A thoroughly sporadic column from astronomer Mike Brown on space and science, planets and dwarf planets, the sun, the moon, the stars, and the joys and frustrations of search, discovery, and life. With a family in tow. Or towing. Or perhaps in mutual orbit.

And the answer is....

Almost a year ago, Eris – the, uh, most massive known dwarf planet -- passed directly in front of an otherwise anonymous star, momentarily causing the star to disappear, as seen from the earth. By carefully measuring the length of time that the star disappeared, astronomers made a very precise measurement of the size of Eris. I care about the size of Eris for many different reasons, but the most trivial yet emotional for me is the fact that, 5 years ago, I measured the size of Eris myself. We used a much more difficult and less accurate technique than watching a star disappear and timing it. We looked at Eris with the Hubble Space Telescope and carefully compared the tiny disk that we saw with a picture of a star (which should show no disk at all) and we claimed that we could tell that Eris had a diameter of about 1.3 pixels on the HST camera. Only 1.3 pixels! It’s hard to imagine that you could tell the difference between something 1.3 pixels across and 1.2 pixels. In fact, it had never been done before. Even we were not convinced at first that our technique was as accurate as it appeared to be. So we spent months on a careful analysis to make sure we had done nothing wrong. In the end our measurement technique passed every test we could dream up for it, and we became convinced that it was correct. We wrote the paper to announce it to the world. The diameter of Eris, we claimed, was 2400 km with an uncertainty of 100 km in either direction (I’ll be writing this as 2400±100 km).

I’ve been waiting for most of the past year for the new results from the stellar occultation, holding my breath. One of the nice things about science is that no matter what you do, eventually, someone is likely to come along and make the same measurement more precisely than you did, and you will get to learn whether you were right or whether you screwed something up. When you screw up, you do so in a very public manner. Everyone knows. You don’t look so good. It’s best to not screw up. The new measurement of the size of Eris would instantly tell us whether we screwed up or not.  

 This week the scientific paper describing the observations is finally out and we now know that the diameter of Eris is 2326±12 km.  (Pluto is a similar size, by the way, but with a larger uncertainty. We have no way of knowing which one is actually biggest until New Horizons gets detailed images of Pluto in 2014. Gory details on the uncertainties in the size of Pluto here). It’s an excellent result from a top-notch team.  Plus, I should point out, 2326 km is well within our 2400±100km that we had estimated. It shows that our very difficult measurement was, in fact, correct. I would like a little gold star, please.

I will admit: the most important implication of this result is not that we did our measurements correctly five years ago. But it’s also definitely not that Pluto and Eris are “nearly twins” although that is what you will read in the press release that accompanies the paper. In fact, the most important implication is that Pluto and Eris – which we used to think of as near twins – are much more different than anyone predicted.

How can I say that they are different when they are essentially the same size?  Because – I hate to say it – [precise] size doesn’t matter. The precise size that an object has tells you almost nothing. Some objects will always be a little bigger, some a little smaller. There is nothing interesting in knowing that. What matters is approximate size. We have known for a long time that Pluto and Eris were approximately the same size. Nothing has changed there. No scientific value in knowing that any more precisely.

What we also thought, though, was that Pluto and Eris had approximately the same density. Unlike precise size, precise density matters, and it matters a lot. Density tells you the composition of the object, which tells you something about where it came from and what happened to it since. If you had asked me a year ago, I would have told you that Pluto and Eris came from similar locations and had similar histories, thus they should have the same density, which would mean that Eris, which is 27% more massive, should have 27% greater volume to have the same density. Thus I would have predicted that Eris was 9% larger than Pluto. That fit our measurement, too; we just needed to assume that Eris was at the high end of the 2300 – 2500 km range.  It seemed the most reasonable expectation. There was no good reason to expect anything else.

We now know that the opposite is true. Pluto and Eris are approximately the same size, but that means they have very different densities (because we know Eris is 27% more massive it must also be 27% more dense. That’s a lot). As I wrote extensively when discussing this earlier, dwarf planets are crazy.  Rather than repeat that entire post again, I just encourage you, when pondering what the size of Eris means, to go back and read that post. I am convinced that understanding why Pluto and Eris are so different is one of the keys to understanding the formation of the entire outer solar system. So I find it particularly funny when people say they are twins.

Though it is clear that precise size has little scientific value, I will admit that certain bragging rights come with whoever can claim to be the biggest dwarf planet. For now my money is still on Eris, as I know that the more trustworthy measurement techniques yield the smaller sizes. But we just don’t know, and won’t until New Horizons flies by in 2014. I think it will be quite funny if one of the legacies of New Horizons is to prove that Eris is, in fact, the larger of the two.

