How We Measured the Universe Without Leaving Earth（原声文稿）

No tape measure reaches another galaxy.

So how do we know how big the universe is, and how fast it is expanding?

The answer is a chain of measurements, each built on the last. And that chain has a crack in it.

Over the next 12 minutes, we'll travel from the moon to the edge of the universe, who invented each method, how it works, and why the two best ones disagree.

There are two ways to measure how fast the universe is expanding today.

One measures objects near us and gets close to 73.

The other works from the oldest light in the universe and gets about 67. The bigger the number, the faster the expansion.

It's like two survey teams measuring the same bridge, each confident in their work, and disagreeing by almost 10%.

That disagreement may be a sign of a gap in our understanding of the universe. To see why, let's go back to the beginning.

Around 240 BC, Eritasthanes heard that at noon on the summer solstice in the town of Sayin, today's Aswan, sunlight reached the bottom of a well.

At the same moment in Alexandria, a stick still cast a shadow about 7 degrees, 150th of a circle.

So the Earth must be 50 times the distance between the two cities around, by the usual conversion, roughly 40 ,000 km.

Using the Earth's radius as his unit, Hipparchus used its shadow on the moon during lunar eclipses to estimate the moon's distance, 60 to 70 Earth radii.

The true value is 60.

In 1969, Apollo 11 left a mirror on the moon.

A laser pulse bounces back in two and a half seconds, 384 ,000 km.

Today, that measurement is accurate to millimeters. The Sun was much harder.

The trick, two observers, far apart, look at the same planet and measure how much it shifts.

In 1672, Cassini in Paris and Richer in French Guiana did this with Mars and put the Sun about 140 million kilometers away.

Halley proposed something better, timing Venus as it crosses the Sun's face.

In 1761 and 1769, astronomers sailed the world to do it, Captain Cook among them.

Guillaume Les Gentiles missed the first because of war.

He waited eight years for the second and was clouded out.

When he got home, his His family had declared him dead.

In 1961, radar echoes from Venus settled it, 150 million kilometers.

Light takes 8 minutes and 20 seconds to cross it.

Hold up a finger and close each eye in turn. It jumps. That's parallax.

For stars, the two eyes are Earth on opposite sides of its orbit, 300 million kilometers apart.

But the shift is so small that for two centuries nobody could measure it.

In 1838, Friedrich Bessel did the size of a coin seen from over 10 kilometers away.

That star is about 11 light -years away.

If Earth were one meter from the Sun, the nearest star would be 270 kilometers away.

Europe's Gaia spacecraft measured nearly 2 billion stars this way, precise enough to see a hair from a thousand kilometers.

But beyond a few thousand light years, the shift is too small to measure precisely, and the sky holds faint smudges that might be much farther still. Nobody knew what they were.

Think of a light bulb, twice as far away, it looks four times fainter, three times as far, nine times fainter.

If you know how bright a bulb really is, how faint it looks tells you how far away it is. astronomers call this a standard candle.

But how do you know how bright a star really is?

At Harvard, a team of women was hired to study photographs of the sky for just 25 to 30 cents an hour.

In 1912, one of them, Henrietta Leavitt, found that pulsating stars called Cepheids follow a rule. The slower they pulse, the brighter they are.

It was a light bulb that tells you its own wattage.

In 1913, Agenar Hertzsprung used nearby parallaxes to calibrate it.

In 1923, Edwin Hubble found a cepheid in a faint smudge called Andromeda.

On the photographic plate, he wrote, Var!

By his measurement, it was nearly a million light years away.

Andromeda was another galaxy, and the universe suddenly got much bigger.

In 1927, George Lemaite argued the universe is expanding.

In 1929, Hubble, using speeds measured largely by Vesto Slifer, found the evidence, the farther a galaxy, the faster it recedes. The speed comes from redshift.

The color lines in a galaxy's light are shifted toward red, like a siren dropping in pitch as it drives away.

For distant galaxies, it is space itself stretching the light.

Like dots on an inflating balloon, the farther apart, the faster they separate.

Today, it is often called the Hubble -Limetre law.

That rate is the Hubble constant.

The 73 from the start of this video means a galaxy a hundred million light -years away is receding at about 2 ,200 km per second.

But Hubble got 500, which made the universe about 2 billion years old younger than the Earth. In 1952, Walter Bade found the mistake.

There are two kinds of Cepheid, and Hubble had misread the label on the bulb.

Every galaxy was twice as far away, the universe twice as old, and the paradox began to unravel.

Every engineer knows that lesson, and error, low in the chain, carries all the way up.

