Last night, Greg Roberts, of Pinelands, Cape Town, South Africa, was a very busy chap!

Greg obtained a series of excellent images of NASA’s Juno spacecraft just before it made closest approach (CA) during Earth flyby; CA came at 19:21 UTC.

The Jupiter probe whizzed past Earth at just 561km – lower than many Earth observation satellites and low enough to experience a tiny drag due to atmospheric friction.

Greg’s images were acquired in the few minutes before CA. After sending them off to NASA, he kindly agreed to share his photos with the Rocket Science blog, and we’re delighted to present them here.

Greg wrote:

I got 23 images in all – starting at 19h12m04s UT to 19H12m28s UT from between a gap in the house roof; range was about 6750kms.

I saw Juno until shortly before shadow entry – point of closest approach to Earth happened about 2 minutes later so it was still ‘coming in’ and invisible
from Southern Africa.

Then got 17 images or so from 19h16m19s to 19h17m54s UT – then equipment misbehaved – I was tracking on predictions so satellite appeared stellar! After the equipment tracking failure, I managed to get a picture as it went into shadow – time of picture start 19h18m19.553sUT.

Equipment failure? Due to the very rapid movement of the satellite as it approached culmination and the rapid change in azimuth, the computer program failed because of insufficient time resolution and the tracking became non-linear.

Equipment: 135mm focal length f/2.8 with FLI8300M CCD camera, exposure 5 seconds in all cases

I also got a video for several minutes – not that great  but could see the satellite and that it was variable. Pity I had equipment failure from 19h17.54s but was lucky to get to the correct position as shadow entry took place. Cheers!

 Greg Roberts

Thank you, Greg, for sharing!

Click on ‘Continue Reading’ for full technical details on Greg’s images.

Images acquired by Greg Roberts, of Pinelands, Cape Town, South Africa.

Greg obtained a series of excellent images of NASA’s Juno spacecraft as it made closest approach (CA) during Earth flyby; CA came at 19:21 UTC.

Juno flyby was tracked by ESA ground stations in Argentina and Australia. Read more via: <a href=”” rel=”nofollow”></a>



As observed by Greg Roberts, Pinelands, Cape Town, South Africa
Longitude 18.5125 east
Latitude  33.9406 south
altitude  10 metres
GPS position

All images 5.00-second auto sequence with 1.0 sec between exposures
Time of each frame is read by the camera shutter from GPS receiver.

Camera FLI 8300M CCD camera running at -25 deg C.
Optics:  135mm focal length f/2.8 old TAMRON lens
Homemade computer tracking setup that has now been in used for approximately 14 years using computer controlled stepper motors and program originally written by Willie Koorts (SAAO)and major changes made by Mike McCants (Texas USA) about 2 years ago.


Observed between small gap in eves/guttering of roof

A25   19H 12M 04.092S UT
A26   19H 12M 10.147S
A27   19H 12M 16.170S
A28   19H 12M 22.368S
A29   19H 12M 28.407S

Observed once clear of roof of house:

From J06 up to and including J27 times are END of exposure

J06   19H 16M 11.665S
J07   19H 16M 17.690S
J08   19H 16M 23.734S
J09   19H 16M 29.778S
J10   19H 16M 35.819S
J11   19H 16M 41.862S
J12   19H 16M 47.890S
J13   19H 16M 53.931S
J14   19H 16M 59.971S
J15   19H 17M 06.013S
J16   19H 17M 12.049S
J17   19H 17M 18.092S
J18   19H 17M 24.136S
J19   19H 17M 30.179S
J20   19H 17M 36.222S
J21   19H 17M 42.252S
J22   19H 17M 48.300S
J23   19H 17M 54.345S
J25   19H 18M 06.345S
J26   19H 18M 12.483S


J28   19H 18M 19.553S  START EXPOSURE
J29   19H 18M 25.597S  END EXPOSURE

FRAMES J24,J25,J26,J27 now included. J24 and J25 show trail of a satellite moving downwards.

NOTE: New exposure added J29. I though J28 was last image obtained but J29 shows more trail at edge. This demonstrates that it was NOT shadow entry and the satellite was quite variable.

I have enlarged J28 and J29 to better show the track of JUNO.

Unfortunately no data other than above. The automatic tracking failed as a result of inadequate prediction points as the angular velocity increased. The tracking alogorithm uses 5 points to compute a LINEAR track in between prediction points. As a result of the rapidly decreasing range the linearity failed and the program could not track on the satellite. When the track was aborted when I saw the drift-off becoming objectionable the program crashed. By the time I had the program up and running again the satellite was already in shadow.

Greg Roberts