Memunic Space Explorers › Book 12: Memoon and the Biggest Question › Chapter 6: Rewinding the Film
Chapter 6: Rewinding the Film

In this chapter: Use the Hubble constant to estimate a time since expansion began, explain what a scale factor is, give an uncertainty for a fitted value, and compare two published values of H0 using their error bars.
On the sixth evening Simbha had a surprise. He had spent the afternoon rolling a long strip of paper across the terrace floor and drawing on it with chalk, and now he sat beside it with the look of a lion who has done something clever and is trying not to say so.
“The film,” he said.
The strip was a row of frames, each showing five dots in a line. In the last frame, on the right, the dots were as far apart as the five galaxies in the table. In each frame to the left, the dots were closer. In the leftmost, they were all in one place.
“I did not draw the middle ones by guessing,” said Simbha. “I let each galaxy keep its speed from the table and walked it backwards.”
“If each galaxy kept its speed all the way back, how long before they met?” said Memoon. She was already writing. “Time equals distance over speed. We have the numbers.”
Maths track: 1 Mpc = 3.086 × 1019 km. Galaxy A: t = (30 × 3.086 × 1019 km) / 2,300 km/s = 4.0 × 1017 s = 12.8 billion years.
Simbha and Beautine unroll the chalk-film strip
Galaxy E: t = (200 × 3.086 × 1019 km) / 14,500 km/s = 4.3 × 1017 s = 13.5 billion years.
“They are almost the same,” said Beautine, stunned. “The nearest and the farthest. They all arrive at the same moment.”
“They have to,” said Kachui. “If speed is proportional to distance, then distance over speed is the same for every galaxy. It is just one over the Hubble constant.”
t(H) = 1 / H0. With H0 = 71.2 km/s/Mpc = 71.2 / (3.086 × 1019) per second = 2.31 × 10−18 per second, so 1 / H0 = 4.3 × 1017 s = 13.7 billion years.
“It has a name,” said Memoon. “The Hubble time. Careful, though. It tells us how long we would need if the speed never changed. That is exactly the assumption we do not believe.”
“Why not?” said Noodles.
“Gravity pulls everything together, so in the early Universe the expansion should have been slowing down. And then, as we will see, something has been pushing it faster in the last few billion years. The speeds changed. So the Hubble time is a scale, not the age. Look at what happens with a different number.” She wrote: With H0 = 67.4, 1 / H0 = 977.8 / 67.4 = 14.5 billion years. But the full model, which keeps track of how the speed changed, gives 13.8 billion years.
“Five per cent off,” said Kachui. “That is what a changing speed does.”
“Is there any other clock to check it against?” Simbha asked. “Rewinding a film is a model. I would like something I can count.”
“There is,” said Kachui. She flipped to a page with a list. “Stars. The Universe cannot be younger than the oldest objects in it. The Hubble Space Telescope found some very old white dwarf stars in a globular cluster that are between twelve and thirteen billion years old. Other methods agree. The first stars formed within a billion years of the start. So the two clocks, the stretching and the stars, both give around thirteen or fourteen billion years, even though they use completely different physics.”
“Two independent clocks agreeing,” said Memoon. “That is what we mean by KNOWN.”
“And I have a sketch,” said Beautine, who had been trying to say something for a while. “If I rewind the film, the whole thing is the size of a point. So we all came from one point, in the middle, and everything exploded outwards, which is what I said.”
“Not quite,” said Nimo, drifting up. “Remember the band. When we rewind it, every dot gets closer to every other dot. They do not get closer to a place. The band shrinks, and everything in it shrinks together.”
Memoon drew a short line on the page. “We need a way to say how big the Universe was. We call it the scale factor, a. Today it is one. When light left a galaxy with redshift z, the scale factor was a = 1 / (1 + z).”
z = 0.0483: a = 0.954. z = 1: a = 0.5. z = 3: a = 0.25.
“A galaxy at redshift one,” said Memoon, “sent its light when the Universe was half its present size. Distances were half. And the matter was eight times denser, because volume goes as the cube.”
Beautine blinked. “Is that how big the Universe is? Half?”
“That is how big it was. Not how big it is. Not even close.” Kachui coughed. “Here is a puzzle. The Universe is about 13.8 billion years old. Light that left the farthest things we can see has been travelling for about that long. So you would think we can see out to 13.8 billion light-years. But the edge of what we can see is about 46 billion light-years away, in today’s distances.”
“That is impossible,” said Beautine, in the voice of a squirrel who had just been shown a tree with its roots in the air.
