:56 Ut, within minutes f the actual time]

Revisit this post for prediction updates in the coming days.

orbit epoch     pred. date     reentry time (UT)
21081.600725    26 Mar 2021    02:33 +- 16.8 hr
21081.922054    26 mar 2021    03:36 +- 15.5 hr
21082.113317    26 Mar 2021    03:59 +- 14.7 hr
21082.216601    26 Mar 2021    03:40 +- 14.1 hr
21082.278149    26 Mar 2021    03:43 +- 13.8 hr
21082.462749    26 Mar 2021    05:29 +- 13.3 hr
21082.585776    26 Mar 2021    05:29 +- 12.7 hr
21082.708770    26 Mar 2021    05:37 +- 12.1 hr
21082.954651    26 Mar 2021    06:13 +- 11.1 hr
21083.138960    26 Mar 2021    05:03 +-  9.9 hr
21083.261785    26 Mar 2021    05:15 +-  9.4 hr
21083.296296    26 Mar 2021    05:20 +-  9.2 hr
21083.507296    26 Mar 2021    05:28 +-  8.3 hr
21083.875164    26 Mar 2021    05:26 +-  6.5 hr
21084.120127    26 Mar 2021    05:59 +-  5.4 hr
21084.181325    26 Mar 2021    05:20 +-  5.0 hr
21084.486963    26 Mar 2021    05:00 +-  3.5 hr
21084.548018    26 Mar 2021    03:19 +-  2.8 hr
21084.974688    26 Mar 2021    04:46 +-  1.1 hr * post-cast
21085.095602    26 Mar 2021    04:34 +-  0.5 hr * final post-cast

UPDATE  26 March 2021  12:30 UT:

The reentry happened last night, over North America, and was widely seen from the US States Washington and Oregon, near 4:00 UT (March 26 UT: that is 9 pm on March 25 local time for that area).

CSpOC's final TIP places the reentry at 03:58 +- 1 min UT. This time matches the reports from Washington and Oregon well, and based on the last orbit it would indeed place the rocket stage near the NW United States coast.

The listed geographic position in the TIP, 24.5 N, 151 W, does however not match well (it is further down the track, near Hawaii, corresponding to the Falcon 9 position about 6 minutes prior to the observed reentry). We have  noted such discrepancies more often in recent TIP messages. In this case, I half suspect the position was that given by their reentry model, and they forgot to update it when the SBIRS detection of the actual reentry fireball came in.

click map to enlarge

My own final "post-cast" places reentry some 35 minutes after the actual reentry.

Here are some of the reentry sightings as reported on Twitter:

DID YOU SEE THIS? I have never in my life seen something so incredible. I am in awe. Just happened over Portland about 10 minutes ago. #LiveOnK2 pic.twitter.com/L9wLEXBrcW

— Genevieve Reaume (@GenevieveReaume) March 26, 2021

uuuuuhhhh holy shit we just saw this meteor shower across the sky?!?! #pdx pic.twitter.com/umaE0joW7J

— lynseylou (@lynseylou) March 26, 2021

 

UPDATE 2 April 2021 23:00 UT:

Debris has been recovered from this reentry. In Grant Country, Washington, a Composite Overwrapped Pressure Vessel (COPV) was found on farmland.

 

 

SpaceX recovered a Composite-Overwrapped Pressure Vessel from last week’s Falcon 9 re-entry. It was found on private property in southwest Grant County this week. Media and treasure hunters: we are not disclosing specifics. The property owner simply wants to be left alone. pic.twitter.com/dEIQAotItY

— Grant County Sheriff (@GrantCoSheriff) April 2, 2021

Apparent failed deorbit of the Starlink-18 Falcon 9 upper stage [UPDATED]

On 4 March 2021, after several delays, SpaceX launched the 18th Starlink batch (Starlink-18 or V1.0-17). While the launch and deployment profile appears to have been similar to other recent Starlink launches, it appears that something went wrong with the Falcon 9 upper stage near the end of its mission.

On March 8th, Polish observer Adam Hurcewicz reported a bright, fast object in the orbital plane of this launch, passing a few minutes before the main Starlink "train". It was seen on subsequent nights and by other observers as well: the video above is from the early morning of March 9. At it's brightest, this fast moving object reportedly reaches mag -3. It does not appear to match a known object from earlier launches. It also didn't match supplementary TLE's for the Starlink-18 payloads from Celestrak (which are based on State Vectors from SpaceX). The Polish observers therefore speculated it was the Falcon 9 upper stage from the launch. 

