The Best Test of General Relativity (by 2 Misplaced Satellites)

Okay... Hello. Hey. So, this is good, this is good. You - you're working, can you see me? I can see you. Do you know what went wrong in-uh.. during the launch..? Yes - it's not complicated, but, it's a long chain of events.

On August 21, 2014, two satellites were launched by the European Space Agency. They're called Galileo Satellites 5 and 6. They were intended to become part of the Global Navigation Satellite System, or GNSS. This is the European version of the American GPS systems.

Now, after successfully blasting off into space, the satellites were launched with the Russian 5-6 rockets. The final stage of the rocket was set to inject the satellites into circular orbit around 23000 kilometers above the Earth. But that - - is when something went wrong. Uhh, there was a thermal breach between a line of cold helium and a line of propellant. So, the propellant, it froze. This caused the failure of the altitude control thrusters. The satellites seemed to have been injected into some random direction. It was launched, but in the wrong direction. This sent the satellites into highly elliptical and seemingly useless orbits.

At their low point, the satellites didn't really get a full view of Earth. Earth sensors, which enabled them to orient their navigation antennas, stopped working because the Earth just filled their field of view. At the other extreme, the satellites went too high, experiencing significant radiation exposure due to the Van Allen belts. There was a threat that they would just shut them off.

Okay, who do we have to talk to, to make sure they keep these satellites alive? They had an idea for how to use them to make the best tests of general relativity to date. This was a strike of luck for us; we had been proposing such missions. When we saw that this accident happened, we were very happy about it, of course. Now, the satellites did have propellant on board, intended to allow for periodic course corrections over their planned 10 year lifespan.

And they could use that fuel to attempt to correct their orbits. So, they used the propellant on board. They did some of these maneuvers to bring them into stable orbit. But they didn't have enough to completely turn their elliptical orbits into circular ones. They couldn't circularize completely, thankfully for our project.

According to general relativity, relative to a reference clock, clocks tick slower in stronger gravitational fields. That is to say, closer to large masses, like the Earth, deeper in gravitational wells, so clocks in satellites should tick faster relative to those on Earth because they're in weaker gravitational fields.

Here, I'm ignoring the special relativistic effect that works the other way - making clocks in satellites tick slower than those on Earth because they're moving so much faster. This velocity effect is very well tested, so we focused on the gravitational parts. The gravitational effect is hard to measure precisely for satellites in circular orbit, but satellites in elliptical orbit have an advantage.

In every orbit, they go from their lowest position - perigee - to the highest position - apogee - and back. If you want to test the gravitational redshift of a clock, you need two things - you need a very accurate clock and you need a large change in gravitational potential. The satellite goes from 17000 to 26000 kilometers. The difference in altitude is almost 9000 (!!!) kilometers, meaning they are rapidly and repeatedly going from lower gravitational potential to higher gravitational potential and back.

So, clocks on board should be ticking slower when they're closer to the Earth, and then faster when they go to their high point, and they'll keep oscillating back and forth; slower and faster relative to clocks on Earth. If you compare a clock on the ground and a clock on the satellite, then you will have this variation of time.

Now, what's great about this is it allows you to eliminate a lot of sources of error because you don't really care about the absolute accuracy of the clocks. All you want to know is the difference between the rate of ticking at the low point compared to the high point. And since it's the same clock making the measurements at both locations, you can eliminate a lot of errors, like the noise in the clock or a systematic drift, and that's what allows the scientists to achieve such incredible precision.

Why we could do it is because it's a very predictable effect. Due to the eccentricity of the orbit, the signal we were looking for is really the modulation. All the other effects that are at other periods will not have an influence on your measurements.

Now, I should point out that if you were traveling with the satellite, for you, time would not speed up and slow down; time would be passing at a constant rate. You wouldn't be able to measure any change in the rate the clock is ticking. The relativity comes when you compare two clocks that are distant enough to feel the curvature of space and time.

The clocks on the satellites - there are a couple of different types - are all atomic clocks. The primary clock typically is a passive hydrogen MASER clock. A MASER is just like a laser, except it uses microwaves. Atoms of hydrogen interact with one specific frequency of microwaves. A photon of precisely this frequency will flip the spin of an electron.

So by tuning the microwaves so that they best interact with the hydrogen atoms, and then counting up exactly this number of cycles of that radiation, that's one second. You can keep track of time with incredible stability. In fact, over a 30 million years, a clock like this would not be out by more than a second.

Again, if you were traveling with the satellite, you would always observe the frequency of this radiation to be the same. But if you sent this radiation out to a distant observer who is not in a strong gravitational field, they would observe the frequency of your microwaves to be slightly lower than those from their hydrogen maser. In other words - red-shifted.

And the closer the satellite is to Earth, the more red-shifted the microwaves would appear. And hence the slower time would pass, relative to that distant observer. Locally, you cannot see any relativistic effect. This is called the equivalence principle. It's only when you compare the satellite clock to one on Earth that you would find the yo-yo-ing rate of the satellite's clock due to its oscillation back and forth in Earth's gravitational well.

This gravitational redshift was previously measured most precisely in 1976. That's right, for over 40 years, we haven't improved our measurement of the gravitational effect on time. In 1976, Gravity Probe A was launched aboard a sub-orbital rocket. It went up in a parabolic trajectory, reaching a maximum altitude of 10000 kilometers, and then it came down. That gives you quite a lot of modulation and gravitational potential.

The whole time it was in contact with the Earth through the microwave signal of the onboard hydrogen maser. And then they did a direct frequency comparison, so they really did a two-way microwave link. This allowed for a direct comparison of the rate at which a clock would tick on a rocket relative to a clock on Earth. The results matched the predictions of general relativity down to 140 parts per million.

The scientists I'm talking to were eventually able to convince those in charge of the satellites to let them use their misfortune to test the gravitational redshift predictions of general relativity. But actually carrying out the tests wasn't easy; one of the biggest sources of error was the positions of the satellites.

You'd think, in the emptiness of space, the satellites would perfectly maintain their orbits. But that neglects the power of sunlight. The photons of the sun, bouncing on the satellites, is the biggest source of error. That's right, the momentum of photons hitting the satellites was enough to significantly impact the measurements. Careful modeling plus laser ranging to the satellites brought the orbital uncertainties down to an acceptable level.

One way to improve the statistics was by collecting data over more than a thousand days. That's almost 3 years, unlike Gravity Probe A, which spent only 2 hours in space. So... what did they find? I think that we both agree that we did not prove relativity.

There isn't that we confirm the General Relativity. Unfortunately, they were able to reduce the uncertainty in the measurement by a factor of 5 over Gravity Probe A. So it's a new high score, the first in over 40 years. But what is the point? You saying 'unfortunately'?

Yeah. You know what we're looking for is deviation from General Relativity. Because we know this is not the ultimate error. History has taught us that new physics always lies at the boundaries. That more and more precise tests sometimes reveal brand new aspects of nature that we never would've observed if we had not made efforts to look, and there are good reasons to believe that General Relativity may not be the whole story.

Both it and Quantum Mechanics are spectacularly successful theories in their own rights. But for nearly a century, all attempts to merge the two have been more or less failures. Plus, our current worldview includes dark energy and dark matter which make up more than 90% of everything there is in the universe. The fact that this is all down is not understood tells me that maybe we don't know everything about gravity already.

When more tests are planned to further prove the General Relativity and determine if there's a test it can pass, a cold cesium atom clock is set to fly in the International Space Station, and it aims to reduce the deviation by a further factor of 10. As for the satellites, their orbits were made more circular using the propellants on board, though they are still elliptical.

The navigation signals have been tested and are within acceptable parameters. For the moment, they are restricted to 'test mode' awaiting some new software and modifications on the ground, but the hope is that they will be useful for navigation after all. And in the meantime, they hold the record for achieving the best test yet of General Relativity.

Hey, this episode of Veritasium was supported by viewers like you on Patreon and by Audible. For viewers of this video, Audible is offering a free audiobook when you start a 30-day free trial. Just go to audible.com/veritasium or text veritasium to 500500.

Now, the book I wanna recommend to you today is called 'Atomic Habits' by James Clear. We are approaching that time of year when everyone makes New Year's resolutions and wants to change some big things about their lives. But what this book recommends is actually making tiny changes consistently is the best way to achieve the long-term changes that you wanna see in your life, and I have found this book already, just a few chapters in, umm, really effective in terms of changing my thinking about how I wanna change my life and the things I wanna get done.

So I highly recommend it, and you can download it for free by going to audible.com/veritasium or text veritasium to 500500. But Audible doesn't just have audiobooks; they also have audio fitness programs and Audible originals, and every month, you can get 2 Audible originals from a changing set of titles. Those are audio programs made specifically for Audible members.

The way membership works is that every month you get a credit which is good for one audiobook, and it also gives you exclusive access to those other fitness programs and to Audible originals, so if you wanna check it out, go try audible.com/veritasium. I wanna thank you to Audible for supporting me, and I wanna thank you for watching.