“It is the band again,” said Memoon. “Imagine an ant crawling at a steady speed along a band that is being stretched. The ant covers some distance. But all the while, the band under its feet grows. By the time it arrives, the place where it started is farther away than the distance the ant has walked. The light is the ant. The galaxy it started from is now much farther than 13.8 billion light-years.”
Mr Rao, from the stairs, lifted a finger. “Then speed matters. The galaxies far away are receding faster than the speed of light? I have read that. It cannot be right.”
“It is right and it is not a contradiction,” said Memoon. “Let us see where it happens.”
Maths track: v = H0 d = c when d = c / H0 = 3.00 × 105 / 71.2 = 4,200 Mpc, about 13.7 billion light-years.
Hubble's law: farther galaxies recede faster
16,000 E
12,000 recession speed (km/s)
D
8,000 C
best-fit line through the origin: slope 71.2 km/s per Mpc B 4,000 A
0 50 100 150 200
distance (Mpc)
Invented practice data for five galaxies. The slope is rounded; its uncertainty from this scatter is about ±1.2.
Hubble’s law for five invented galaxies. The slope is rounded.
“Beyond that distance, the stretching adds up to more than the speed of light,” said Kachui. “Nothing is racing through space faster than light. The rule against that is about moving through space. Here the space between is growing. Whether light from such a galaxy can ever reach us depends on how the expansion rate changes over time, which is a delicate matter. We will leave it there.”
“I feel dizzy,” said Beautine happily.
“Then here is something solid to hold on to,” said Mr Rao. “I have two numbers.” He held up the tablet. “I looked at the papers while you were all drawing. There are two groups who measure this Hubble constant very carefully. One group uses the glow from the early Universe, which we have not met yet, and a model, and gets 67.4, give or take 0.5. The other uses a ladder of distances: pulsating stars, then exploding stars, and measures the expansion today. They get 73.2, give or take 0.9.”
“They disagree,” said Kachui at once.
“Do they, though?” said Memoon. “Each has an error bar. What does it mean?” She waited for the answer, and when nobody gave one, she gave it herself. “The uncertainty is a statement that we have repeated the measurement many times in our heads and got scatter. The true value is probably within one or two uncertainties. If the gap is much bigger than the uncertainties, something is wrong.”
Maths track: Difference = 73.2 − 67.4 = 5.8 km/s/Mpc. Combined uncertainty = √(0.52 + 0.92) = √(1.06) = 1.03. Gap / combined uncertainty = 5.8 / 1.03 = 5.6.
“The gap is about five and a half times the combined error,” said Kachui, slowly. “If the two measured the same thing and the errors were independent and bell-shaped, a gap that big would arise by chance far less than once in a million.”
“This is the Hubble tension,” said Memoon. “It is one of the real open problems in astronomy at the moment. Some groups using other methods get values in between. One group using the brightest red giant stars gets about 70, and says there is no problem. Other groups disagree with their analysis. It is not settled.”
“Which one is right?” said Noodles.
“That is the wrong question to start with,” said Memoon. “There are three ways it can go. One: a hidden systematic error in one of the methods. Two: the model used to turn the early glow into a value today is incomplete, and the true history of the Universe differs from our simple version. Three: both measurements are fine and we have misjudged one uncertainty. We do not know which. We write both numbers, with error bars, and we say we do not know.”
“I like the second one,” said Simbha. “It would mean there is something new.”
“You like it because it is exciting,” said Kachui, not unkindly. “That is not evidence.”
Nimo’s bubble bobbed. “A picture, if I may. Two people want to know how fast a river runs. One stands at the spring, uses what she knows about rivers to predict the speed downstream. The other drops floats in the water here and times them. If they disagree, either her model is missing something between the spring and here, or his timing is off. My river picture fails where a real river has one speed at one place. The Universe’s expansion rate has changed over time, and what each person measures is a different slice of it.”
“Now ours,” said Kachui. “We have five galaxies. We got 71.2. What is our uncertainty?”
They worked it out together. For each galaxy they found the gap between its measured speed and what the line predicts, squared the gaps, added them up, and divided by four, because five points minus one fitted number leaves four degrees of freedom.
Maths track: Residuals (km/s): +165, -270, +284, -474, +268. Sum of squares = 4.77 × 105. s = √(4.77 × 105 / 4) = 345 km/s. Uncertainty of slope = s / √(sum of d2) = 345 / √(77,000) = 345 / 277 = 1.2 km/s/Mpc.
Result: H0 = 71.2 ± 1.2 km/s/Mpc.
“Our number sits between the two,” said Kachui. “It is less than three combined uncertainties from the first and about one from the second. With only five galaxies we cannot decide anything. We can say that our data are not inconsistent with either one.”
