We often say that the planets orbit the Sun, which is usually a reasonable approximation. But in reality both Sun and the planets orbit the center-of-mass/center-of-gravity of the whole solar system, not the center of the sun.

The Sun is by far the most massive body in the solar system, so it's plausible that the center-of-gravity lies within it. On the other hand Jupiter is rather massive as well and pretty far from the Sun. Thus the center-of-mass might not actually be inside the Sun, only near it. The influence of Saturn might not be negligible either.

  • 4
    Of course, in the N-body problem (with N>2), " orbits" are an approximation. The solar system apparently is quite stable; Earth is unlikely to be ejected in our lifetimes. – MSalters Jun 5 '15 at 10:36
  • Yes - the planets really do orbit the sun - in accordance with the law of universal attraction (which means they each exert a force on each other which can distort a circular orbit.) – Tracy Cramer Jun 5 '15 at 21:34
  • Since the Earth and other smaller planets are orbiting a barycenter which is moving about somewhat chaotically, wouldn't that have a tendency to destabilize their orbits, or induce significant long-term changes, or make them difficult/impossible to model accurately? Or have the orbits all found stable "homes"? – Anthony X Jun 6 '15 at 22:49
  • 1
    I think that the assumptions in the question or the top-voted answer are quite misleading. See this question on physics.SE: physics.stackexchange.com/questions/188650/… – Cedric H. Jun 10 '15 at 12:35
up vote 127 down vote accepted

You are correct, the centre of the Sun is not the Solar System's centre of gravity.

Solar system's COG changing over time

A diagram (courtesy Wikimedia Commons), showing how the barycentre of the Solar System has changed over time.

The Sun is affected by the gravity of all planets in the Solar System, but you are right, it is most affected by the two most massive ones; Jupiter and Saturn. You can see in this animation (a representation, not a simulation), how two bodies affect each other in a normal orbit like that between the Earth and the Sun:

Orbit diagram showing oscillation of the Sun.

Courtesy Wikimedia

Apply this relation to every body in the Solar System (adjusting for mass and distance of course) and you get a rough idea of how the Sun is affected by the rest of the Solar System.

Due to this somewhat chaotic dance the centre of mass of the Solar System continuously moves around, sometimes under the Sun's surface and sometimes outside it. The further away from the centre of the Sun this barycentre is, the more the Sun appears to oscillate.


Another example of this phenomenon is the Pluto-Charon system:

Pluto-Charon System

Courtesy Wikimedia.

Charon is about one-tenth the mass of Pluto (thank you for the correction @Hobbes), and yet exerts a significant gravitational pull on Pluto. Hence they both orbit around the centre-of-gravity of their system, well outside the surface of Pluto.


Bonus: We use this phenomenon to find planets outside the Solar System! If a distant star is observed to 'wobble' or oscillate about it's mean position, we can use that data to infer the presence of one or more exoplanets, and calculate their mass.

Further Reading:

http://homepages.wmich.edu/~korista/solarsystem_barycenter.pdf http://spaceplace.nasa.gov/barycenter/en/

  • 13
    That first diagram, of how the barycenter moves relative to the sun, is pretty cool. While I read that it's sometimes inside and sometimes outside the sun, I didn't expect it to vary that much. – CodesInChaos Jun 5 '15 at 7:54
  • I can't verify the scale of the diagram @CodesInChaos, but the variation is quite significant :) – Vedant Chandra Jun 5 '15 at 8:14
  • Can the wobbling of a distant star really be measured directly? At the multiple light years distances we are talking about, I'd expect the movement across the sky to be too small a fraction of a degree to be measurable. I recall being told that the wobbling is measured indirectly through the Doppler shift. – kasperd Jun 6 '15 at 9:33
  • @kasperd Exactly, it's still a rough measurement though. – Vedant Chandra Jun 6 '15 at 10:32
  • 4
    Solar System Live shows where the planets were on a given date. On June 1 1990 all the outer planets except Jupiter were within 15 degrees of each other on one side of the sun, with Jupiter directly across from them. Result: the barycenter was almost dead center inside the sun on that date. – kim holder Jun 6 '15 at 15:26

Using the approximation

$$\Delta \mathrm{center}=\frac{m_p}{m_\odot}\cdot\frac{\mathrm{dist}_p}{r_\odot}$$

and data from List of gravitationally rounded objects of the Solar System - Wikipedia:

