Kosmos 482 Descent Craft reentry forecasts [PERIODICALLY UPDATED]

Click diagram to enlarge

Last update: 11 May 2025 10:15 UTC 

 

(this post is being periodically updated with new reentry forecasts/postcasts)

 

* The last Tudat nominal forecast is reentry on 10 May 2025, 6:39 UTC ± 1.5 hr *

* The Tudat nominal aftercast is reentry on 10 May 2025, 6:40 UTC ± 1.5 hr *

 

 

 

In the second week of May 2025, an unusual object reentered. It concerns the Kosmos 482 Descent Craft (1972-023E, cat. nr. 6073).

This object is the lander module from a 1972 failed Soviet Venera mission to Venus. Because of a failure of the upper stage of the rocket that launched it, it got stuck in a very elliptical orbit around Earth in 1972, instead of going to Venus. 

 

I published an extensive discussion and analysis of this object and its history three years ago in The Space Review. The identification of this object as the lander module was initially suggested by Jonathan McDowell (see his brief interesting history here).

Recently declassified Russian historic documents unearthed by Anatoly Zak point out that after failure to get to a heliocentric orbit, the lander was deliberately separated from the main bus by the spacecraft operators. My analysis published in 2022 suggests that this happened in June 1972. The lander is encased in a semi-spherical shaped Titanium protective shell, a kind of rounded metal bucket so you will (see image below).

As this is a lander that was designed to survive passage through the Venus atmosphere, it is possible that it will survive reentry through the Earth atmosphere intact, and impact intact. It likely will be a hard impact: I doubt the parachute deployment system will still work after 53 years and with dead batteries. There are many uncertain factors in whether the lander will survive reentry though, including that this will be a long shallow reentry trajectory, and the age of the object.

 

 

Venera 7 lander mock-up. The Kosmos 482 Descent Craft is probably similar. Photo: NASA

 

The risks involved are not particularly high, but not zero: with a mass of just under 500 kg and 1-meter size, risks are somewhat similar to that of a meteorite impact. A Tudat reentry analysis to ground level suggests an impact speed (after atmospheric deceleration) of about  65-70 meter per second (~242 km/h), assuming the reentering lander did not break up or extensively ablate during reentry (see the diagram below: note the logarithmic scale of the x-axis). The kinetic energy at impact is similar to that of a 40-55 cm large (after ablation) meteorite fragment. As it will likely reach earth surface as only one single object, the risks involved are lower than for example those created by a Falcon 9 upper stage reentry, which showers multiple meter-sized objects over a large area (as we saw recently with the impact of Falcon 9 remains in Poland).

 

Click diagram to enlarge

 

The diagrams below shows the change in altitude of apogee and perigee over the past 1.5 years and the past 4 months: notably apogee has been coming down steadily, but in the past few months, perigee has started to come down too. Early May 10 around 00:35 UTC, the object was in a 166 x 120 km orbit, with apogee coming down by 70 km/day and perigee by 25 km/day (and these values increasing each day, see diagram below).

 

Click diagram to enlarge

 

Click diagrams to enlarge

 

The reentry is an uncontrolled reentry. Even close to reeentry, we cannot say with any degree of certainty when and where the Kosmos 482 Descent Craft exactly will reenter.

With an orbital inclination of 51.95 degrees [note: I initially erroneously stated 51.7 here], the reentry can occur  between latitude 52 N and 52 S.

Over the past months, together with my colleague Dominic Dirkx,  we have been developing a reentry model for this object in Tudat.

 

Tudat, the TU Delft Astrodynamics Toolbox is open source, multi-platform Astrodynamics software developed and maintained at the Aerospace faculty of Delft University of Technology (TU Delft) in the Netherlands (where I work). The Tudat script we use for our analysis is here, while the Tudat software itself is available here.

 

The Kosmos 482 Descent Craft is probably similar to the descent craft of Venera 8 (which was launched only a few days earlier in 1972). Literature values suggest that the object is about 1 meter in size and semi-spherical, with a mass of ~495 kg. 

 

Using our Tudat model and a 1-meter diameter, and the NRLMSISE00 model atmosphere with historic space weather data, we find that the orbital evolution of  the object from mid-1972 to early 2025 is actually best matched when we use a mass of 480 kg, 15 kg less than the literature value. All our forecast predictions are therefore done using a mass of 480 kg. 

 

For our forecasts we employ the NRLMSISE00 model atmosphere, and historic spaceweather data plus estimated future space weather (solar flux). Orbit updates for each run are sourced from the US Military tracking network CSpOC (so as a note: our forecasts are not based on our own tracking data. They are based on our own custom reentry model developed at Delft University of Technology, with external tracking data from CSpOC as input).

 

Click diagram to enlarge

The diagrams in top of this post and below (which I will periodically update) give the evolution of our reentry predictions, based on orbits issued for the object since November 2024. Over the past half year, the model has consistently pointed to reentry within a few days of 9-10 May 2025. 

 

The latest nominal forecasts currently center on early May 10, but are still fluctuating with each orbital update. Early May, there was a short-lived tendency in the model runs of a shift towards a later time of reentry (i.e. May 11 rather than May 10), as can be seen in the below diagrams and the table near the end of this post. But lately, the new forecast runs have been returning to nominally May 10. 

 

 

Click diagram to enlarge

 

Click diagram to enlarge

 

The brief shift towards a later date in the early May model runs was due to space weather forecasts underestimating future solar activity at that time. I had started to note this (when comparing earlier forecast solar flux values to actual solar flux values for the last days of April  and the first days of May) and already expected the nominal forecasts to return to May 10, as they now indeed have done.

