A pale blue marble: The first accurate scale model of the solar system

September 19, 2015 • 1:49 pm

This video, made by Wylie Overstreet and Alex Gorosh, is apparently the first attempt to make a model of the solar system completely to scale: that is, the sizes of both the planets and their orbits must be on the same scale of measurement. The one constraint is that they start with a model Earth the size of a marble: the rest follows automatically.

The model was constructed in the Nevada desert, and they needed seven miles of empty space to do it—and even that is excluding Pluto (it’s a planet, dammit!), an inclusion that would have made the model much larger (readers can calculate how much larger). It’s a laborious and fascinating task, and has garnered nearly 900,000 views in the last three days.

The final test, in which they compare the image of the real rising Sun with that of the model Sun, both from the vantage point of the model Earth, is lovely. Kudos to the filmmakers!

h/t: Kevin H.

73 thoughts on “A pale blue marble: The first accurate scale model of the solar system

  1. That may be the first model with the orbits drawn in. There are scale models of the solar system, with the sun and planets at their proper scale distances, at Westerbork in the Netherlands and near Zurich in Switzerland.

    1. I have seen size and distance scale models outside of museums. I think one was at the Smithsonian, but i am not sure now.

    2. There is also one on the Falkland Islands, it starts with the Sun at the war memorial on the outskirts of Stanley and I got as far as Saturn but Pluto is a reputedly a full day trip.

      1. Hmmm, Saturn would be about half-way across East Falkland (Mt Tumbledown?) if you went all the ay to Pluto on the furthest islet of West Falkland

  2. You can have Pluto as a planet, so long as you’re also willing to accept Ceres and a bunch of other asteroids, KPOs, and TNOs, some of them larger than Pluto.

    In any case, calling Neptune’s orbit “the edge of the Solar System” is wrong by almost any definition. The Kuiper Belt extends 20 AU beyond Neptune, and the scattered disk another 60 or more AU beyond that. The heliopause (where the solar wind meets the interstellar medium) is around 100 AU. The Oort Cloud starts at 1000 AU and extends to perhaps 100,000 AU.

    At the scale of the video, you’d need another 10,000 miles of desert for all of that.

    1. The heliopause/terminal shock region is the dominant boundary, though we can also contemplate the gravitational boundary.

      [Which is past the Oort cloud I think and depending on encounters with other stars. It is estimated that 10 % of the cloud is not indigenous though most of the exchange should have been in the packed cluster where the solar system was born.]

      Perhaps one should see the EM/interstellar media boundary as the harbor area for ‘sailing’ bodies, while the Oort cloud is boats that go out into national waters.

    2. I saw a note by Hal Levinson recently explaining some of the reasoning behind the IAU’s decision. I still don’t really agree with it, but he makes a better explanation than the usual “just-so” formulation.
      Long story short … “In this simple system, there are two ways in which a planet can remove a population of small bodies: 1) It can accrete all of them, or 2) It can gravitationally eject them from the Solar System. I will treat each of these separately. ” ;
      “A planet can eject a small body if it is able to change the body’s orbital energy by an amount at least equal to its original value. Tremaine (1993) shows that this occurs if ” [inequality in planet mass and orbit radius]
      ” Assuming that the small body has the same semi-major axis as the planet, we can show that for a planet to accrete all the nearby small bodies in the age of the Solar System (t*) it mass must be:” [another inequality in mass and orbital radius]
      Plotting up those two inequalities he shows an area of radius-mass space separating those objects that will accrete nearby small bodies, those that will eject nearby small bodies, and those that will do neither. Pluto and Ceres are in the “neither” category ; the classical 8 planets are in the “one or both” methods categories, with Mars and Mercury in the accrete-but-not-eject category.

      1. Very nice. I had a feeling there was some parameter space in which the difference would be clear. Thanks for digging it up.

  3. apparently the first attempt to make a model of the solar system completely to scale

    The first apparent US scale model, perhaps.

    Sweden has had a national one, Sweden Solar System, starting to cover the whole nation since the 90’s. The round, white Globen in Stockholm is the Sun, and the terminal shock is situated close to the northern end of Sweden. [ https://sv.wikipedia.org/wiki/Sweden_Solar_System ]

    I assume other nations may have something similar [ADDED @ POSTING: And I see from Stephen’s post that I assumed correctly], and that Sweden’s national model wasn’t the first.

  4. Very well done. I have some nascent ideas for how this same idea might be done as a permanent exhibit in Phoenix South Mountain Park, with the planets situated on the peaks of hilltops so you can see them all from each other. I think some combination of scaling and placement should get it — especially if you just care about distance to the Sun and not putting them all in a straight line.


    1. That would be very cool.

      Some cities have actually done that, i.e., placing scale models of the planets at various locations or along a road to show relative distance.

      Here is one of these orreries.

