All-sky sat monitoring

From: Greg Roberts via Seesat-l <>
Date: Sun, 08 Feb 2015 09:41:51 +0200
Finally got an idle moment to say something about this.

Of course the set-up used by Cees, Leo and Scott are basically already 
doing this but with narrower field instruments but the principle is the 
same - change the focal length of the camera lens and you get a larger 
field and as can be seen from the long lists of observations from these 
gentlemen the system is very productive and quick , but unfortunately 
limited to Linux systems and apparently not that straight forward to 
install the software developed by Cees, so a steep double learning curve 
for those not familiar with Linux.

I see no point in setting up an all-sky camera for purely satellite 
work. Because of the relatively low magnitude limit and the image scale 
of an all-sky camera it is only sort of effective during a short period 
after sunset and before sunrise when we have low altitude, and thus 
bright, satellites. Once the earths shadow takes over and removes these 
low objects one is only left with higher altitude objects, in the 
majority of cases too faint for an all-sky system,so one will get a very 
low return.

To me it would be far more practical and useful to use a monitoring 
system covering say 50-60 degrees wide and monitoring say about 20 
degrees ahead of the shadow entry curve for satellites at say a maximum 
altitude of around 800 kms.   Of course this shadow entry curve will 
depend on season and location but at least one will be aiming for the 
area where best to see satellites because of the more favourable phase 
angle close to shadow entry. In the winter months it could well turn out 
there will be little or no return after say 2-2.5 hours after sunset, 
(or before sunset) - saving thus on equipment usage and wear and also 
saving on the observers time in not having to go through images  over 
possibly hours of no observations at all.    It should be relatively 
easy to compute these pre-eclipse visibility contours - for example this 
has already been done in tabular form in SIMPLIFIED SATELLITE PREDICTION 
USING MODIFIED ORBITAL ELEMENTS issued in the early 60's and available 
on the Internet  and  a graphical plot of shadow entry  appeared in SKY 
AND TELESCOPE about a decade later.

Next comes the question of the sensor.   Comparing DSLR images against 
images secured with a video or CCD detector there is an obvious 
difference against a DSLR - especially if in colour - which generally 
has less sensitivity than a mono system.   Besides a video detector 
system will be cheaper than a reasonably good DSLR camera and experience 
far less "mechanical/physical" wear.    It is a relatively simple matter 
nowadays to record video to a computer/laptop hard drive and hard drives 
are cheap so storage of a sessions observations is very practical - to 
my mind its easier to go through a sessions video observations than 
examine hundreds of DSLR images.   It is also possible to time stamp 
video observations -  my Nikon camera has to rely on its internal time 
keeper which frankly is pretty putrid and, as far as I am aware there is 
nothing I can do about it!

Even a pc ( again in my particular case ) handles time accuracy better 
than my Nikon camera,   and even if one does not have a GPS time 
stamping device one can use the PC clock connected to one of the several 
Internet time systems available ( I do not know the precise details as 
my particular PC runs directly from a GPS receiver) but I am sure 
several SeeSat readers can elaborate on their means of accurate time, so 
at least 0.1 second accuracy is possible. This should be regarded as the 
minimum time accuracy for objects in leo orbits.

Finally comes the question of image scale. As pointed out an all-sky 
camera with a fish-eye lens has an essentially useless image scale for 
any sort of accurate satellite positional work , so one wants an image 
scale where a timing accuracy of 0.1 sec minimum is meaningful.  With a 
field of view of 50-60 degrees one may be pushing the limit here but its 
certainly better than a fish-eye system,   and with some astrometric 
measuring programs available one can get sub pixel accuracy.    Besides 
,   with a video system being cheaper than a good DSLR camera one could 
set up two or three video cameras with slightly overlapping fields ( but 
this produces complications in having to record three streams of data 

Anyway enough said - I am sure others will add their views.

Seesat-l mailing list
Received on Sun Feb 08 2015 - 01:43:04 UTC

This archive was generated by hypermail 2.3.0 : Sun Feb 08 2015 - 07:43:04 UTC