On Earth is possible to define the position of something using WGS, and the same system is used for navigation. What is used in space instead? Are different systems used for Earth-orbiting objects (the ISS for example) as opposed to objects beyond LEO (Voyager 1/2) due to practical reasons?

  • 5
    $\begingroup$ This is a very broad question. To solve practical problems in space exploration, it is necessary to use many reference frames: fixed and rotating, with many points as origin - Earth, other bodies, Solar System Barycenter (or other barycenters), the spacecraft, ground station, etc. Coordinates may be expressed as spherical, geodetical, cartesian. You may start reading from this article. $\endgroup$ Aug 4, 2013 at 13:30

1 Answer 1


While coordinate systems are used for all sorts of things in space flight, from your position in the solar system down to individual instruments, I'll assume for the purposes of this question that you're talking specifically about coordinate systems used for determining satellite position relative to something.

Also, I'm going to assume you're really interested in reference frames, instead of coordinate systems. For the reference frames I'll talk about, Cartesian coordinates are typically used (Keplerian elements are, of course, another popular alternative).

Earth-orbiting satellites

There are two main types of reference frames: inertial and rotating. Specifically, when they're Earth-centered (origin at the center of the Earth, as they nearly always are, except when dealing with measurements from the Earth's surface), the two main classes are Earth-centered inertial (ECI), and Earth-centered, Earth-fixed (ECEF). ECI frames have axes defined by some inertial reference, such as pointing one of the axes at a specific star or constellation. The Earth rotates in these axes (i.e. the ECI axes do not correspond to any fixed location on Earth's surface). Alternatively, ECEF frames rotate with the Earth.

There are many variations on these two frames, and most of them have to do with how they account for the precession and nutation of the Earth.


This is a much more broad category, but generally they tend to be barycentric frames. You could define a reference frame with the origin at the barycenter, or center of mass, of several bodies (usually two, e.g. Earth-Sun or Earth-Moon).

The International Celestial Reference Frame (ICRF) is valid throughout the solar system, and has its origin at the barycenter of the solar system.

Of course, if you have a mission to, say, Mars, you would define a Mars-centric reference frame to work in.

  • $\begingroup$ What distinction are you drawing between reference frame and coordinate system? $\endgroup$ Apr 26, 2021 at 18:49
  • $\begingroup$ @Acccumulation Coordinate system transforms cannot change the magnitude or direction of a vector, only the numbers assigned to it. Frame transforms can, for they can change equations of motion. For example, they can add rotating-frame effects to velocities and accelerations. Coordinate systems are really just nothing more than bookkeeping, but picking the wrong frame for your problem can be quite the pain! $\endgroup$
    – Cort Ammon
    Apr 27, 2021 at 0:20
  • $\begingroup$ @CortAmmon What do you mean by "magnitude" and "direction"? Frame transformations can't change proper time. $\endgroup$ Apr 27, 2021 at 4:40
  • $\begingroup$ @Acccumulation Consider an object in geosynchronous orbit around the Earth. Its velocity vector in the ECEF frame is 0. Its velocity vector in an ECI frame is a non-zero vector, whose speed is the orbital speed of the craft. Its acceleration vector in the ECEF frame is 0, while its acceleration vector in an ECI frame is a vector pointing at the center of the Earth, with a magnitude that is the acceleration of gravity. $\endgroup$
    – Cort Ammon
    Apr 27, 2021 at 15:09
  • $\begingroup$ These are not simply scaling and/or offsetting of numbers in a coordinate vector (an [x, y, z] triple). The vectors themselves actually change when one does a frame transformation on them. If it were merely a coordinate transform, all I would need to know about a vector is its value as a coordinate vector [x, y, z] and a function which transforms [x', y', z'] = f([x, y, z]). In a frame transform, however, I also need to know what the vector was describing, as velocities may transform differently than accelerations. $\endgroup$
    – Cort Ammon
    Apr 27, 2021 at 15:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy