Spoiler Alerts

Think about this: every piece of information you’ve ever heard about the cosmos — the Sun’s temperature, the age of the universe, the composition of a distant star — none of it came from collecting samples. We can’t scoop up a piece of a star. We can’t fly to the edge of the universe with a tape measure. We’re stuck here on Earth (mostly), looking up at points of light.

And yet, somehow, we know that:

• The Sun’s surface temperature is about 5,800 K
• The nearest star is 4.2 light-years away
• Stars are mostly hydrogen and helium
• The universe is about 13.8 billion years old and expanding

How is any of that possible? The answer is the thesis of this course:

Every gorgeous astronomical image you’ve ever seen is actually a dataset. The colors, the shapes, the bright spots and dark lanes — all of it encodes physical information. The job of an astronomer is to decode that information using physics and mathematics.

So astronomy is inference under constraint: we never touch our subjects, and we work from a handful of signals.

The Four Things We Can Actually Measure

Here’s the humbling truth that makes astronomy both challenging and beautiful: from our cosmic vantage point, staring at points of light billions of kilometers away, we can directly measure only four types of things.

Reading tip: treat the tables in this reading as reference maps — don’t memorize them. You’ll revisit each one when it matters.

Four direct observables

Let’s sharpen what each one means:

Brightness is the rate at which light energy arrives at your detector — photons per second per unit area. We literally count photons. This is the raw signal, and it depends on two things we can’t disentangle from brightness alone: how luminous the source truly is, and how far away it is. A dim nearby candle and a brilliant distant lighthouse can appear equally bright.

Position is where an object appears on the sky, measured in angles. This is purely geometric — we’re mapping locations on a celestial sphere. But position becomes powerful when we track changes: the annual wobble from Earth’s orbit (parallax), the slow drift across the sky (proper motion), or the periodic wobble caused by an unseen companion.

Wavelength is the key to spectroscopy. By spreading light into its component wavelengths, we transform a single brightness measurement into thousands of data points — brightness as a function of wavelength. This spectral information encodes temperature (through the overall color), composition (through absorption and emission lines), and motion (through Doppler shifts).

Timing is special: it’s both a measurement in its own right (when did this pulse arrive?) and a way to extract additional information from the other three. A star whose brightness varies periodically tells us something a constant star doesn’t. An object whose position shifts tells us about motion. Timing transforms static snapshots into dynamic movies.

What’s NOT on This List

Notice what you cannot directly measure:

Every single one of these fundamental properties must be inferred by combining measurements with physical models. The gap between “what we measure” and “what we want to know” is bridged by physics. That bridge takes the form of models — mathematical relationships that connect observables to physical quantities. We’ll formalize this in Section 1.3.

A Taste of What “Decoding” Means

Here’s a concrete example of the decoder ring in action.

Suppose you observe two stars of the same type (same intrinsic properties), but one appears 4× fainter than the other. Why might that be?

If they’re truly identical, the fainter one must be farther away. But how much farther? This is where physics enters. Light spreads out as it travels, and the amount of light you receive decreases with the square of the distance. So if Star B is 4× fainter than Star A, Star B must be 2× farther away (because $2^2 = 4$).

That’s one measurement (relative brightness), one physical law (the inverse-square law), one inference (relative distance). This is what “decoding” means.