I wouldn’t be surprised, however, if, at that point, the question is even less interesting than it is now. There are surely even larger dwarf planets out there. It is only a matter of time before both Pluto and Eris are supplanted. Perhaps New Horizons will tell us which of the two is the second largest dwarf planet and which is the third largest. Perhaps they will be even further down the list. Scientifically, it will still not matter, which is good, because explaining why Pluto and Eris are so different is going to keep us busy for many years, I suspect.


  1. "keep up busy"!

    learning how to spell "us".

  2. I would have expected that the writer at Space.com would know better and avoid the usage of the word twin (or at least had put a qualifier like "quasi" or pseudo" before the word twin) but they chose to go with "almost perfect" instead!


    Dwarf Planet Eris is 'Almost Perfect' Pluto Twin

    by Mike Wall, SPACE.com Senior Writer


  3. From the mythic realm: Eris is no twin to Pluto. She is the goddess of chaos and discord. He is the god of the underworld, and some say money as well. It's not about size. It's about archetypes.

    Sorry science guys, this comment may drive you crazy! but Mike Brown, the maverick, understands.

    1. Unfortunately, the myth does not matter. Eris is the planet of success and good fortune in astrology. Unfortunately you are not a good enough astrologer to prove that or understand that or even comment on that. Therefore science does not listen to astrology. One day it will change. But as long as astrologers believe in myth, it will not change.

  4. Your uncertainty limit of +/- 100 km: 1 standard deviation? 95% confidence limits? Please clarify.

  5. How do you know the chord across which Eris' disc crossed the star? Are your 'ephemera' for Eris (? orbital data) good enough to be certain where it is between a diameter and a tangent?

  6. I've heard Triton referred to as "Pluto's twin." How does it really compare to Pluto and Eris?

  7. the amount of objects greater than an half thousand miles are growing in the solar system! I know that nerd scientifics do not care about name of scientific object! but when 2007 OR10 will have a real name for this object came into the common world and not only into the nerd world where the nature are only numbered sampler for scientific purpose and glory !

    1. hiiiiiiiiiiiiiiiii !!!!!!!!! :)

  8. Size does matter. I have a bright yellow t-shirt with bold black letters that says it does :D.

    Jupiter and Saturn vary by mass, size, and density. They vary so much that it would be difficult to say they share a common history and origin, but they do. Saturn is a little less than a third the mass of Jupiter. Unlike KBO's, the mass of a gas giant is quite dependent on mass. Extra mass means extra gravity and that compresses the gas Jupiter is made out of so it's denser.

    The assumption you make in this post, Mike, is that once Pluto and Eris formed, they would keep the same density they formed with. I beg to differ. Pluto has moons and the arrangement of them suggests that it has undergone a collision. Eris has one moon and it's smaller than Pluto's. Perhaps it has also undergone a collision. Haumea has very definitely undergone a collision.

    At this point I would like to respond to the anonymous poster who has heard that Triton is referred to as "Pluto's twin". That idea has been around longer than Eris and Haumea (at least in the consciousness of Man). At the time this was most popular, it wasn't considered important that in fact, Triton is about 60% more massive than Pluto. They were still considered twins.

    The mass of Triton is equal (or about as equal as sort of narrow error bars can be) to the mass of Pluto plus the mass of the four major moons of Uranus. Since Neptune has a ring system that orbits around the planet's equator and a moon, Proteus, that does the same, how does Triton's backwards orbit fit in to all this?

    Why is the Neptune moon system so different from that of the other three gas giants?

    If Triton is in fact an interloper from the Kuiper belt, then it's mass can be accounted for by a Pluto sized object accreting the original moons of Neptune. Triton is orbiting at the same place with regard to Neptune as Uranus's four moons are with respect to Uranus. That's a strange coincidence, isn't it? Or is it?

    Uranus and Neptune are pretty much considered twins, too, mostly because they're about the same size. The axial tilt is much greater for Uranus, though, in fact it spins slightly retrograde. I used to think it was hit by four Pluto-sized bodies but then I took angular momentum in Physics class and was awakened from my stupor by an internet question by Miss Julie Williams from the United Kingdom. Four Pluto sized bodies, even with perfect limb-grazing shots, will not turn Uranus over on it's side. It would take an Earth mass body and even then you need a perfect limb shot. There's some other explanation for Uranus' tilted spin. And that explanation is, uh, gee, I don't know :?

