More accurate look angles

[mpielikis]

There were some inaccuracies in time conversions and angle calculations that leaded in loosing several seconds on AOS and LOS times for ground station. Some of these changes breaks backward compatibility.

As a reference I used gPredict, pyorbital and sgp4

[joshuaferrara]

As a reference I used gPredict, pyorbital and sgp4

Some of the modified methods have citations in the comments prior to them, for example:

// Reference: The 1992 Astronomical Almanac, page K11.
func LLAToECI(obsCoords LatLong, alt, jday float64, gravConst GravConst) (eciObs Vector3) {

I think it'd be wise to update these comments to include where the algorithm modifications were ported from.

[joshuaferrara]

Please include citations alongside the methods that were modified to indicate where the changes were ported from.