“A careful sentence,” said Memoon, “and a useful one.” She wrote three lines.
KNOWN: Two careful methods give different values of H0, and the gap is much larger than their stated errors.
HINT: A hidden error, or an incomplete model, or both.
GUESS: Which one. We do not know.
Beautine was quiet for a while. “So when we rewind the film, we go back to a time when everything was close together. And hot?”
“Squeeze anything and it heats up,” said Noodles, who knew that from a bicycle pump.
“If the early Universe was hot, it glowed,” said Memoon. “And if that glow was released when the Universe was still small, then it has been stretching, like everything else, ever since.” She turned back a few pages in the notebook, to the line she had underlined on the second evening. T = 2.9 K. “Remember this? A one-millimetre peak.”
“Beautine asked why anything would be at 2.9 kelvin,” said Kachui.
“Mr Rao,” said Memoon, “you said you had a story about a pigeon.”
“I did,” said Mr Rao, pleased. “I will tell it tomorrow, if there is a biscuit.”
Science Corner The Hubble time 1 / H0 is the time it would take to rewind the expansion if the speed had never changed. For H0 = 71.2 km/s/Mpc it is 13.7 billion years. The full model, which tracks how the speed changed, gives about 13.8 billion years, and the oldest stars measured are 12 to 13 billion years old. The scale factor a is the size of the Universe relative to today: a = 1 / (1 + z). Distances scale with a, and matter density scales as 1 / a3.
The observable universe extends about 46 billion light-years in today’s distances, even though the Universe is 13.8 billion years old, because space expanded while the light was travelling. The uncertainty of a fitted slope through the origin is s / √(sum of d2), where s is the scatter of the points. Two values with uncertainties can be compared by the gap divided by the combined uncertainty, √(sigma12 + sigma22), if the errors are independent. The Hubble tension: the Planck-based value, 67.4 ± 0.5 km/s/Mpc, and the distance-ladder value, 73.2 ± 0.9 km/s/Mpc, differ by about 5.6 combined uncertainties. Other groups get values near 70. As of late 2026 it is unresolved. Story note: The Planck value is an inference from the early-Universe glow using a model; the distance-ladder value uses nearby standard candles and needs no such model. The ’less than once in a million’ statement assumes independent bell-shaped errors, which hidden systematic errors do not follow. The Hubble time is not the age, and c / H0 is not the edge of the Universe; recession faster than light involves growth of space, not motion through it, and which distance to quote for a far galaxy depends on the definition (here, the distance in today’s space). The ant-on-a-band picture fails because the ant moves at a fixed speed relative to the band, while light’s speed is fixed locally; it shows only that the source ends up farther than the distance covered. The Club’s 71.2 ± 1.2 comes from five invented points with four degrees of freedom, so the uncertainty is itself uncertain. Nimo’s river picture fails because the expansion rate itself changes with time.
Space Detective 1. Explain why 1 / H0 is not exactly the age of the Universe, and name one independent check on the age. 2. A galaxy has redshift z = 0.5. Find the scale factor when its light left, and how many times denser the matter was then than now. 3. A new measurement gives H0 = 70.4 ± 1.9 km/s/Mpc (an invented value). How many combined uncertainties is it from 67.4 ± 0.5, and from 73.2 ± 0.9? What can you conclude?
Try it: Two Ways to Measure Time: 30 minutes. Adult: An adult should stand clear of the band’s path and check the band for wear before each stretch. You need: the rubber band from the previous activity, a ruler, a stopwatch, a phone with slow-motion video.
Do: Mark two dots 10 cm apart. Stretch the band so that those two dots move to 20 cm apart. Time the stretch five times with a stopwatch, and film five more with slow-motion video and read the time from the frames. For each set, find the mean time and the standard error (the scatter divided by the square root of five). Divide the gap between the two means by the combined standard error. What to expect: The two means are close but may differ more than their random errors suggest, because reaction time is a systematic error in the stopwatch. A gap larger than about two combined errors points to a systematic effect, not bad luck. Cloudy or low-resource day: This activity is indoors. With no phone, use two different people with stopwatches. Limitation: two people share the same method, so a reaction-time bias common to both would not show up.
Next time: If the early Universe was hot, its glow should still be arriving. Tomorrow Mr Rao tells the Club about an antenna, some pigeons, and a signal that would not go away.
← Chapter 5: The Stretching Light · Book 12 contents · Chapter 7: The Oldest Light →
Keep exploring with Memunic Space Explorers
Continue with Book 12: Memoon and the Biggest Question, or see all twelve books in the series. Find more Memunic ebooks on MemuniCity.
Comments
Post a Comment