The other night, I observed what I believe must have been a satellite, however it was at 1:30am so I was not sure that this would be possible. Being in the Southern Hemisphere we are just past Summer Solstice, so theoretically it could have been reflecting the sun's light.

There were no flashing lights, just a star-like point of light. As it slowly traveled in a NE direction across the sky, it began to gradually fade as typical of a satellite passing into Earth's shadow.

Based on my observation, I'd like to get an understanding of the approximate minimum altitude of the object that I observed, allowing for it to be out of the shadow of the Earth sunlight. I'm not a mathematician, but using Space Engine with the correct time & location, I had to position the camera to some 7000km in altitude before the sun was visible over the horizon. The object was N to NE from where I observed it, whist the midnight sun would have been shining away at the South Pole.

Date/Time: 27-Dec-2017, ~01:30 local time (GMT + 8) Location of observation: Perth, Western Australia (~32S, 116E) Position in sky : ~ 80 to 60 degrees elevation, moving north due northeast Duration of observation : ~2 to 3 minutes


1 Answer 1


I'll venture a guess that the 7000 kilometers (your software's camera position) was with respect to the center of the Earth. Using 6371 km as a value for the Earth's radius, that would make the altitude only 629 km, which it turns out is quite a popular altitude for satellites in LEO.

For the simplifying case of midnight, one can draw a simple diagram. I get

$$h = R_{\textrm{Earth}} \left(\frac{1}{\cos(35°)} - 1\right)$$

which is about 1400 km at midnight, and the height will be lower before and after.

I'll wait to hear about the camera "height" and try to find another tool for you to use. Do you ever use Python programming language?

Red dots are Perth at local noon and midnight.

enter image description here

  • $\begingroup$ Thanks, I appreciate the response!I cannot vouch for the accuracy of my coordinate accuracy except to say when I ascend to the required altitude I position the camera as tangential to the horizon as possible (i.e. directly to the ground) before moving the camera backwards. $\endgroup$
    – razeezar
    Jan 7, 2018 at 9:28
  • 1
    $\begingroup$ Altitude in the software is indeed ~7,500km above sea level. Below is a link to a screenshot to demonstrate. In hindsight I should have referred to lat & lon in the HUD to get a more accurate positon. Note that the camera has been angled after ascent to include both the sun on the horizon as well as Perth city night lights; FOV has also been adjusted to fit both in the shot: link $\endgroup$
    – razeezar
    Jan 7, 2018 at 9:39
  • $\begingroup$ @razeezar that looks great! i.sstatic.net/myVkH.jpg You can consider posting a second, supplemental answer and include it the screen shot, confirming what you've tested. $\endgroup$
    – uhoh
    Jan 7, 2018 at 10:01
  • $\begingroup$ Thanks uhoh. I have also linked the following screenshot using in-the-sky.org link; I have selected a moment of 1:16am; This is around the time when satellites were in the same vicinity of the sky travelling in a similar direction as what I saw (i.e. three FLOCK satellites), except these should have all been invisible as they were at too low an altitude to be illuminated at that time of night. I did not notice any other illuminated moving objects at the time that I observed the object ~2 mins it as moved, then began to fade. $\endgroup$
    – razeezar
    Jan 7, 2018 at 14:46
  • 1
    $\begingroup$ I have only just noticed that Space Engine automatically changes distance from sea level to the center of the planet at =>1000km above sea level so you are correct! Therefore the altitude at which sun becomes visible is ~1,130km, which seems to check out Mathematically for ~1:30am. (btw I know a little C/++ but not Python). Based on available satellite data I am yet to identify any satellites that were anywhere near this altitude at this time. $\endgroup$
    – razeezar
    Jan 7, 2018 at 16:01

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .