The “Astronomical Unit” May Need an Upgrade as the Sun Loses Mass

Article written: 7 Feb , 2008
Updated: 26 Dec , 2015

The Sun is constantly losing mass. Our closest star is shedding material through the solar wind, coronal mass ejections and by simply generating light. As the burning giant begins a new solar cycle, it continues to lose about 6 billion kilograms (that’s approximately 16 Empire State Building’s worth) of mass per second. This may seem like a lot, but when compared with the total mass of the Sun (of nearly 2×1030 kilograms), this rate of mass loss is miniscule. However small the mass loss, the mass of the Sun is not constant. So, when using the Astronomical Unit (AU), problems will begin to surface in astronomical calculations as this “universal constant” is based on the mass of the Sun…

The AU is commonly used to describe distances within the Solar System. For instance, one AU is approximately the mean distance from the Sun to Earth orbit (defined as 149,597,870.691 kilometres). Mars has an average orbit of 1.5AU, Mercury has an average of about 0.4AU… But how is the distance of one AU defined? Most commonly thought to be derived as the mean distance of the Sun-Earth orbit, it is actually officially defined as: the radius of an unperturbed circular orbit that a massless body would revolve about the Sun in 2Ï€/k days (that’s one year). There lies the problem. The official calculation is based on “k”, a constant based on the estimated constant mass of the Sun. But the mass of the Sun ain’t constant.

As mass is lost via the solar wind and radiation (radiation energy will carry mass from the Sun due to the energy-mass relationship defined by Einstein’s E=mc2), the value of the Astronomical Unit will increase, and by its definition, the orbit of the planets should also increase. It has been calculated that Mercury will lag behind it’s current orbital position in 200 years time by 5.5 km if we continue to use today’s AU in future calculations. Although a tiny number – astrophysicists are unlikely to lose any sleep over the discrepancy – a universal constant should be just that, constant. There are now calls to correct for this gradual increase in the value of the AU by discarding it all together.

[The current definition is] fine for first-year science courses. But for scientific and engineering usage, it is essential to get it right.” – Peter Noerdlinger, astronomer at St Mary’s University, Canada.

Correcting classical “constants” in physics is essential when high accuracy is required to calculate quantities over massive distances or long periods of time, therefore the AU (as it is currently defined) may be demoted as a general description of distance rather than a standard scientific unit.

Source: New Scientist

22 Responses

  1. David Madison, Sr. says

    For all the talk in the article, they did not give the critical number. How much mass is lost per year? It also equated two different mass loss mechanisms, one by conversion of mass into energy, and the other by solar wind and mass ejections. The energy conversion is easy enough to calculate, but how much mass does the Sun lose by expelling particles?

  2. Alay says

    Yes, I Agree, the defintion of AU should change as engineering students and researchers need to keep up with the loss of mass of the sun and this also should be done in realtime, for example, the constant k should be given a dynamic status, where on could calculate the mass lost in a given period.

  3. Ralph Kuhn says


    I see another problem.

    A “massless” object would not orbit the sun – even by Newtonian theory, with or without the mass of the sun changing.

    Am I missing something?


  4. Steen Jorgensen says

    Ralph, I can understand your confusion. Bear in mind, however, that a fleck of dust inside the space shuttle follows the exact same orbit around the Earth as the entire space shuttle does. For even though its mass is very small, and thus the gravitational pull from the Earth, only a very small gravitational force is required to keep it in that orbit.

    The orbit of a small object (mass m) around a large object (mass M) is well-defined, as long as m«M. That’s why they have the word “massless” in the definition. Probably, only a swift, mathematical trick is required to show that if this holds for m→0, it holds for m=0 as well.

  5. ddk says

    I’m a bit confused. What is wrong with keeping something like this constant like inch, gram, etc. as long as it’s origin is clearly defined? We need some stable things for reference in this world. If you need something more precise just calculate a second version and call it something like AUnow and update it at the same time as the earth day time or even dynamic calculated to the nanosecond in time.

  6. Brack the barbarian says

    I can understand that no-one will lose sleep over the demise of the AU but where will that leave the parsec? The AU is an essential part of it’s definition and whilst the AU may be of only local importance the parsec, kiloparsec etc most certaily is not.

  7. Dark Gnat says

    I agree. It’s like saying an inch is “this long” in 2008, but it will be “that long” in 2108.

