Walter Nissen wrote: >> Now the question that goes begging is: why did Trakstar generate=20 >> a pass for 0325 UT while Quicksat did not? My hunch is that it has=20 >> to do w=F9th the fact that Quicksat ignores B* while Trakstar does not.= =20 >> Would like your thoughts or anyone else's on this.=20 >=20 >I'm sure there are people who read SeeSat-L who are far better able to=20 >answer this question than I. I would guess that many good programs would= =20 >generate similar output, given the same input. =20 Yes, but Trakstar and Quicksat didn't. This inspired me to take a close look at the formulae in Spacetrack Report No. 3, which contains the algo- rithms for SGP, SGP4 and SGP8. Quicksat must be using SGP because the SGP method ignores B* and uses the mean motion terms. Spacetrack says "NORAD element sets are currently generated with either SGP4 or SDP4... For element sets sent to external users, the value of mean motion is altered slightly and a pseudo-drag term (n-dot/2) is generated. These changes allow an SGP user to make compatible predictions [because] the pseudo-drag term used in SGP simulates the drag effect of the SGP4 model." But the SGP4 model is simplified for low perigees. Spacetrack says "It should be noted that when epoch perigee height is less than 220 kilometers, the equations for a and IL are truncated... and [other] terms are dropped..." The SGP model makes no such adjustments for perigee height. >the acceleration is not applied continuously, but only at perigee,=20 >more or less as a step function That's right. If you look at the SGP formula for semi-major axis, for example, it simply multiplies n-dot and n-dot2 times [t - t0] and [t - t0]2 respectively. Like you say, there is no way n-dot and n-dot2 can keep up with with the rapid change in MM at perigee. SGP8, like SGP4, ignores the pseudo-drag terms and uses B*. It does not truncate terms or make adjustments for low perigee heights. It appears at first glance that maybe SGP8 will handle high e / low perigee objects a little better than SGP or SGP4. The conclusion I come to after reading Spacetrack is that no prediction program based on SGP or SGP4 can handle high e / low p objects well. I'm not surprised now that Trakstar and Quicksat diverged.=20 Are there any programs out there that implement SGP8? I'd be interested in playing with it. >> "Celestial mechanics is entitled to be regarded as the most=20 >> perfect science"=20 >> Forest Ray Moulton=20 >=20 >More perfect than number theory with its gems such as Euclid's=20 >finite proof of the infinitude of the primes?=20 Moulton was saying that no other science has theory match physical observations so precisely. He wrote that circa 1900 so he had no idea what a mess decaying LEO sats make :) -- Jim ----------------------------------------------------------------------------= -- Jim Varney | 121^ 23' 54" W, 38^ 27' 28" N | Sacramento,= CA Member, SeeSat-L| Elev. 31 ft. = |jvarney@mail2.quiknet.com ----------------------------------------------------------------------------= --