      1. The Boston one takes quite I while to see. It’s interesting that the Desert Scale model is visible all at once in principle. although something the size of a marble is not going to be too obvious from one or two orbits away. It must be quite special to see at night when lighted. Also, it is not likely permanent.

  5. I had never thought about how there is no true scale model or image of the solar system. I guess looking at the way it is portrayed would give one the impression that that is how it is in reality.

  6. Since they put Neptune at 5.6 km (3.5 mi), and it is 4.5 x 10^9 km at aphelion, then Pluto at aphelion would be about 9.1 km (5.6 mi) on this model. Maybe.

  7. The immensity of our smallness should shatter anyone’s notion of a personal, loving god. Too bad it’s only ‘should’.

      1. This is something that should be discussed in an upper level course at seminaries. Perhaps they use the extent of God’s love to define the boundaries of the universe – in terms of light years, or perhaps only out as far as the Oort Cloud. Myself, I suspect it is confined to the shape and size of the skull.

  8. There’s a similar model in New Zealand near the Mt John observatory, adjacent to Lake Tekapo in the South Island. It’s actually on two scales – a small one you can walk around, and a much larger one that requires some days to tramp to Pluto, where the planets are also to the same scale. It was under construction when I was last there and may have stalled for lack of funds. Great to see it also being done elsewhere. The world needs such models because of the way it gives true perspective (‘total perspective vortex’) on our personal scale vs that of “outer space”.

  9. Something just occurred to me…many have noted that they’re far from the first to make a scale model of the solar system with its distances…but they may well be the first to have made a scale model of the entire orbits. At least, I don’t think I’ve even heard of any attempts to do so.

    It’s one thing to line up all the models on a football field or along main street or the like. It’s a bit more of a challenge to sweep out the full circle….


      1. Relative to what? Angular or linear velocity? You could come up with arguments that the Earth proxy should take an entire year to carve out its path to that it should be going 67,000 MPH to even faster than light.


        1. Why ami thinking of the Monty Python skit with the swallow right now with Ben playing the person Arthur encounters?

        2. Relative to each other. Whatever linear speed they choose for Earth, Neptune should travel roughly 1/6 that speed, and take 165 times as long to complete one orbit.

          Deviating from those ratios (as they clearly have; they have all the planets moving at essentially the same linear speed) undermines their claim to have built an accurate scale model.

          1. Might be possible. Scale km/s to km/h and you’ve got Neptune at about minimum / engine idle speed, and Mercury doing about 30 MPH on a 500-foot-radius circle, if I’ve got the math right. That might be faster than is safe for the desert area, but I don’t think you’d have a traction problem at that scale. You wouldn’t be able to film long enough for a full orbit of Neptune, I don’t think.

            In any such exercise, it would be nice to include something traveling at the speed of light as a reference. But that would still mean 186,262 MPH…a bit of a problem.

            Another approach might be to scale it so that light would take 8 minutes to reach Earth…but that would be a bit under 2 MPH for the light, and the planets would be too slow to see movement.

            Maybe there’s an happy medium…?


              1. No; the assumption was that the second would get scaled to the hour. Mercury’s orbital velocity is 47.362 km/s, according to Wikipedia, which is 29.429 miles per second. If we scale that to miles per hour — my ~30 MPH — then we’d have to similarly scale light from 182,262 miles per second to 182,262 miles per hour similarly.

                Even dropping it down another three orders of magnitude still leaves problems…light going 180 MPH, but the fastest planet going 0.03 MPH…half an inch per second.


          2. I think the primary advantage of a scale model of the solar system including accurately scaled velocities of the planets is that it illustrates just how relatively slowly they are moving. Since there are 365 days in a year and 360 degrees in a circle the Earth moves about 1 degree along it’s orbit in a day. With a radius of 579 ft. as given in the video the circumference of the orbit would be about 3637 ft. which means that every day this scale model of the Earth would travel approximately 10 ft. along this circumference. This would be imperceptible to the naked eye. It takes the Earth several minutes to traverse it’s own diameter!

  10. Spectacular! It does give one a sense of how silly the thought is of a god that controls all this.

    I went to a talk & slide show by David Levy a couple of weeks ago in Kansas City. Of course he is the astronomer David Levy of Shoemaker-Levy 9 fame, the comet named for the co-discovers, Levy along with astronomers Eugene & Carolyn Shoemaker. That comet famously collided with Jupiter in 1994.

    Levy was quite an entertaining & witty speaker. He was a friend of Clyde Tombaugh, the discoverer of Pluto, and Levy humorously insists that Pluto is still a planet. He had a slide of Clyde at his home, and on Clyde’s lap was his beautiful tuxedo cat named – what else? Pluto! Clyde claimed Pluto was a Pluto-cat. Lol!

  11. Should probably be required watching for those who claim that we’ll be able to colonise Mars or fly to another star system if we only try hard enough. (More and more it seems to me that what people are missing the most isn’t intelligence, rationality or education per se but perspective, both spatially and temporally, to appreciate what policies are important and what policies aren’t.)

