The best proof that the Earth spins | Hacker News Reader

The best proof that the Earth spins

profmattstrassler.com

91 pointsnyc11144 comments

The gyroscopes described in the article actually, if we measure with sufficient precision, will not keep pointing in exactly the same direction as determined by distant objects. They will precess. Heuristically, the precession has three components:

(1) Thomas precession;

(2) De Sitter or geodetic precession;

(3) Lense-Thirring precession.

The first two are due to the rotation of whatever platform the gyroscope is riding on, in this case the Earth, so they are also evidence for the Earth's rotation. The second is due to the spacetime curvature produced by the Earth. (The third also involves spacetime curvature, through effects on it due to the Earth's rotation.)

(Note that these effects are separate from the effects of imperfect isolation of the gyroscope, which the article discusses.)

These effects are technically present but negligible for all practical purposes (well, unless you're building the next Gravity Probe B). The geodetic precession for a satellite in polar LEO is roughly 1 deg per 500 years [1]; the other types are orders of magnitude smaller.

You've made a good point, however, that two of these types of precession are evidence that the Earth spins (whether you could convince a flat-Earther of Thomas precession is a separate matter). But let me offer a gentle correction: Thomas and Lense-Thirring precessions are caused by the rotation of the Earth, whereas the geodetic precession is merely caused by the presence of a central mass.

---

[1] Gravity Probe B final report: https://arxiv.org/abs/1105.3456

> Thomas and Lense-Thirring precessions are caused by the rotation of the Earth, whereas the geodetic precession is merely caused by the presence of a central mass.

Oops, yes, I misstated this. Thanks for the correction!

> if you watch carefully, you’ll see none of them are heading west.

I've seen this posted twice in different places by different people this week. It's not true.

While most satellites are west to east (or in polar orbits), most Israeli sats orbit east to west. This is because their launch center is on the west coast of their country and it's not good for them to be launching rockets over their neighbors.

I'm not an expert but looks like you're right.

> Shavit rockets are launched from Palmachim Airbase by the Israel Space Agency into highly retrograde orbits over the Mediterranean Sea to prevent debris coming down in populated areas and also to avoid flying over nations hostile to Israel to the east; this results in a lower payload-to-orbit than east-directed launches would allow

https://en.wikipedia.org/wiki/Shavit_2

One of many hidden costs of conflict in the middle east.

Likewise launches from Vandenburg AFB, where prograde orbits are disallowed for safety reasons. (Most Vandenburg launches are actually more polar than retrograde however.)

My understanding is that the first conclusive proof that the Earth rotates was the astronomical observation of the aberration of starlight[1] in ~1730. The rotation of the Earth causes about 20 arcseconds of deflection from a stars known position, which is pretty large. The phenomena itself is due to relativistic beaming, and lead to one of the first measurements of the speed of light along with Roemer's earlier work using the moons of Jupiter as a clock in ~1680.

[1] https://en.m.wikipedia.org/wiki/Aberration_(astronomy)

I think the 20 arc second deviation is caused by Earth's orbital velocity around the Sun. Did you intend that instead of rotation? The latter causes less than 1 arc second deviation.

Why the need for a perfect gyroscope that can spin for 24 hours? Why not use a motor to counteract the slow-down without affecting the rotation? Science museums sometimes do this to their Foucault Pendulums so they don't need to keep re-swinging them: https://www.pugetsound.edu/about-puget-sound-0/leadership/20...

I always thought the direct parallel transport explanation worked best for Foucault's pendulum: https://www.youtube.com/watch?v=jNQCOknMZKg

Also: https://physics.mcmaster.ca/phys3c03/notes/FoucaultPendulum....

The wand and gyroscope analogy is basically just constructing this scenario I think.

Ring laser gyros are fairly well known, but there's another way you can make a gyro with no mechanically spinning parts using the right combination of magnets. This paper describes an experimental apparatus that incidentally serves as a solid-state gyrocompass: https://arxiv.org/abs/hep-ph/0606218

(Disclaimer: I did my PhD in this group but started there after this paper was published.)

High precision gyroscopes showing a 15 degrees per hour drift?

Thanks Bob! :-)

I wonder why Prof. Strassler does not consider star trails, like here, https://en.wikipedia.org/wiki/Star_trail definitive proof of Earth's rotation. In historic times astronomers invented crystalline spheres [1] that carried stars to save the stationary earth doctrine. But now we know that crystalline spheres do not exist (stars have their own proper motions) so since stars do not rotate, the earth must rotate. What's wrong with this argument?

[1] https://en.wikipedia.org/wiki/Celestial_spheres

> In historic times astronomers invented crystalline spheres [1] that carried stars to save the stationary earth doctrine.