 

This highlights the influence of uncertainties in estimated future solar activity on the forecasts. Future solar activity is not well predictable. Even the best reentry models will be off in their forecast if actual solar activity in the window between the moment of forecast and the moment of reentry develops differently than the predictions. Solar flux variations are an important driving factor of short-term variations in density and extent of the upper atmosphere. The latter determine how many drag the object experiences over its orbit, and with that how quick the orbit decays. Stronger solar activity will mean an earlier reentry, lower solar activity a later reentry.

 

Of course, at this point the uncertainty in the forecast is still +/- 1.5 hours. Yet, the uncertainty window has now decreased to the point that we can rule out certain parts of the world: Africa is safe. So is Japan, East Asia, and the USA.

 

The map below gives the trajectory of the object over the current uncertainty window of the Tudat reentry forecast (red dots are cities with over 1 million inhabitants, between latitude 52 N and 52 S). In our model the spacecraft could have reentered anywhere along the blue line:

 

Click map to enlarge

 

 

CSpOC, the US military tracking network, has started to provide TIP (Time of Impact Prediction) messages for the object since May 6. They are depicted as red circles in the diagram earlier in this post.

 

The last orbit update published by CSpOC has epoch 10 May 2025, 00:36 UTC, i.e. dates from about 6 hours before the time we think it actually reentered. There are later radar detections from Europe, per ESA, dating to as late as 6:04 UTC, but no orbit based on that has been published.

A roundup of current reentry forecasts from various sources (each running their own reentry model) compiled 10 May 15:30 UTC:

 

Tudat forecast:              10 May   06:39 ± 91  min UTC

Tudat postcast:              10 May   06:40 ± 91  min UTC

Aerospace Corp forecast:     10 May   06:29 ± 120 min UTC

RosCosmos postcast:          10 May   06:24 ± ?   min UTC

ESA postcast:                10 May   06:16 ± 22  min UTC

EU-SST forecast:             10 May   06:04 ± 20  min UTC

CSpOC TIP postcast:          10 May   05:32 ± 12  min UTC *

 

* this cannot be correct given that a positive radar detection from Germany at 6:04 UTC was reported by ESA.

 

note: forecast = issued before the reentry; postcast = issued after the reentry after re-analysis.

 

ESA reports that they had a positive radar detection of the spacecraft, still on-orbit, at 6:04 UTC while it was passing over Germany (presumably with the TIRA radar); and a negative detection one orbital revolution later, at 7:32 UTC. This implies reentry during the window 6:04 - 7:32 UTC. Five of the forecasts fall in that window: our Tudat forecast, and those by RosCosmos, ESA, the Aerospace Corporation, and also EU-SST when taking the error margins on the latter into account.

 

Roscosmos, the Russian Space Agency, on Telegram states reentry at "6:24 UTC" over the Indian Ocean. That is not too far (15 minutes) from our latest TUDAT model results. Three remarks: (1) this is not a reentry detection but likely another model result (just as the others), based on an earlier groundbased detection during a pass over presumably southern Russia. (2) It does not come with a stated uncertainty window: is that 10 minutes, 30 minutes, an hour? (3) It remains a question how serious one should take Russian State statements these days as sometimes pragmatic considerations (such as deliberate denial of responsibility or risk) are leading. 

 

While many news media seem to treat the RosCosmos announcement as 'authoritive' and the 'most accurate' or 'final' say on the matter, there is in reality no clear reason to do so. It is not clear at all whether the RosCosmos position is more accurate than the other model estimates, and how meaningful the stated positions and time are given the lack of error margin information.

A final accurate TIP from CSpOC has not yet appeared (and might not appear if they have no space-based detections of the reentry fireball). CSpOC sometimes, several hours to a day after a reentry, publishes a very accurate (with quoted +- 1 minute uncertainty) final TIP which Jonathan McDowell and I believe is not based on a reentry model, but on space-based (SBIRS satellite) detections of the reentry fireball, hence the accuracy. They have not issued such a TIP so far for this reentry, unfortunately. Note that their last issued TIP cannot be correct, given the reported positive radar detection of the spacecraft passing over Germany at 6:04 UTC.

 

This is how the nominal reentry points have been placed by various organisations, each based on their own modelling:

 

Click map to enlarge

 

It is clear that the assessments spread. Ignoring the clearly incorrect CSpOC TIP, the other models all suggest reentry over either southwest Asia or the Indian Ocean. My verdict is that the latter area is the most likely place.

 

Below is the evolution of the reentry forecast from our TU Delft Tudat model in tabular form, latest forecast at the bottom. Please take note of the uncertainty values listed in the last column!

 

The 'postcast' value is a re-run of the model after-the-fact based on the last published orbit and the actual rather than estimated space weather of May 10.

 

--------------------------------------------------------------------
TUDAT REENTRY FORECAST EVOLUTION for KOSMOS 482 Descent Craft
M. Langbroek & D. Dirkx, Delft University of Technology