    If the explanation for Pluto's and Eris' moons' existence is a collision then they would not neccessarily wind up with the same density. There might be other reasons for different densities, also, for instance Pluto approaches closer to the Sun than any of the objects mentioned here except Uranus, so sunlight sublimating ices off it's surface would result in different densities.

    Since this is gettin almost as long as Mike's post, I would like to close with the kicker:

    Charon does not have the same density as Pluto.

  9. John D,
    in reports I have seen, there were multiple measurements from a few observatories so combining them revealed longest transit as diameter.

    JohnD said...

    How do you know the chord across which Eris' disc crossed the star? Are your 'ephemera' for Eris (? orbital data) good enough to be certain where it is between a diameter and a tangent?

  10. Re: transit estimation of diameter. If you have two observations of the transit from different locations, the timings of the star winking out and back on give two chords. The four endpoints of the chords uniquely define a circle (even if neither of the chords corresponds to the diameter). Estimate quality would increase with the number of observations/chords, but I think even one full chord and a tangent point would suffice.

  11. @mikeemmert: The Moon doesn't have the same density as the Earth, either.

    This situation is expected when a satellite is created by a collision of a smaller body with that of a differentiated larger body. The satellite ends up being composed of bits of the smaller body (usually less dense than the larger body) plus surface (less dense) bits of the larger. Hence it's no surprise when a satellite is less dense than its primary.

  12. Congratulations on the accuracy of your measurement!

    I would have given you the gold star you request, but I have this feeling you'd actually prefer a ball of reddish, dirty ice instead... :-)

  13. did mike brown has studied why there is a lot of Objects around 67 UA?????

  14. Hi Mike,...what do you think about these works:http://arxiv.org/PS_cache/arxiv/pdf/1102/1102.0212v6.pdf ?
    (On the anomalous secular increase of the eccentricity of
    the orbit of the Moon,...)
    Did you find or could you find any similar disproportions in motions of planetoids (Eris, Sedna,Pluto,..)? Pavel Smutny

  15. It would be really ridiculous to the point where I LOL, if in three years New Horizons identifies and proves that Pluto is not only larger than Eris, but the largest of the known dwarf planets, even if by an inch. If this is the case, all who identified Eris as the larger of the two, will have egg on their face for bringing the sudden and urgent need to create new definitions and classifications for planets to the IAU.

    Okay, so density is far superior and important to me, as a drop of matter from the center of a black hole, may contain more matter than the entire Milky Way (except for the black hole at the center of The Milky Way) . Most interesting prospects for life sustaining worlds (in our solar system) include Enceladus, Europa, Ceres and Titan, all of which are much smaller than a gas giant such as Saturn, whose density is less than that of water.

    I do not care if a world is the size of a marble if it harbors life. Although it is possible that life exists on gaseous worlds, I think the scientific community believes it to be improbable (perhaps based on our narrow- minded view of what life needs to exist), therefore less likely to exist on such a world.

    Is it not mass that determines a world's shape, size, ability to clear its neighborhood etc.; it's certainly not a condition of size alone; as with Saturn, of great size does not mean of great density.

    So what if Eris is 27% more massive than Pluto, what does it point to regarding differences in their origin, age, composition or other things of interest? Specifications of a world can be interesting, but what science can do with such information can be very interesting. Answers to questions such as "how can worlds which reside in the same region be so different (if this is the case)," "how were such worlds made" and "what is the natural history of these worlds," is of greater interest to me.

    I am not going to be waiting on the edge of my seat for New Horizons to make precise measurements of small frozen, red, blue or purple worlds. There is a lot more out there than we can know at this time, but one thing is for sure, we will continue to find worlds, because there are an infinite amount of them in the universe, just waiting to be discovered.

    May the search continue, forever...

  16. I've also heard Venus described as "Earth's Evil Twin". It occurs to me that twins don't have to be identical. They merely have to be siblings born from the same mother at the same time.

    Presumably, Earth and Venus formed more or less together, and likewise Uranus and Neptune. How about Pluto and Eris?

    Bob Shepard of Denver

  17. What I appreciate most about you and your work, Dr. Brown, is that you put great care and emphasis on accuracy, no matter how popular/unpopular that stance is.

    Seems to me that saying Eris and Pluto are alike because they are roughly the same size is like saying two people are alike because they are both 5'-11" tall--even though one weighs 300 pounds with blond hair and blue eyes, and the other is a dark-eyed brunette who is height-weight proportionate. Seems like it's a weird way to try to correlate two very dissimilar objects via one of their very few common denominators...