My understanding is that the ancients used the celestial spheres to explain the movements of the planets around the Earth - the spheres themselves moved relative to each other hence the different speeds and courses of the planets.

The spheres could be made to work regardless of whether the Earth spins or not. If the Earth is static, then the spheres can still use their relative motions to explain the motion of the planets. In contrast, if the Earth is spinning, you add its spin to that of the spheres and an observer on Earth couldn't tell the difference.

What's wrong with your argument is it assumes facts that themselves need to be proven. "Stars have their own proper motions" is not an easy fact to prove, and it took a very long time for people to prove that "stars do not rotate" (I presume you mean around the earth). The sort of thing being discussed here is the sort of thing you need to do to prove that crystalline spheres do not exist. There are of course other ways to prove that crystalline spheres do not exist, but they are things like the existence of black holes or the speed of light. And in general, all these sorts of things require a much broader knowledge base/trust in our current understanding to prove the idea that the earth rotates.

This article is saying that if you can prove how a gyroscope works (itself requiring proof of the Conservation of Angular Momentum), you can show that the earth rotates. Which is ultimately a much simpler thing to prove than things about the nature of the stars themselves.

We know stars are distributed in various distances, this we know by observation. If not the earth but the stars rotate we must assume that billions of crystal spheres at different distances somehow rotate in exact coordination to trace the star trails that we observe. I don’t think there is such a huge “conspiracy” of crystal spheres to fool us.

> We know stars are distributed in various distances, this we know by observation.

How do we know? Establishing the cosmic distance ladder requires far more elaborate tools than the ones discussed here (Special and General Relativity, for instance) and the necessary measurements and observations also presume in one way or another that Earth is rotating. Even if you're only talking about stars in our immediate neighborhood for which a distance measurement merely needs the parallax effect (instead of SR or GR): In this case the parallax effects literally rests on the rotation of Earth and its orbit around the sun.

> the necessary measurements and observations also presume in one way or another that Earth is rotating.

Yes, the lack of star parallaxes were the main argument offered by Tycho Brahe against a moving earth theory. When star parallaxes were observed the motion of earth was accepted. The way I understand it, to measure star parallaxes the earth must move. This is not an assumption, there is no circular reasoning. And we observe different parallaxes for different stars, meaning that stars are not all on the same distance from earth. I don't know if this is a trend like flat earthers to say that all stars are the same distance from the center of the universe?

> In this case the parallax effects literally rests on the rotation of Earth and its orbit around the sun.

So, this way you prove that the earth moves and stars have different parallaxes and are not all attached to a crystal sphere. I don't see a problem here.

Stars can still have parallaxes while bolted to the spheres. Maybe the spheres themselves wobble a little. Maybe the motion of the sun around the earth imparts a slight motion to closer spheres. The fixed stars are obviously on spheres far enough away that the sun doesn't shake them.

> Yes, the lack of star parallaxes were the main argument offered by Tycho Brahe against a moving earth theory.

Much earlier than Brahe. Aristotle in De Caelo (Book II, part 14):

> The earth, then, also, whether it move about the centre or as stationary at it, must necessarily move with two motions. But if this were so, there would have to be passings and turnings of the fixed stars. Yet no such thing is observed. The same stars always rise and set in the same parts of the earth.

* http://classics.mit.edu/Aristotle/heavens.2.ii.html

* https://en.wikipedia.org/wiki/On_the_Heavens

> This is not an assumption, there is no circular reasoning.

Of course it is an assumption. A star moving across the sky is not a proof by any means.

> And we observe different parallaxes for different stars, meaning that stars are not all on the same distance from earth.

No, we don't measure different parallaxes for different stars. We measure different trajectories for different stars across our night sky and then, using the assumption that Earth rotates and that it is on an orbit around the sun, we conclude that there must be a parallax effect involved in the observed trajectory of the star which we can then use to estimate its distance to us.

> So, this way you prove that the earth moves and stars have different parallaxes and are not all attached to a crystal sphere. I don't see a problem here.

To see that what you're referring to as "proof" is no proof at all, consider that Ptolemy's epicycles (i.e. the elaborate version of the crystal spheres) are just a Fourier series and can therefore be used to describe any periodic movement across the sky. (Without you having to know anything about whether Earth rotates or not.) For a very impressive example see https://m.youtube.com/watch?v=QVuU2YCwHjw .

“Stars have their own proper motions" is not an easy fact to prove. . .”

No, proper motions are well known, https://www.britannica.com/science/proper-motion#:~:text=pro....

So we must further assume that a star attached to a crystal sphere will have a separate motion which also need to be explained.

It's well known that the earth rotates too though.

>> a star attached to a crystal sphere will have a separate motion

Two spheres. Proper motion can be explained completely if we allow each star to exist on its own crystal sphere. Layers of spheres could also extend great distances. Maybe the universe is vast, but it is all spheres. What does disprove spheres are collisions of objects having different proper motions.

> Two spheres. Proper motion can be explained completely if we allow each star to exist on its own crystal sphere.

Wow! I can believe we are discussing in the age of Hubble space telecope and Webb etc. the existence of crystal spheres and how they can explain motions of stars. I don't believe crystal spheres exist but I respect your belief in them. To me they are an unnecessary and cumbursome explanation.

As cumbersome as unobserved massive particles representing the majority of matter in the universe?

Nobody believes in crystal spheres. The point is they are not so trivial to disprove.

> This article is saying that if you can prove how a gyroscope works (itself requiring proof of the Conservation of Angular Momentum)

Which means we're already in the domain of Newtonian mechanics, which came almost 80 years after Kepler's laws of planetary motion which gave us a clear understanding of the motion of a bunch of “stars” (yes, planets are not stars but they have the same kind of “star trails”).

The really cool thing about star trails is that it's something really intuitive and easy to observe, and as such it makes a pretty convincing “proof” of earth rotation. Unlike the “proof” in this article, which will leave most people unimpressed (and which as I said earlier, uses a much more advanced physics theory than what's needed if you really want to prove earth rotation with star trails).

The author also seems think Foucault pendulums aren't conclusive. That one is much harder to argue against than star trails (there's no need to prove the absence of celestial spheres or other kinds of star movement).

Honestly, I don't see much difference between a gyroscope experiment and the Foucault pendulum. Except that the later is much easier to do.

As the article notes, the Foucault pendulum is slightly more work to prove. Nothing beyond high school math, but it's less obvious than "once you point this gyroscope at any star, it keeps pointing at that star no matter how that star moves, how could that be if the stars are moving but not the earth?"

>> how could that be if the stars are moving but not the earth?

Because maybe the earth is fixed, the heavens all move, and the pendulum is reacting to the heavens. Maybe a tidal force from the stellar spheres is imparting a slight motion on the pendulum. Maybe only objects like gyroscopes are sensitive enough to feel this force.

Oh it's definitely possible to have alternative theories but they get increasingly outlandish as you make small modifications. E.g. "how does a gyroscope know which star to point at? What happens if the gyroscope points at the middle between two stars? What happens if you point it at a star that has not yet been discovered (e.g. more powerful telescopes reveal it in the future)? Do different stars have different effects?"

Trying to answer these questions makes the stellar spheres theory require more and more ad-hoc additions to support.

Newtonian mechanics does not have absolute speed. We can only say A is moving at 5m/s relative to B.

But Newton does have absolute acceleration. So we can say A is accelerating at 5m/s² without even thinking of B.

Since a rotating body accelerates, we should be able to say that the earth rotates without even thinking of the sky.

The proof does exactly this. Looking at a gyroscope is sufficient to prove that the earth rotates. No need to look at the sky

But the gyroscope looks at the sky, it points at the same star, right? This is how the proof works.

Imagine that you live in a starless void. Could you still prove that the earth spins?

Yes, simply point the gyroscope towards an arbitrary direction and repeat the same process as in the proof.

If the earth doesn't spin, then the gyroscope would point at the same direction always. But if it does spin then it would show those circles.

It doesn't account for the option that the Earth is stationary and the sky is spinning, much it like it would be implemented in a game or CGI scenes for example.

Gyroscopes can rule this out as well since they're showing rotational change directly.

In this post and your subsequent replies you rely on the existing astronomical body of knowledge to show that the Earth spins.

This misses the point of the OP, which is to try to provide a simple physical demonstration that anyone could perform or witness which shows that the Earth spins, in particular one that's more intuitive than Foucault's pendulum.

In 2001 he wrote a paper (unfortunately behind a paywall) entitled “Detection of Earth Rotation with a Diamagnetically Levitating Gyroscope”. This paper was co-authored with his hamster, Tisha, who apparently participated in the levitation experiments.

https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.10...

Time lapse with a webcam.

Watch the sky with a notepad.

Bring up a sat video feed.

Mix in some common sense with observations? We've know this for 1,000+ years, probably way way more EG: navigation at sea.

This doesn't settle whether it's the earth or the stars that are doing the spinning. That was the controversy in 1620. It remained an open question until Newton came along. The proofs discussed here are tests based on the consequences of Newtonian mechanics.

Coriolis will actually effect a long range projectile, but not by much.. (basically the earth spinning under a bullet with a sufficiently high arc).

Yes! While it's technically true, this has evolved into an urban legend about southern hemisphere naval battles during WWI.

https://skeptics.stackexchange.com/a/44760

Coriolis Effect

Reply on news.ycombinator.com