Date/times in UTC

REFERENCE ORBIT    ORBIT EPOCH      REENTRY FORECAST   +/-   
----------------------------------------------------------------
15-11-2024 05:43   24320.23870400   05-05-2025 23:33   42.9 days   
01-12-2024 05:32   24336.23116452   08-05-2025 09:09   39.5 days   
15-12-2024 18:58   24350.79080401   07-05-2025 11:51   35.7 days   
01-01-2025 11:20   25001.47260254   09-05-2025 07:20   32.0 days   
15-01-2025 03:23   25015.14140884   10-05-2025 20:40   28.9 days   
02-02-2025 08:54   25033.37115746   13-05-2025 17:52   25.1 days   
15-02-2025 03:20   25046.13941015   11-05-2025 09:52   21.3 days   
01-03-2025 00:07   25060.00535989   10-05-2025 17:51   17.7 days   
15-03-2025 05:56   25074.24770046   10-05-2025 07:57   14.0 days   
30-03-2025 12:05   25089.50360681   09-05-2025 21:11   10.1 days   
13-04-2025 21:32   25103.89775709   09-05-2025 22:01    6.5 days   
20-04-2025 01:39   25110.06916305   09-05-2025 11:31    4.9 days   
22-04-2025 21:24   25112.89204293   09-05-2025 12:48    4.2 days   
23-04-2025 22:57   25113.95657237   09-05-2025 19:43    4.0 days   
27-04-2025 00:27   25117.01893077   10-05-2025 04:52    3.3 days   
28-04-2025 00:24   25118.01685903   10-05-2025 06:33    3.1 days   
28-04-2025 22:50   25118.95143786   10-05-2025 06:01    2.8 days   
01-05-2025 20:57   25121.87323063   10-05-2025 14:30    2.2 days   
02-05-2025 09:12   25122.38386703   11-05-2025 02:36    2.2 days   
02-05-2025 12:16   25122.51167762   11-05-2025 03:41    2.2 days   
02-05-2025 17:52   25122.74476620   11-05-2025 06:50    2.1 days   
03-05-2025 18:07   25123.75533175   10-05-2025 20:48    1.8 days   
03-05-2025 21:05   25123.87876286   10-05-2025 23:06    1.8 days   
04-05-2025 20:47   25124.86648285   10-05-2025 19:05    1.5 days   
05-05-2025 20:29   25125.85361170   10-05-2025 07:26    1.1 days   
06-05-2025 09:39   25126.40254557   10-05-2025 08:37    1.0 day
06-05-2025 21:26   25126.89335737   10-05-2025 07:51   20.6 hr
07-05-2025 21:00   25127.87559184   10-05-2025 07:34   14.6 hr
08-05-2025 11:31   25128.48053679   10-05-2025 07:24   11.0 hr
08-05-2025 13:13   25128.55136979   10-05-2025 07:34   10.6 hr
08-05-2025 19:13   25128.80083450   10-05-2025 07:54    9.2 hr
09-05-2025 12:37   25129.52575525   10-05-2025 06:23    4.4 hr
09-05-2025 14:12   25129.59226104   10-05-2025 06:30    4.1 hr
09-05-2025 15:43   25129.65543839   10-05-2025 06:35    3.7 hr
09-05-2025 19:55   25129.82989846   10-05-2025 06:34    2.7 hr
10-05-2025 00:35   25130.02495443   10-05-2025 06:39    1.5 hr

                               nominal location 35.7 S 126.5 E

POSTCAST:

10-05-2025 00:35   25130.02495443   10-05-2025 06:40    1.5 hr

                               nominal location 38.0 S 129.8 E

 

I will periodically update this table with new forecasts, more frequently so when the reentry dates comes nearer.

Here is footage I shot of the Kosmos 482 Descent Craft with my tracking camera in Leiden, in 2020:

An added note: about that parachute....

To muddy the waters further, a story is spreading that the parachute of the landing craft might already have deployed in space. This is based on telescopic imagery purportedly showing this.

I have strong doubts that the imagery in question shows any meaningful detail. The imagery has the same origin, and the same problems attached, as the imagery I discussed in 2022 in my Space Review article. I think the "detail" is the result of camera/telescope shake and atmospheric distortion.

 

UPDATE 28 May 2025:
With my colleague Dominic Dirkx, I wrote a 'post mortem' on the reentry for The Space Review that can be read here.

This post has been updated with new forecasts and additional background information several times

Delfi-C3 reentry forecast updates (periodically updated post)

 

[ Post last updated:  14 Nov 2023  12:45 UTC ]

In my October 25 blogpost I presented imagery of the iconic 3U cubesat Delfi-C³ (2008-021G) taken by my tracking camera. I also provided a reentry forecast in that post, that I updated several times.That reentry is currently (mid November 2023) very near.

I am now consolidating the reentry forecasts in this current, periodically updated blogpost.

While late October the reentry forecast was still shifting to a later date with each orbit update, that shift is now flattening out, but the forecast is still fluctuating. 

A geomagnetic storm that caused aurora at middle latitudes on Nov 5/6 has had a clear effect in speeding up orbital decay (the 'dip' can be seen in the digrams below). Over the weekend 11-12 November, solar activity was more mild than forecast, making the reentry forecast slowly shift to a later time with each orbital update.

My current "aftercast", based on a mid-November 13 orbit modelled in GMAT, is reentry on 13 November 2023, nominally near 20:39h UTC +- 1.5 hours (please note the uncertainty interval!), i.e. between 19 - 22 UTC (and likely  between 20 - 21 UTC)

A SatEvo/SatAna analysis results in a very similar, if slightly earlier, time (nominally 19:34 +- 1.5h UTC). I do not expect it survived to November 14.

The orbit determination on which these 'aftercasts' are based, is from about 3 orbital revolutions before the forecasted reentry moment. This orbit can be of bad quality, as in these final stages, orbit determination is tricky.

Here is how the reentry forecast has been developing so far (note the sudden downward developing trend due to the Nov 5/6 geomagnetic storm):

click diagrams to enlarge

 

This is probably the last reentry forecast update (or rather "aftercast" by now), as I do not expect new orbital elements to be released between this update and the reentry

(Note: forecast data in tabular form are at the bottom of this post)

This is the Delfi-C³ ground-track over the current uncertainty interval (spanning two orbital revolutions) in the forecast:

 

click map to enlarge

These reentry predictions are based on a tumbling satellite with, due to the tumbling, an average drag surface of about 60% of the maximum drag surface. Previous experience with modelling reentries in GMAT has shown this to be a reasonable value for a tumbling object. 