But that would be against expectations. The Falcon 9 upper stage normally does not stay in orbit: it is de-orbitted soon after payload release, usually about 1.5 revolutions (about 2.5 hours) after launch. So if this object is the Falcon 9 upper stage, this suggests  something went wrong and it failed to deorbit.

The speculation that this object is the Falcon 9 Upper Stage can now be bolstered by additional information. The first orbital element sets for this Starlink launch have appeared on the CSpOC portal  Space-Track late yesterday (11 March), with catalogue numbers ranging from 47722 to 47786. And they show an extra object!

With Starlink launches, 64 objects are usually catalogued: 60 payloads and four 'Falcon 9 debris' pieces. The latter 'debris' pieces are the payload stack retaining rods: four metal rods which keep the satellite stack together on top of the upper stage. They are jettisoned upon payload release.

An elset for the Falcon 9 upper stage is usually not released by CSpOC: as it normally stays on-orbit for barely more than 1 revolution, it is not catalogued.

But this time, not 64 but 65 objects have been catalogued. The extra 65th object must be the Falcon 9 upper stage, and it indicates it stayed on orbit for more than a few revolutions. Which lines up with the observations by the Polish (and later also other) observers.

Although the 65 objects, at the moment of writing, do not have been individually ID-ed by CSpOC yet (all have the temporary designation "TBA - TO BE ASSIGNED"), the 60 payloads, four retaining rods and the upper stage as such can be clearly identified among them. The objects separate in 3 groups in terms of orbital altitude. The 60 payloads all have (for orbits with epoch 12 March) a perigee above 280 km. The four retaining rods have clearly lower orbits: their perigee is near 243-246 km and apogee near 268-278 km.

The 65th object, which by inference must be the Falcon 9 upper stage, is in a still lower orbit . It has the smallest semi-major axis of all of them with perigee near 237 km and apogee near 270 km. The orbit for this object, catalogue nr 47782 (2021-071BN) also closely matches the observations by the Polish observers.

So why is the Falcon 9 upper stage still on-orbit? It suggests of course that the deorbit went not as planned, i.e. it failed for some reason (e.g. the rocket engine refusing to restart).

That the Falcon 9 upper stage should have deorbitted on March 4, after 1.5 revolutions, is clear from the Navigational Warnings that were issued in connection to this launch. Navigational Warning HYDROPAC 695/21 delineates the usual elongated deorbit zone in the Indian Ocean familiar from earlier Starlink launches:

 

021948Z MAR 21
HYDROPAC 695/21(GEN).
SOUTHERN INDIAN OCEAN.
1. HAZARDOUS OPERATIONS, SPACE DEBRIS
   041024Z TO 041326Z MAR,
   ALTERNATE 051004Z TO 051306Z MAR
   IN AREA BOUND BY
   29-43S 060-07E, 24-55S 064-27E,
   38-45S 084-30E, 45-12S 099-45E,
   49-46S 119-13E, 50-42S 138-19E,
   48-50S 156-44E, 51-46S 158-08E,
   54-42S 148-32E, 56-20S 131-03E,
   55-52S 107-50E, 49-11S 085-05E,
   34-32S 064-13E.
2. CANCEL HYDROPAC 685/21.
3. CANCEL THIS MSG 051406Z MAR 21.

I have plotted the zones from the Area Warnings connected to the launch in this map, along with the groundtrack for the first 1.5 orbital revolutions. The large elongated red zone in the southern Indian Ocean is the planned deorbit area from Navigational Warning HYDROPAC 695/21:

click map to enlarge

The position of the reentry hazard zone indicates a reentry was planned around 10:55 UT (March 4), 1.5 revolutions (2h 30m) after launch, following a deorbit burn some 30 minutes earlier.

But the deorbit evidently did not happen as it should have: the upper stage is still orbiting as we speak, a week after launch. The issued Navigational Warning for the deorbit hazard zone strongly suggests this is not intentional.

So how long will the upper stage stay in orbit? The current orbit is low (237 x 271 km), and the object is large (16 x 3.66 meter, with a mass of 4.5 tons) so eventually the rocket stage will have an uncontrolled reentry, somewhere between latitudes 53 deg N and 53 deg S. 

A first assessment using both SatEvo and a GMAT simulation suggests that the reentry will probably happen in the last few days of March or the first few days of April.

UPDATE 14 March 2021:
CSpOC has now added identifications to the objects, and indeed object 47782 is now listed as "Falcon 9 RB"