Name       Distance   Mass/[kg]  Mass       Distance   ΔCenter
           /km        /kg        /sun mass  /sun radius
--------------------------------------------------------------
Sun        0.00E+000  1.99E+030  1.00E+000  0.00E+000  0.00000
Mercury    5.79E+007  3.30E+023  1.66E-007  8.32E+001  0.00001
Venus      1.08E+008  4.87E+024  2.45E-006  1.55E+002  0.00038
Earth      1.50E+008  5.97E+024  3.00E-006  2.15E+002  0.00065
Mars       2.28E+008  6.42E+023  3.23E-007  3.27E+002  0.00011
Jupiter    7.78E+008  1.90E+027  9.55E-004  1.12E+003  1.06735
Saturn     1.43E+009  5.69E+026  2.86E-004  2.05E+003  0.58576
Uranus     2.87E+009  8.68E+025  4.37E-005  4.12E+003  0.18007
Neptune    4.50E+009  1.02E+026  5.15E-005  6.46E+003  0.33278

Which sums to a shift of 2.17 sun radii if all planets were aligned. Jupiter clearly dominates (about half of the total effect) and would be enough to move the center-of-gravity outside the sun. Only the heavy outer planets (Jupiter, Saturn, Uranus, Neptune) have a non negligible influence.


C# source code for reproducibility:

void Main()
{
    // http://en.wikipedia.org/wiki/List_of_gravitationally_rounded_objects_of_the_Solar_System
    Body[] planets={
        new Body{Name="Mercury",Distance=57909175,Mass=3.302E23},
        new Body{Name="Venus",Distance=108208930,Mass=4.8690E24},
        new Body{Name="Earth",Distance=149597890,Mass=5.9742E24},
        new Body{Name="Mars",Distance=227936640,Mass=6.4191E23},
        new Body{Name="Jupiter",Distance=778412010,Mass=1.8987E27},
        new Body{Name="Saturn",Distance=1426725400,Mass=5.6851E26},
        new Body{Name="Uranus",Distance=2870972200,Mass=8.6849E25},
        new Body{Name="Neptune",Distance=4498252900,Mass=1.0244E26},
    };
    var bodies=new[]{Sun}.Concat(planets);
    bodies.Select(planet=>new{
        Name=planet.Name,
        Distance=planet.Distance.ToString("E2"),
        Mass=planet.Mass.ToString("E2"),
        DistanceInSunRadii=planet.DistanceInSunRadii.ToString("E2"),
        MassInSuns=planet.MassInSuns.ToString("E2"),
        CenterOfGravityShift=planet.CenterOfGravityShift.ToString("n5")
    }).Dump();
}

static Body Sun=new Body{Name="Sun",Distance=0,Mass=1.98855E30};
const double SunRadius=696342;

class Body
{
    public string Name{get;set;}
    public double Distance{get;set;}// in km
    public double Mass{get;set;}// in kg

    public double MassInSuns{get{return Mass/Sun.Mass;}}
    public double DistanceInSunRadii{get{return Distance/SunRadius;}}
    public double CenterOfGravityShift{get{return MassInSuns*DistanceInSunRadii;}}
}
  • I'm always surprised that Neptune's influence on the Sun's position is almost double that of Uranus' even though they have similar masses. – uhoh Sep 29 at 12:16

We often say that the planets orbit the Sun, which is usually a reasonable approximation. But in reality both Sun and the planets orbit the center-of-mass/center-of-gravity of the whole solar system, not the center of the sun.

That depends on what you mean by "orbit". If you mean that the equations of motion take on their simplest form in a non-rotating frame centered at the solar system barycenter, that is correct. Ignoring perturbations from the galaxy and nearby stars, this barycentric frame is an Newtonian inertial frame of reference. (Taking those small perturbations into account means that a barycentric frame is only approximately an inertial frame.) A heliocentric (Sun-centered) frame is also very close to inertial, but less so than is a barycentric frame. You have two choices with a frame that is known to be rotating and/or accelerating: Account for that rotation/acceleration via fictitious forces, or ignore them. Accounting for them makes the equations of motion more complex.

If on the other hand you mean that each body in the solar system is attracted toward the solar system barycenter, that is incorrect.

Mach's principle (which is somewhat controversial, but has led to some inspired thinking - see Wikipedia) can be interpreted as saying there is no center really, it is the overall distribution of mass that determines motion. And that is where Einstein's General Relativity takes over and says things just follow their local geodesics which are determined by the overall mass distribution. In that view some (well, many) configurations look like one object is orbiting another, but really every object is just doing its thing oblivious to all other objects. For a great book about this with wonderful illustrations (but lots of math) see Gravitation by Misner, Thorn and Wheeler and here.

I don't think this is always the simplest and most useful way to look at things, but it certainly has its uses (more so in complex configurations) and I just wanted to put it out there because no one else has in this thread.

protected by Deer Hunter Jun 8 '15 at 7:43

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.