What makes these reentry forecasts challenging, and creates the large uncertainties in the reentry position is a combination of things. One is the, even at short timescales, varying solar activity and it's influence on the upper atmosphere, which cannot be well predicted or captured by the model. Second, during the last orbits just before reentry, all kind of gas flow mechanisms around the spacecraft are taking place, which alter tyhe cdrag it expriences. Moreover, in this final stage where the spacecraft meets a more thicker resistance of the upper atmosphere, it sometimes auto-stabilises itself in a least-drag orientation, prolonging survival slightly. Last but not least: in the last moments of the spacecraft's life, when it is in a very low, fast changing orbit, orbit determination becomes difficult, affecting tehe quality of the last orbital elements available for the forecast.

I do not expect a TIP to be issued for this object. TLE updates might end a few hours before the actual reentry, after which CSpOC issues an 'administrative decay'.

Delfi-C³ 's orbit had decayed to below 187 x 177 km by mid 13 Nov 2023, and the cubesat was coming down increasingly fast, as can be seen in the diagrams below (the orbit was dropping by almost 100 km/day mid 13 Nov 2023):

 

click diagrams to enlarge

click diagram to enlarge

Note that Delfi-C³ is very small and lightweight (2.2 kg and 30 x 10 x 10 cm) and the reentry will have been completely harmless. The friction with the molecules in the atmosphere will heat up the spacecraft during its final moments, untill it burns up completely, at altitudes above 50 km. The current ground-track where it can come down, is almost completely over Ocean.

The operators of Delfi-C³ have put out a call to radio amateurs to try to receive and decode telemetry from the cubesat. Software to decode is provided here. The goal is to try to follow Delfi-C³ and get telemetry as close to reentry as possible.

Note that Delfi-C³ only sends telemetry when it is in sunlight, and not continuously as after 15 years the system is showing issues. The frequency is 145.870 MHz, it is sending 1200 Bd BPSK packages. See also here. 

[EDIT:  as off 11 november, the cubesat seems to have shifted to a backup frequency near 145.934 MHz, as can be seen in this spectogram from my detection in the morning of November 11. The 145.870 MHz frequency is also still intermittently active. So pay attention to both frequencies]

click spectrogram to enlarge

[end of edit]

On the morning  of  Nov 8, during the 7:43 UTC pass, I received thise radio signal of Delfi-C3 from Leiden, the Netherlands (but could not decode the signal), clearly a bit early on the then available TLE:

click image to enlarge

The "stepped" effect is due to attempts to correct the Doppler shift. That Doppler shift (the "s"-shape in frequency due to the Doppler effect in the signal) can be very nicely seen in this spectogram from the same pass as received by my colleagues at Delft University of Technology (image courtesy Bart Root, TU Delft):

click image to enlarge. Image courtesy Bart Root, TU Delft

 

Radio signals from the cubesat were also received by me and the TU Delft station on later dates, most recently in the morning of November 13th during the 7:24 UTC pass:

Nov 13 ~7:24 UTC radio spectrogram from Leiden. Click image to enlarge

Delfi-C³ is a 3U cubesat developed and built at Delft Technical University in the Netherlands, and launched 15 years ago, in 2008.

More on Delfi-C³ on the website of Delft Technical University. There is also a nice background article on it's 15-year anniversary here.

The last (and only!) visual image of Delfi-C3 on-orbit, taken by me on 24 October 2023 from Leiden, the Netherlands, using a WATEC 902H2 Supreme and Samyang 2.0/135 mm lens at 25 fps (image is a frame-stack of 16 frames: the faint trail lower right is Delfi-C3 moving through the field of view. See also blogpost here which also features some video footage):

 

TABLE:  Forecast development in tabular form (most recent one at bottom).
All dates and times are in UTC. Quoted times are nominal times, please take into account the uncertainty intervals that go with them!

