Where can I find details regarding the SGP4 orbit propagator (satellite function of Aerospace Toolbox)?

Related help page: https://www.mathworks.com/help/aerotbx/ug/satellitescenario.satellite.html#namevaluepairarguments

What assumptions does SGP4 propagator make for spacecraft ballistic coefficient?

What other perturbations are included or excluded? What assumptions are made for those?

Answer 1

[Youssef Noureddine]

Hi Eric,

The SGP4 Propagator is used in conjunction with two-line element sets produced by NORAD and NASA. It essentially models the motion of your current spacecraft based on other TLEs. I'll leave out the explanation of how exactly it works as it is quite complex, but generally because of the way the positions are estimated from other TLEs, you can expect to see errors on the order of approximately 1 kilometer at epoch which grows by 1-3 kilometers every day. So although I can't list out all the assumptions made, you can sort of get some insights on the degree of accuracy to expect from this propagator.

If you would like to see the exact underlying equations, I would recommend taking a look at some popular SGP4 propagators in the MATLAB file exchange and/or some reports such as the 'Spacetrack Report No. 3' where the equations were originally introduced.

Sincerely,

Youssef

[Eric Lund]

Thank you. I am familiar with the equations and math. I am asking for the assumptions Mathworks made when implementing the equations. SPG4 accounts for atmospheric drag, but that is not equal for all spacecraft. There is a drag term built into a TLE, but what does Mathworks assume for that if I under Keplarian elements and select the SPG4 orbit propagator? Is drag zeroed out when the user inputs Keplarian orbit parameters?

[Harish Saranathan]

The drag term BStar is assumed to be 0 when adding satellites using Keplerian elements. For SGP4 propagator, the keplerian elements are assumed to be the mean elements. The other elements that you see in a TLE file (such as mean motion) are deduced from these elements. The epoch is assumed to be satelliteScenario StartTime. The first and second time derivatives of mean motion are unused by the SGP4 algorithm.

In essence, adding a satellite using Keplerian elements is equivalent to adding a satellite using a TLE file with BStar set to 0.

When adding a satellite using a TLE file, the MATLAB algorithm does not make any additional assumptions outside of those already made by the SGP4 algorithm.

Sincerely,

Harish Saranathan

[Kurt]

Some TLEs include a negative BSTAR value. What does that mean?

It appears that function tleread() blows up when it encounters a negative BSTAR value. For example, this TLE is acceptable:

1 38480U 00000A   22125.33472451  .00182152  00000-0  09453-0 0    10
2 38480  73.8231 342.6630 0019834 309.3639  50.5805 15.30761709681550

While this one throws an exception:

1 38480U 00000A   22125.33472451  .00182152  00000-0  -9453-0 0    10
2 38480  73.8231 342.6630 0019834 309.3639  50.5805 15.30761709681550
spacecraft:tleread:InvalidTLEFile

Apparently the minus sign MUST be in column 54. This may require leading zeros. For example, this is

valid:

1 38480U 00000A   22125.33472451  .00182152  00000-0 -09453-0 0    10
2 38480  73.8231 342.6630 0019834 309.3639  50.5805 15.30761709681550

I have found that some TLE generators do not insert leading zeros in the BSTAR term, and can blow up on you.