Sephiods fade from view beyond about a hundred million light years.

Further out, we need Type IA supernovae, exploding white dwarf stars.

Once corrected, they explode with nearly the same brightness as bright as their entire galaxy, like stadium floodlights in the dark.

To calibrate the floodlight, astronomers find galaxies with both Sephiods and a supernova like a room with two kinds of lamp.

The Cepheids give the rooms distance, and that gives the floodlights true power.

In 1998, two teams using these floodlights discovered the expansion is speeding up. It won the 2011 Nobel Prize.

Parallax, Cepheids, Supernovae. Together they form the cosmic distance ladder.

In 2001, Wendy Friedman's team used Hubble to get 72, give or take 8, right, so the true value could be anywhere from 64 to 80.

By 2022, Adam Reese's shows team had it down to 73, give or take 1.

But there is a completely different way, with no latter at all.

In 1964, at Bell Labs, Arno Penzias and Robert Wilson picked up a hiss in their antenna. It came from every direction.

They even blamed pigeon droppings and cleaned them out. the hiss stayed.

It was the cosmic microwave background, the afterglow of the Big Bang predicted back in 1948 by Ralph Alfer and Robert Herman.

It's the oldest light we can see, released when the universe was about 380 ,000 years old. That light has traveled for 13 .8 billion years.

Because space kept stretching along the way, the place it came from is now some 46 billion light years away.

the edge of the observable universe.

The Kobe, WMAP, and Planck satellites mapped it, revealing tiny, hot, and cold spots that differ by only about one part in 100 ,000. Why are the spots there? The early universe was a hot fog.

Gravity pulled matter together, light pushed it apart, and the gas kept pulsing like sound waves, like ripples spreading on a pond after you throw in a handful of pebbles.

As early as 1970, physicists in the United States and the Soviet Union predicted that those ripples would leave exactly such spots.

The waves traveled at more than half the speed of light.

After 380 ,000 years, the fog cleared, and the ripples froze in place.

Knowing how fast they ran and for how long, physics can work out how big the spots should be. That makes them a standard ruler.

On the sky, the spots appear about one degree across, roughly twice the width of the full moon.

Like looking down at a football pitch from a plane, you know it's 105 meters long, so how small it looks tells you how high you are.

Knowing the true size and the size we see gives the distance to that fog.

Put it into our standard model of cosmology, the recipe for what the universe is made of and how it grows.

Run it forward to today and out comes the expansion rate, about 67.

The first works outward from where we stand, linking standard candles to measure the universe today.

The second works forward from the oldest light, using a standard ruler and a model.

If the model is right, they should meet at the same number, they don't.

Other methods split the same way, roughly 70 to 76 from the nearby universe, 67 or 68 from the early universe.

Astronomers call this disagreement the Hubble Tension.

For an engineer, the first instinct is to suspect the measurement, and there was an obvious suspect.

At great distances, Hubble sees each cepheid blended with neighboring stars, like one bulb inside a cluster of lamps.

The bulb looks brighter, so it seems closer, and 73 could be pushed too high. Then came the James Webb Space Telescope.

With a bigger mirror, an infrared vision, it can separate individual stars from the crowd.

In 2023, Ries' team checked with Webb.

The Safiids came out clean, and the answer was still 73.

Friedman's team used a different kind of bulb, red giant stars, and got between 68 and 70.

Webb also caught a supernova nicknamed Hope, split into three images by a galaxy cluster.

That independent measurement leans toward 73, though with large uncertainty.

In 2026, many groups pooled their data.

Like a proper engineering review, they fix their methods before looking at the numbers.

The result? 73 .5, uncertain by less than one. The microwave background still says 67.

The odds that this is just chance are vanishingly small.

When two independent checks disagree, an engineer doesn't pick the number they like.

They question the assumptions. The first method rests on few assumptions.

The second rests on a whole model of the universe.

If the rulers are right, maybe the blueprint needs fixing.

Perhaps the early universe held an extra kind of energy, making the sound rulers shorter than we think.

There are also tentative hints that dark energy, whatever is making the expansion speed up, the 1 ,998 discovery changes over time.

None of this is proven, and there may still be an error nobody has found, but after more than a decade, the gap is still there.

From a shadow 2 ,000 years ago to the James Webb Space Telescope, we've linked measurement to measurement out to the edge of the universe, and it's because we measured so carefully that we can see the crack.

The expansion rate also sets the age of the universe.

All else equal, those two numbers give ages about a billion years apart.

Next time, how do we know the age of a star, a galaxy, or the universe itself?