orbit date         orbit epoch     reentry forecast + uncertainty      
08 okt 2023 13:52  23281.57810544  05 nov 2023 02:23  ±  8.3  day
09 okt 2023 14:28  23282.60344413  05 nov 2023 08:17  ±  8.0  day
10 okt 2023 19:41  23283.82047902  05 nov 2023 11:00  ±  7.7  day
12 okt 2023 00:53  23285.03692402  05 nov 2023 20:20  ±  7.4  day
13 okt 2023 01:27  23286.06083878  06 nov 2023 18:43  ±  7.4  day
14 okt 2023 00:29  23287.02034889  06 nov 2023 18:53  ±  7.1  day
15 okt 2023 08:42  23288.36296159  07 nov 2023 00:59  ±  6.8  day
16 okt 2023 00:02  23289.00204702  07 nov 2023 04:44  ±  6.7  day
16 okt 2023 15:23  23289.64097529  08 nov 2023 02:01  ±  6.7  day
17 okt 2023 18:58  23290.79063916  08 nov 2023 12:14  ±  6.5  day
18 okt 2023 19:29  23291.81212343  08 nov 2023 21:43  ±  6.3  day
20 okt 2023 00:35  23293.02450124  09 nov 2023 16:03  ±  6.2  day
21 okt 2023 01:04  23294.04489630  09 nov 2023 19:49  ±  5.9  day
21 okt 2023 19:26  23294.80984121  09 nov 2023 22:19  ±  5.7  day
22 okt 2023 19:54  23295.82928392  09 nov 2023 23:43  ±  5.4  day
24 okt 2023 07:03  23297.29389307  10 nov 2023 23:54  ±  5.3  day
26 okt 2023 00:17  23299.01200402  11 nov 2023 11:29  ±  4.9  day
27 okt 2023 00:42  23300.02946319  11 nov 2023 12:23  ±  4.6  day
28 okt 2023 01:06  23301.04628230  11 nov 2023 16:02  ±  4.4  day
28 okt 2023 19:24  23301.80847586  11 nov 2023 18:46  ±  4.2  day
30 okt 2023 00:20  23303.01441157  12 nov 2023 09:54  ±  4.0  day
31 okt 2023 00:41  23304.02910990  12 nov 2023 13:04  ±  3.8  day
01 nov 2023 01:02  23305.04306068  12 nov 2023 17:28  ±  3.5  day
02 nov 2023 11:58  23306.49918974  12 nov 2023 17:44  ±  3.1  day
03 nov 2023 07:42  23307.32139835  12 nov 2023 23:36  ±  2.9  day
04 nov 2023 07:58  23308.33247480  12 nov 2023 19:20  ±  2.5  day
05 nov 2023 00:38  23309.02698061  12 nov 2023 21:03  ±  2.4  day
06 nov 2023 00:51  23310.03600547  12 nov 2023 05:15  ±  1.9  day
07 nov 2023 01:02  23311.04321962  12 nov 2023 09:23  ±  1.6  day
07 nov 2023 13:06  23311.54617034  12 nov 2023 14:38  ±  1.5  day
07 nov 2023 14:36  23311.60900895  12 nov 2023 14:25  ±  1.5  day
08 nov 2023 01:10  23312.04868443  12 nov 2023 15:48  ±  1.4  day
08 nov 2023 13:13  23312.55073768  12 nov 2023 19:28  ±  1.3  day
08 nov 2023 19:14  23312.80157938  12 nov 2023 20:14  ±  1.2  day
08 nov 2023 20:44  23312.86427030  12 nov 2023 20:32  ±  1.2  day
09 nov 2023 01:15  23313.05229412  12 nov 2023 21:16  ±  1.2  day
09 nov 2023 13:16  23313.55332040  12 nov 2023 20:49  ±  1.0  day
09 nov 2023 14:46  23313.61590941  12 nov 2023 20:54  ±  1.0  day
09 nov 2023 19:17  23313.80362555  12 nov 2023 21:54  ±  22.4  hr
10 nov 2023 01:17  23314.05377944  12 nov 2023 23:07  ±  20.9  hr
10 nov 2023 14:47  23314.61604116  13 nov 2023 01:14  ±  17.5  hr
10 nov 2023 20:46  23314.86565574  13 nov 2023 01:32  ±  15.8  hr
11 nov 2023 01:15  23315.05274319  13 nov 2023 02:37  ±  14.8  hr
11 nov 2023 13:13  23315.55109903  13 nov 2023 06:40  ±  12.4  hr
11 nov 2023 19:11  23315.79995400  13 nov 2023 08:42  ±  11.3  hr
12 nov 2023 01:09  23316.04856854  13 nov 2023 10:21  ±  10.0  hr
12 nov 2023 14:34  23316.60695182  13 nov 2023 14:53  ±   7.3  hr
13 nov 2023 00:57  23317.03998349  13 nov 2023 17:11  ±   4.9  hr
13 nov 2023 11:18  23317.47133062  13 nov 2023 19:55  ±   2.6  hr
13 nov 2023 12:47  23317.53274184  13 nov 2023 20:06  ±   2.2  hr

FINAL 'aftercast':
13 nov 2023 15:53  23317.65534723  13 nov 2023 20:39  ±   1.5  hr

A Goodbye to Delfi-C3

Delfi-C3 tracklet on a stack of 16 video frames, 24 Oct 2023

 (note: new reentry forecasts are now published in a separate dedicated blogpost)

Earlier this year, my TU Delft colleague Stefano Speretta asked me if I could try to I imaged a pass of the TU Delft-built 3U cubesat Delfi-C³ (2008-021G). 

It next took a long time due to initially unfavourable observing geometries and - when the observing geometries got better - the Dutch weather, but I finally managed to successfuly image it in late evening twilight of 24 October 2023. 

Above is a stack of 16 video frames showing the tracklet created by the cubesat; below is the actual video footage, shot with a WATEC 902H2 Supreme camera and Samyang 2.0/135 mm lens at 25 fps (the object is very faint in the footage, due to it being very small in size):

 

Delfi-C³ (2008-021G) is a 3U cubesat and was the first cubesat built in the Netherlands. It was launched on 28 April 2008 from Satish Dawan in India on a PSLV rocket, as part of a rideshare mission.

It was built by students of Delft Technical University (my current employer) as the first in what was to become a line of Dutch-designed-and-built cube- and pocketsats. It carried at that time experimental technology (autonomous star sensors and thin-film solar panels) and an amateur radio responder. More information on the cubesat, its mission and the technology onboard can be found at the TU Delft website.

Although no longer operational, there is occasionally still radio telemetry received from the cubesat by our TU Delft Rooftab Radio lab.

 

Delfi-C³ in stowed condition (image: TU Delft)

Delfi-C³in deployed condition (image: TU Delft)

Fifteen years after launch, it is time to say Goodbye this very successful cubesat. It has less than half a month to live.

 

[post NO LONGER UPDATED below. Instead refer to this new dedicated post here for new reentry forecast updates]

Initially launched into a 615 x 635 km, 98.0 inclined orbit, the orbit has now decayed to 321 x 324 km (status as of 3 Nov 2023), and the cubesat is coming down increasingly fast, as can be seen in the diagrams below (currently, the orbit is dropping by 4 km/day (status 3 Nov 2023)):

click diagrams to enlarge

click diagram to enlarge

 

Delfi-C³ will probably reenter into the atmosphere and burn up somewhere mid-November 2023 (depending on how solar activity develops over November).

Here is the evolution of my GMAT reentry forecast so far (ignore the error bars and quoted  reentry date uncertainty for now, as this far before reentry they still have little meaning, due to the uncertainty in future solar activity):

click diagram to enlarge

The initial shift of the forecast over time towards a progressively later date is slowing down. My best guess at this moment (3 Nov2023) is reentry mid-November, around 12-14 November 2023, plus-minus a few days.

I will update this figure over the coming days and weeks, as the reentry forecast develops. 

New reentry forecasts are now published in a separate dedicated blogpost.

I do not expect a TIP to be issued for this object. Recent experience shows that TLE updates will probably cease some 2-3 days before the actual reentry, after which CSpOC issues an 'administrative decay'.

Delfi-C³ is very small and the reentry will be completely harmless, with the cubesat burning up completely.

The upcoming (?) reentry of Kwangmyŏngsŏng 3-2 (KMS 3-2), North Korea's last remaining satellite

KMS 3-2 (image: KCNA)

Somewhere this week, North Korea's last remaining - albeit probably not functional - satellite on orbit, Kwangmyŏngsŏng 3-2 (KMS 3-2, 2012-072A), will reenter (or has it already? See later...).

Kwangmyŏngsŏng 3-2 was launched on 12 December 2012 and was North Korea's first succesful satellite launch. Following in the footsteps of a later launch, KMS 4 from 2016 which reentered earlier this year on 30 June 2023, it is now set to have an uncontrolled reentry. 

With this, North Korea will lose the last satellite they have on orbit (the UNHA-3 upper stage of this launch is still on orbit).

It is not clear whether KMS 3-2, said to be a remote sensing satellite to monitor crops and weather, was ever functional. To my knowledge, no independent reception of signals from the payload has ever been reported.

Unfortunately, orbital updates for KMS 3-2 stopped to appear in the CSpOC catalogue after September 12. The last available orbit has epoch 12 Sept 2023 15:20:30 UTC. My reentry forecast based on that orbit was 16 Sept 2023 18:00 UTC ± 20 hrs.

 

click diagram to enlarge

 

A TIP was not issued (it wasn't for KMS-4 either at the time), but a reentry notice was suddenly added to the KMS 3-2 entry in the CSpOC SATCAT on Sept 15, stating 13 September 2023 as the reentry date.

I think that date is not correct, and merely an 'administrative decay'  - meaning they either don't really know, or are not willing to share, when it really reentered. 

The last known orbit from September 12 is a 226 x 235 km orbit and even using extremely high values for solar activity (higher than they in reality were) I can't get it to reenter earlier than Sept 15-16 in my GMAT model, nor in SatEvo. I simply don't see it coming down from that altitude within a day, unless something drastic happened (e.g. a structural disintegration of the satellite before the actual reentry). For reference, it took KMS 4 some four-and-a-half days to come down from that altitude and reenter in June.

 

click diagram to enlarge

 

So it looks like either CSpOC lost the object and doesn't really know, or they do not want to to share the last whereabouts of KMS 3-2 (e.g. from a notion that this would help the North Koreans in some way).

Meanwhile, after two failed launch attempts the past months using a new carrier rocket, North Korea said it will attempt to loft a new satellite into orbit in October.

I recently talked to the website North Korea News about the upcoming KMS 3-2 reentry, which you can read here.

The reentry of a Soyuz rocket stage over southern Australia on August 7

click map to enlarge

 

On 7 August 2023 at 13:20 UTC, Russia launched the first of it's improved GLONASS-K2 navigation satellites from Plesetsk Cosmodrome. The launch employed a Soyuz 2.1b rocket with a Fregat upper stage. The payload and the Fregat upper stage were subsequently catalogued in 19156 x 19135 km resp 19182 x 19005 km, 64.8 degree inclined Medium Earth Orbits (MEO), as catalogue numbers 57517 and 57518.

Some 40 minutes after the launch, people from southern Australia and Tasmania were treated to a spectacular sight of a bright slow-moving, fragmenting fireball that crossed the sky. Many eyewitness video's were posted on social media and poicked up by the News media: for a few fine examples see here, here, here and  here. Immediate suspicions were raised that this was space debris.

Indeed, the fireball was the Soyuz 3rd stage reentering the atmosphere. A Navigational Warning for space debris connected to this launch had been published earlier (HYDROPAC 2502/23), for an area south of Australia and Tasmania:

021113Z AUG 23
HYDROPAC 2502/23(75,76).
TASMAN SEA.
WESTERN SOUTH PACIFIC.
TASMANIA.
DNC 05, DNC 06.
1. HAZARDOUS OPERATIONS, SPACE DEBRIS
   071300Z TO 071600Z AUG, ALTERNATE
   1300Z TO 1600Z DAILY 08 AND 09 AUG
   IN AREA WITHIN 35 MILES OF TRACKLINE JOINING
   43-10.00S 148-55.00E, 53-30.00S 163-20.00E.
2. CANCEL THIS MSG 091700Z AUG 23.

The time window matches well with the Australian reentry sighting. The area defined by the Navigational Warning matches a launch into a ~63 degree inclined parking orbit from Plesetsk:

click map to enlarge

 

The Soyuz 2.1b rocket consists of four side boosters ('stage 1'), a core stage ('stage 2') and a third stage. On top of that is (for this launch) the Fregat upper stage. The Soyuz rocket brings the Fregat upper stage and GLONASS payload in a low parking orbit. From there, a series of firings of the Fregat stage bring the payload to 19150 km Medium Earth Orbit (MEO). The Fregat upper stage is left on orbit, but the Soyuz stages deorbit downrange from the launch site: the last of these stages, is the stage that reentered over southern Australia about half a revolution after the launch.

