Improving elsets ; Was:Re: Norad 23907 MIA

bjorn.gimle@saf.se
Tue, 27 May 1997 11:48:35 +0100

     To: SeeSat-L
     From: b_gimle@algonet.se (Bjoern Gimle)
     Subject: Re: Norad 23907 MIA
    =20
    =20
     Craig Cholar    3432P@VM1.CC.NPS.NAVY.MIL wrote:
    =20
     >Observed #23907, Noss 2-3 Rk again, running 4m 8s late;  It was=20
     predicted
     >to pass 6.5 deg above Alphard at 04:15:22 26May UTC according to=20
     Skymap
     >(SGP4), but passed about 3.5 deg above Alphard at 04:19:30 instead.
     >
     Russell Eberst posted on May 24 an elset by Pierre Neirinck.
     Here it is, in more normal form:
     NOSS 2-3 r       9.8  3.0  0.0  4.8              549 x 312
     1 23907U 96029  B 97132.88974560  .00007031  00000-0  70000-4 0    01
     2 23907  63.4002  69.3795 0173760 151.8282 209.1228 15.45266993    02
    =20
     >Have the observations over the past few days, taken together,=20
     provided
     >enough information for a new elset to be created, or do the=20
     observations
     >need to be more precise?  I've never tried to modify elements to make=
=20
     them
     >fit observations, so excuse my ignorance on this subject.
     >
     For the purpose of improving the time of prediction, to reduce the=20
     risk
     of missing it, a single obs like yours is sufficient to adjust ndot2
     (or MM). Two of those spaced a few days would determine both.
     More accurate observations are required to adjust all other elements.
    =20
     You can easily attempt the first kind of adjustment with SkyMap:
     Load the old elset twice on the F7/F8 menu (delete other objects)
     select SGP on the F7 menu, and save the the .cfg file with a new name.
     Edit the .cfg file; change one of the elset copies to have USSPACECOM
     e.g. 43907 instead of 23907, to differentiate the predictions better.
    =20
     NORAD( 1)=3D23907
     1 23907U 96029  B 97114.89739697 0.00013000  00000-0  13893-3 0    01=20
     2 23907  63.4294 133.2060 0186000 156.1079 203.8921 15.44968413    00=20
    =20
     The observed delay of 4.1 minutes is about .044 orbits.
     The elset age at the observation is 146.18-114.89 =3D 31.3 days
     You will have to reduce ndot2 from 0.00013000 to achieve a delay;
     a 0.00001 decrease results in about 0.00001*31.3*31.3=3D0.01 orbits.
     This suggests a reduction to .000130-.000044 =3D .000086
     (The corresponding value of B* for SGP4 is about .13893E-3 *=20
     .000086/.000130 =3D .09191E-3 =3D 91910-4 in elset notation )
    =20
     SatName( 2)=3DNOSS 2-3 r adj.
     NORAD( 2)=3D43907
     1 43907U 96029  B 97114.89739697 0.00008600  00000-0  91910-4 0    01=20
     2 43907  63.4294 133.2060 0186000 156.1079 203.8921 15.44968413    00=20
     SatJD( 2)=3D2450594.678998
     SatDur( 2)=3D    6.54792
     Autolocate( 2)=3D242.9964, 30.9166
    =20
     In SkyMap, just reload the same (edited) .cfg file and press F10
     to see the effect of the changes.
     If you are not running Windows (or not using SkyMap ) and have to
     exit + restart the program for each try, you can just as easily
     modify an elset file the same way.
    =20
     Run a new prediction and verify that the change is appropriate
     (and switch back to SGP4 to se the corresponding change)
     The suggestions above seem to give the 4 m 08 s you mentioned.
     I get the appulse time as 04:19:34, 2.9 degrees above Alphard=20
     (alpha Hya)
    =20
     Pierre's elset gives a pass 28 seconds later (0.5 degrees closer)
    =20