Mon, 22 Jan 1996 12:51:50 +0100 (Matthew France) wrote:

>> I was testing with a file of about 240 objects, and it found
>> about 130 "events" (minimum separation <50km) in a single day(*)
>> 7 of those were less than 10km.

Could you augment your program to use different limits in altitude
and along-track ? A small collision-avoidance mainly makes
a small change in period, and a prediction uncertainty is usually
caused by a bad estimate of the drag term. In both cases, the
effect on the vertical position of the orbits, and their relative
longitudinal position in space, is much smaller than the effect of
reaching the orbital cross-over a few seconds earlier or later.

As a result, NASA uses different safety limits in two (three?)
directions, and I think any COLA program should allow this
(even estimate them automatically?), to avoid listing those
that cannot possibly collide because of a large vertical
separation, yet find all that may come close because of
prediction errors.

Will you publish your source ? I guess it would easily convert to DOS
and/or Windows. I program in all these, but I have Unix available only
at work, where I don't want to run my satellite computations.

 --    +46-8-7428086  (home)  --
 --                +46-8-59095739 (office)--
 --(      59.22371 N, 18.22857 E --