GLONASS is the Russian equivalent of GPS.

[UPDATED] North Korea's KMS 4 has only a few days left on orbit

 

[post update nr 8, 4 July 2023  09:40 UTC]

Recently, on May 30, North Korea tried to launch a new satellite. The satellite however did not reach orbit, because the second stage of the launch vehicle failed (see my earlier post here).

There are currently two older North Korean satellites on orbit: KMS 3-2 (launched on 12 December 2012) and KMS 4 (launched on 7 February 2016). One of these, KMS 4, is now very close to reentry.

Kwangmyŏngsŏng 4 (KMS 4, 2016-009A) was launched on 7 February 2016, into a 501 x 465 km, 97.5 degree inclined sun-synchronous orbit. It is said to be an optical reconnaissance satellite, as also implied by its sun-synchronous orbit: but it was never clear whether it ever actually functioned after launch. 

A lot of noise was made in some parts of the US Media back in 2016 about it "tumbling out of control" through space, but optical observations did and do not show clear signs of tumbling.

Over the past 8.5 years, natural orbital decay has lowered the orbit of the satellite, to the point where it at the last available orbit  (orbit epoch 29 June 2023, 21:14 UTC) was down to a 149 x 168 km orbit. The video above was filmed by me on June 14, 2023, and shows the current fast angular movement due to the low orbital altitude.

The diagram below shows the orbital evolution since 1 January this year: it is was coming down fast. At the moment you are reading this, it likely only has a few hours on orbit left it has reentered already.

 

click diagram to enlarge

[Updated and editted 3 July 2023] My current "aftercast", based on a late published orbit with epoch 29 June 21:14 UTC, is reentry in the early hours of June 30, 2023. The nominal modelled reentry time is 30 June, 03:54 UTC ± 1.4 hr (note the large uncertainty interval, about one full orbit!).

The diagram below shows how the reentry forecasts for KMS 4 were developing with each orbit update.

click diagram to enlarge

 

With the current still widely uncertain "aftercast", with an uncertrainty of one full orbital revolution, the blue line on the map below is where it could have come down: 

 

click map to enlarge

Note that this reentry is NOT something to worry about. The reentry will have been harmless: KMS 4 is small, it is a box of about 1 x 0.65 meter with solar panels, weighing maybe 200 kg. It will likely have burned up completely upon atmospheric reentry. Yet because it is a North Korean satellite, the reentry might garner some interest.

[UPDATED] Curiously, CSpOC (18th SPSS) never published a TIP for KMS 4. Early on July 3, 2023, the catalogue listed it as if still on orbit, albeit with days old elements. Late on July 3, a reentry date (but no reentry time) of 30 June 2023 was entered into the catalogue.

Apparently 18 SPSS told Jonathan McDowell (see this tweet) that no TIP was issued because the satellite was "too small" to warrant one. Which is not very believable, as the same week they did issue a TIP for Skysat C18 (2020-057BR), which is of similar size as KMS 4, and TIP's have been issued for even smaller objects in the past....

We are not talking about a cubesat here, KMS 4 was about one meter in size, not counting deployed solar panels.

Perhaps - but this is my speculation only- they wanted to keep the reentry quiet because anything North Korean is accompanied by unbridled panic and paranoia around 'nefarious intentions' in the US. Some US media pundits, fueled in this idea by a US Government Task Force and a Congressional Hearing, have tried to push the narrative that the KMS satellites are EMP devices. Something which others have characterized as "grossly overstated" (see this debunk on 38North and further comments by others here in this Newsweek article) as North Korea does not seem to have nuclear weapons that are powerfull enough to cause serious EMP effects.

It will be interesting to see whether TIP's will appear for the other North Korean satellite, KMS 3-2, which I currently forecast to reenter somewhere in September-October 2023.

[The text of this post has been repeatedly updated. Post last updated: 4 July 2023, 09:40 UTC]

CZ-5B rocket stage 2022-085B: reentry forecasts

 (this post is periodically updated when new tracking data become available and have been analyzed)

last updated 30 July 19:05 UT   

My latest forecast is: 30 July, 17:10 UTC +- 15m

The latest CSpOC forecast was:  30 July, 16:49 UTC +- 10m 

The FINAL CSpOC TIP is 16:51 UTC +- 1m, near 3.4 N, 113 E

A possible reentry observation has been made from Kuching (Sarawak, Borneo) near 16:50 UTC 

Click diagram to enlarge

 

On 24 July 2022, China launched a CZ-5B rocket carrying the Wentian module to the Chinese Space Station. The huge CZ-5B core stage, 2022-085B, with a dry mass of 18 tons and a length of 33 meter, is now set for an uncontrolled reentry on July 30.

The diagram above shows the evolution of apogee and perigee altitudes so far.

I am running reentry models in the General Mission Analysis Tool (GMAT) for the CZ-5B stage, that currently point to reentry in the second half of July 30 UTC. At this point in time it is still too early to pinpoint a more precise time or location - and for this reason I do not give geographic coordinates for the nominal reentry point yet.

The map below shows the current risk area, based on the latest forecast and its uncertainty window. Within the uncertainty window, the rocket stage can come down anywhere on the blue line in the map. The yellow circle is the nominal reentry position, i.e. the center of the uncertainty interval, but not necessarily the most likely location.

 

Current risk area. Click map to enlarge

Movement of the spacecraft is from west to east = left to right on the map. Cities between 41.5 N and 41.5 S with 1 million or more inhabitants are shown as yellow dots.