    I think it would be best to stick with the current definition of an AU, and use it for general purposes.

    It may be a good idea to decide on a real-time AU for precise measurements, but it should be given an easily identifiable name and acronym. CAU (Current AU) for example.

  8. eileen says

    Alan Bennett and the AU

    Other than the fact that we inhabit the Earth, is there any logical reason to base these relationships on the third planet of the system? Are we still geocentric, and backwards, in our thinking?

    Take a new look at the solar system and see the incredible relationships that result when Mercury, the first planet, is used as the base measure for the other planets.

    Consider the relative values for planetary distances provided by Solar Geometry when Mercury equals 1:

    Planet Calculated
    Value from
    Solar Geometry
    Mercury 1.0000000000000
    Venus 1.8660254037844
    Earth 2.5833057115410
    Mars 3.9364580433138
    Jupiter 13.4399084391945
    These values produce some absolutely astonishing relationships between Earth and the other planets. Keep in mind that the following relationships are not approximations. When computed using values from the Solar Geometry calculations for each planet they produce exact mathematical equalities.

  9. UkMan says

    eileen –

    not sure how you can call mercury ‘1.0000000000000″ to 13 decimal places – then calculate the other planetary distances referenced to this with the same accuracy when no planet has a perfectly circular orbit – not even your reference planet.

  10. none says

    Mercury will lag behind it’s current orbital position

    It’s a contraction. Really.

  11. Kevin says

    It has no affect on a parsec. That is based solely on the speed of light.

  12. Edward says


    While it can be calculated in terms of light-years, a parsec (PARallax of one SECond of arc) is indeed based on the AU. For more info, please see the below link

  13. Brack the barbarian says

    Has the definition changed recently – seriously? The original definition of the parsec was something to the effect that ‘it was the distance that which an object would have a parallax of one second of arc when viewed from a baseline of one AU.’ Hence the rather odd name PARallax, SECond of arc. No mention of the speed of light. If the definition has changed then I would be grateful if someone could confirm this.

  14. justsomeguy says

    Why we measure the distance from the sun to us in ‘AU’ instead of light seconds is merely a reflection on our stubborn vanity. Man will ever be the measure of all things.

  15. Brack the barbarian says

    Sorry to be hogging this but no-one yet has answered Ralph and Steen’s point about a ‘massless’ object feeling the effect of the sun’s gravity. The essential misunderstanding Ralph is in your reference to ‘Newtonian’. In Relativity a massive object (like the sun) bends time and space around itself. All other objects, whether they have mass or not, will feel the effect of this bending. Indeed, this was one of the first predictions of Einstein’s theory to be checked out and it was found, during the course of an eclipse, that light (which is massless) does indeed bend when it passes close to the sun.

  16. pantzov says

    i’ll never look at the AU the same way again 🙂

  17. ron says

    Is there a tolerance associated with the AU? If so, would this make the loss of mass unimportant?

  18. Keith Atkin says

    There is a very simple answer to all this terrible confusion and worry over the AU.
    Get rid of it, along with the light-year, and the parsec , and replace them all with the metre and and its multiples. e.g. the gigametre and terametre are ideal for solar system distances, while the petametre, exametre, zettametre and yottametre will take us to the stars and furthest galaxies. Solved!

  19. Richard Holland says

    Excuse me, but I need to provoke a riduculus argument: some web sites are suggesting that our planet is experiencing an orbital variance, and that the AU is actually decreasing! This could be the true cause of global warming. Our global rotational speed has decreased by 1/4 second, and our magnetic shield is growing weaker (both symptons of a changing AU). I’m wondering if anyone is trying to calculate changes in the AU. It could be important!

  20. saito says

    if it’s true….that means this universe will be ‘game over’ in many years later?

  21. Isaac NO. says

    How about using the transformation speed from IR to microwave, the distance would then be measured from any fixed spot in space, and then relate to any certain spot that would be appropriate to use, such as the center of our galaxy for a directional guidance f.e.?
    Then you would only have to concern yourself about if the centre would stay fixed for eternity….that shouldn`t be a big problem in the measurable future at least!

  22. How to Get Six Pack Fast says

    After reading through the article, I feel that I really need more info. Can you suggest some more resources please?

Comments are closed.