  12. Any knitters out there? I have knit several “solar system” scarves. If each row is 10M km, the planets are beads and the overall length ends up being about 5.5 feet, including the 4 Kuiper belt dwarf planets. Planets not to scale, but distance is. Keeps the neck warm, too.

  13. There’s one in Finland in Helsinki and Espoo and has a scale of 1:

    A website for the project (in Finnish) is at https://www.ursa.fi/tahtitieteesta/aurinkokuntamalli/yleista.html

    At the site of Pluto there’s a construction site for the western extension of the Helsinki metro and the planet has not been there for ages.

    A map of the project can be found here: https://www.google.com/maps/d/viewer?ll=60.196156%2C24.853477&spn=0.061431%2C0.178528&hl=fi&msa=0&z=12&source=embed&ie=UTF8&mid=zT8gitEQfzQs.kHWnu9mC4WC4

    1. Yeah, I did a rough calculation of the liveable volume of tho solar system at one point. IIRC, it was something like 1E-21%.

  14. I used to wonder about a similar thing: a basketball court is 93 feet long and the Earth is 93 million miles from the Sun, so if you put a model Sun on one end of the court, and a model Earth on the other, a million miles to the foot, what size would Earth and Sun be? Using my imperfect arithmetic, the Sun would be (roughly) the size of a basketball – and our planet? An apple seed. Picture that.

    1. The Exploratorium has one of the many online calculators you can use for that sort of thing:


      A regulation basketball has a circumference of 75 – 76 cm, which works out to a diameter of 24 cm.

      When the Sun is scaled to a diameter of 26.83 cm = ~10.3 inches, the scale is about one foot per million miles, putting the Earth at 93 feet away, and making it less than 1/10″ in diameter — about 2.4 mm. Jupiter is about 1″ in diameter and almost 500′ away. The entire model would be about half a mile in radius.

      At the same scale, the nearest star is almost 5,000 miles away and a model of the entire galaxy would need almost as much space as there is between the real Earth and the real Sun.



    1. I suspect it will be seen in many high school science classes. My observation has been that science teachers are pretty good at assembling teaching resources like this. It helps the students but also helps the teacher accomplish important goals.

  15. We’ve celebrated our son’s 7 years old birthday party by inviting some of his little friends to an afternoon in a local observatory (Observatoir de Tauxigny) where the organise fun and educational events. After a short presentation about the solar system and some views of our sun with a solar telescope, they have taken all kids outdoors for a solar system walk. They’ve started the walk showing us a model of the Sun of about 1.75 meters in diameter and then walked the kids to an information board some 50 meters away where they found some info about mercury. Then, after some 50m more, Earth and son on. We stopped in Saturn, some 1.5km away from the observatory because the kids were showing severe signs of boredom and could not listen any more, but according to the guide we could walk about 4km more towards Pluto if we could keep the kids under control, which we could not.
    I am sure I have learned more with the experience than most of the kids. It is really hard to put such large distances into perspective.

    1. That sort of scale is going to be too big for kids that young. The thousand yard / meter “Earth as a peppercorn” model will be about the limit.


      For kids that young, I’d start establishing the size of the Earth. Use a campus map that shows the surrounding buildings. Pull out a city map and have them find the outline of the campus. Pull out a regional map and have them find the outline of the city. Pull out a globe and have them find the outline of the region. Then pull out the peppercorn….

      Also, if it’s your own child, I’d start even before then with a trip to the grocery store. Don’t give any hint of what you’re doing or why; just make it otherwise normal…but make sure you buy a cantaloupe or other yellowish / orangish melon that’s the right size for the Sun, grapes and blueberries and the like the right size for the gas giants, peppercorns for the terrestrial planets, and poppyseeds for the minor planets.

      You need about half a mile (or about a kilometer) for the model at this scale. Far enough to impress a young mind, but not far enough to bore to tears. Also…small enough that you can see the whole span at once. For a rerun, but not the first time, find a stretch of sidewalk that’s suitable. Mark out the locations ahead of time. Assign the planets to kids and have them sprint to their designated locations — obviously, with the fastest runner for the farthest planets. (Obviously, only for kids old enough to not run into the street, sidewalks not too crowded, etc.) Have everybody hold up their planets and look around at how far away everybody else is. This version you could do in under ten minutes — and, as a bonus, the kids burn off a lot of energy running the whole way. For bonus points, send the kids in opposite directions so there’s a mile between Neptune and Pluto…obviously, with Jupiter and Saturn in opposite directions, too, and so on….



  16. My father and I tried to construct a solar system model once when I was small, but we didn’t do the planets to scale, only the orbits and fit them in the basement thereby.

    Having read some astronomy books for kids, I asked if we could put Proxima Centauri near by somewhere …

Leave a Reply