Below is a diagram of the evolution of the reentry forecast so far. Below the diagram is the same data in tabular form, with the most recent forecast at the bottom of the table.

- My latest GMAT based forecast (black dots) is: 30 July, 17:10 UTC +- 15 m

- The last pre-reentry CSpOC TIP forecast (red dots) is: 30 July, 16:49 UTC +- 10 m

- The FINAL CSpOC TIP is 16:51 UTC +- 1 m, near 3.4 N, 113 E

The latter time, with its 1 minute uncertainty, and position are likely based on an infra-red detection of the reentry fireball by a SBIRS Early warning satellite. So it seems my last forecast was about 20 minutes off. It reentered more or less at the start of my uncertainty window.

Below is a map with the final revolution and actual reentry location:

 

click map to enlarge

EXPLANATION OF THE DIAGRAM BELOW:
Each dot is a prediction based on a new orbit release. On the x-axis is the date/time (in decimal days) of the orbit on which the forecast is based. On the y-axis is the predicted reentry time/date resulting from a GMAT model run using  this orbit. These too are given in  decimal days (e.g., "30.5" = "30 July, 12:00 UTC". Note: for dates beyond July 31, "July 32.0" means "August 1, 00:00 UTC" etc.). 

Error bars show the uncertainty in the prediction. Note how the uncertainty window become smaller once we get closer to the reentry.

The black dots are the results of my GMAT model runs. The red dots are predictions by CSpOC, the US Military tracking organization (and "the" authoritive source of orbital tracking data), given for comparison.

click diagrams to enlarge

My reentry modelling in GMAT uses the MSISE90 model atmosphere, current, predicted and past spaceweather, a dry mass of 18000 kg and a drag surface of 62% of the maximum value, which is the best estimate for the average drag surface of a tumbling rocket stage of these dimensions. The depicted uncertainty window is a fixed 20% of the timespan between orbit epoch and the reentry time forecast resulting from the orbit.

 

GMAT model forecasts:

ORBIT DATE/TIME    REENTRY         UNCERTAINTY   REMARKS/Lat Lon
(UTC)              FORECAST (UTC)  INTERVAL

 25-07 11:52:33    31-07 08:28  +-  28.1 h
 25-07 13:01:55    31-07 02:51  +-  26.8 h
 25-07 14:30:57    31-07 02:42  +-  26.4 h
 25-07 17:29:00    31-07 02:45  +-  25.9 h
 25-07 19:29:01    31-07 02:08  +-  25.3 h
 26-07 03:04:47    31-07 02:39  +-  23.9 h
 26-07 11:16:28    31-07 03:40  +-  22.5 h
 26-07 13:11:03    31-07 02:22  +-  21.8 h
 26-07 15:43:05    31-07 02:34  +-  21.4 h
 26-07 18:40:46    31-07 02:44  +-  20.8 h
 26-07 19:12:28    31-07 01:04  +-  20.4 h
 27-07 14:41:30    30-07 21:29  +-  15.8 h
 27-07 15:55:00    30-07 20:57  +-  15.4 h
 27-07 17:18:31    30-07 21:01  +-  15.1 h 
 27-07 18:20:32    30-07 23:39  +-  15.5 h
 28-07 17:05:39    30-07 23:46  +-  10.9 h
 29-07 01:18:22    30-07 21:36  +-   8.9 h
 29-07 08:55:31    30-07 20:12  +-   7.1 h  
 29-07 13:03:52    30-07 19:20  +-   6.1 h   
 29-07 16:34:29    30-07 18:58  +-   5.3 h  
 29-07 18:00:07    30-07 19:00  +-   5.0 h  
 30-07 01:31:37    30-07 20:06  +-   3.7 h 
 30-07 08:22:22    30-07 17:50  +-   1.9 h  
 30-07 09:54:18    30-07 17:29  +-   1.5 h  
 30-07 12:28:44    30-07 17:28  +-   1.0 h 
 30-07 12:58:01    30-07 17:30  +-    54 m  
 30-07 14:17:27    30-07 17:20  +-    37 m   33 N 126 W  
 30-07 15:48:42    30-07 17:09  +-    16 m   40 N 178 E
 30-07 15:57:24    30-07 17:10  +-    15 m   41 N 173 W  latest

 

UPDATE 30 July 17:15 UT:

Footage purportedly shot today from Kuching (Sarawak, Borneo) appears to show the reentry of 2022-085B. It would place reentry near ~16:50 UTC. The footage is here, in a tweet by Nazri Suleiman.
I say 'purported' because for the moment the footage has not yet been verified. But it seems likely it is the reentry indeed.

 

UPDATE 30 July 19:05 UT:

CSpOC gives a final TIP of 16:51 +- 1m UTC near 3.4 N, 113 E, which matches the Kuching sightings well (multiple video's from the Kuching region have since turned up on social media, so it is now quite certain that this was the reentry).

The map below gives the final orbital revolution and location of reentry over Borneo.

 

click map to enlarge

 

Other forecast sources:

CSpOC  (https://www.space-track.org/)

Aerospace Corporation  (https://aerospace.org/reentries/cz-5b-rb-id-53240)

Josep Remis  (https://twitter.com/jremis)

ESA  (https://reentry.esoc.esa.int/reentry) 

EU-SST (https://twitter.com/EU_SST) 

 

Added note on 'Live' satellite tracking websites

BEWARE: contrary to what many people think, purported 'Live' satellite tracking websites such as n2yo.com are NOT showing you satellite locations based on 'live' tracking data!

In reality, they show predictions based on older orbital element sets, that can be hours old actually.

This can be deceptive. They sometimes happily show an object apparently still "on orbit" after it in reality already reentered!