  18. Pluto and Eris are neither twins nor rivals. Why does everything have to be a contest? Earth and Venus have similarities, as do Uranus and Neptune. And how can we be certain which object is bigger unless and until we visit BOTH of them? Yes, Eris is a lot further away, but we should at least start thinking about a mission there. Every mission has taught us that we don't really know much about a world until we go there. The whole notion of having to make a decision based on which object is bigger ends up being downright sily, just like the idea of "bragging rights." These are all planets of our solar system; they are all different, but they are all fascinating. And I disagree about their not being interesting to people. Clearly, all this coverage and all the books and articles that continue to be written show that the outer solar system is a place that interests a great number of people. I think there will eventually be a reconsideration of the IAU vote to allow for the inclusion of dwarf planets and of exoplanets as well.

  19. I believe that the matter of size (if by size you mean "diameter") is in itself problematic.

    Let me elaborate a bit more.
    Unless both Pluto and Eris are perfect spheres and not someting more akin to a triaxial ellipsoid or even a pear shaped body, there in not only one dimension that can emcompass size.

    If you want to go by "size", meaning the M in the Système international of units, then you would have to have it as "diameter of a sphere with equal volume". So there is no Pluto being "bigger" than Eris by one centimeter, inch, foot or meter. It could well be that Pluto's largest dimension is larger by 2 kms. than on of the "diameters" of Eris but contrarywise there is a shorter dimension accross Pluto which 1 Km. shorter than the longest dimension on Eris.
    At the end of this comment the issue is that the answer to the question of which is bigger? Might be "what is it meant by bigger?
    Bigger equatorial diameter? Bigger ESD? (Equivalent spherical diameter)
    If One of the bodies is 2326 +/-12 Kms. and the other is 2400 +/- 50 Are this numbers refered to ESD (equivalent spherical diameter)or to one maximal chord?
    What is uncontroversial is that ERIS IS 27% MORE MASSIVE THAN PLUTO.
    As a corollary we can say that the eridian density is at least 20% greater than the plutonian.
    Lets just hope that New Horizons can give us estimates of Pluto's J2 at the end of the day.


  20. Hi Mike,..Didn't you participated on atempt to observe asteroid 2005 YU55?

    much smaller eccentricity but direction cca like Elenin,..Pavel

  21. The primary planets in the Solar System are: Mercury, Venus, Earth, Mars, Vesta, Juno, Ceres, Pallas, Jupiter, Saturn and Uranus.

  22. Screw that, the Solar System is a huge place! There's room for everybody, even Sedna (and maybe Tyche)!

  23. Thank you, KenN and djlactin, for your explanation of diameter estimation from star transits.

    Rudy, that didn't help!

  24. Johh. I'll try harder next time :)


  25. Mike, I've noticed that the numbers you have published recently for Eris' percentage of eccentricity, degree of inclination and semi-major axis are a little different than those that are commonly cited. Can you tell us why this is?

  26. Hi Mike,
    Last week my in-laws dragged me to the Kennedy Space Center tour and since I passed on the IMAX show because it was 3D, I stumbled upon your book at the gift shop while waiting and bought it on a whim, because quite frankly it was the only book in there that looked remotely enjoyable to me. Glad I did! I now have an appreciation for the sky and am almost tempted to buy a telescope? and try to figure out where the heck all those planets/stars are so I can show my 5-year old son. Before, I would just point out the moon to him and now I feel ashamed of my nonchalance towards the night sky. Anyway, great book even for a non-techy like myself. Even my CS-professor husband paged through it with interest, and we laughed at some similarities of what must be the researcher state-of-mind (same coffee shop, same order, etc). Thanks for opening my eyes to a totally different world. And best to little Lilah!

  27. Mike Brown! Your blog is my crack! You've deprived me for many moons now! Write more! Please?

  28. Just bought your book and am really enjoying it! I know how it ends, with this "dwarf planet" nomenclature that I frankly think will be tossed by a future generation. For me, Pluto will always be a planet, and it sounds like Eris is the tenth :)

    I know there are other, smaller TNOs but my admittedly lame and purely poetic definition is a sphere that orbits that sun, and is at least as big as Pluto. Not IAU caliber, I admit.

    Eric Martin

  29. Is there a map showing where Eris, Sedna and company are in the Solar System?

  30. Hi Mike,

    Of as much interest as the relative diameters of Eris and Pluto are, the diameters of their orbits is of even greater interest.

    You got a mention in this article I wrote a while ago:


  31. I enjoy your blog, and have nominated you for a Very Inspiring Blogger Award. The details are at http://jameswight.wordpress.com/2013/01/12/very-inspiring-blogger-award/

    Hoping you will resume blogging soon!