At 11:09 28/07/98 +0000, Jason Hatton wrote: >Hi All, > I was wondering what the practical limits of the prediction algorithms >used in tracking software are for satellites in "deep space" orbits. I >know that SDP4 works quite well for satellites in geosynchronous orbits >(ie. 36,000km alt) & in geosync. transfer (ie 600km x 36,000km), but at >what altitude do solar / lunar perturbations become too significant for >reliable predictions to be made? There are some satellites, such as the >Russian "Granat" & "Astron" astrophysical satellites which are in high >eccentric orbits with 200,000km apogees, but perigee's of only a few >thousand km's, so they might be visible in a telescope near perigee. To reply to both Jason & Harald in one reply. Jason, the observations of objects like Granat & astron are obviously made in the lower part of their orbits, [mostly by the GEODSS stations I imagine] where the sdp4 algorithm is meant to be used, so I think the elements would be reasonable. One way is of course to check out the predictions for one element set against the next. If one is only considering a small number of satellites, a methos which does a full numerical integration of the equations of motion is bound to be accurate. Remember SPACECOM uses all these algorithms because its handling observations & predictions for some 8000 objects every day, and needs all the help it can get to handle the calculations. Harald, There are lurkers on this list from the JPL navigation teams, like George Lewis, who could answer much better than I. However I think you will find that they maintain a spacecfradft ephemeris using direct numerical integration of the equations of motion, allowing for the accelarations of object by inner and outer planets and whatever moons are appropiate. Since they are not using a keplerian model your question has no meaning. That doesnt mean they might for practical reasons have written the programme so that for example for galileo in orbit around jupiter , Eart & moon are lumped together. Tony Beresford