A work in progress: I'm confident in the headline number and less so in some of the details I've written up so far. The isochron-maths section is still a stub (more on that below) — that's the main gap.

The Earth is about 4.54 billion years old, give or take roughly 1%. That number is one of the great settled facts of science — and the way we got it is the whole Kronion idea in miniature: time, measured with ions.

Why you can't just count the rocks

The intuitive method — count rock layers, or find the oldest rock you can — fails badly, because Earth erases its own history. Plate tectonics, erosion, and melting constantly recycle the crust. The oldest intact rocks are about 4.0 billion years old, and the oldest mineral grains we have — tiny zircon crystals from the Jack Hills of Western Australia — are about 4.4 billion years. Both are younger than the planet. We need a clock that doesn't get reset.

The clock: radioactive decay

Some atoms are unstable. A parent isotope decays into a stable daughter at a rock-steady rate set by its half-life — the time for half of any sample to convert. Crucially, that rate is immune to heat, pressure, and chemistry. Two clocks matter here, both built on uranium turning into lead (an atom changing identity — the literal ion at the heart of the name):

• ²³⁸U → ²⁰⁶Pb, half-life ≈ 4.47 billion years
• ²³⁵U → ²⁰⁷Pb, half-life ≈ 0.70 billion years

Measure how much daughter has accumulated relative to surviving parent, and you read off the elapsed time.

The meteorite trick

Here's the elegant part. We don't date the Earth's surface — it's too disturbed. We date meteorites, specifically the iron meteorite from Arizona's Meteor Crater (Canyon Diablo). The reasoning:

1. The Earth, the meteorites, and everything else in the Solar System condensed from the same cloud at essentially the same time.
2. So dating an undisturbed leftover from that event dates the Earth too.

In 1956, Clair Patterson did exactly this and got 4.55 billion years — a number that has barely budged in seventy years of refinement. Moon rocks returned by Apollo, dated independently, agree at about 4.5 billion years.

Earth ≈ 4.54 ± ~0.05 billion years. The load-bearing evidence is the convergence of independent methods — meteorites, Moon rocks, and Earth's own lead — not any single date.

Stub: how an isochron actually works

The honest gap in this draft. The real measurement doesn't assume you already know the starting amount of daughter isotope — it cleverly extracts both the age and the initial conditions from the slope of a line through several samples (an isochron). I haven't written this up properly yet, and it's the part that turns "trust me" into "here's why." It's the first thing I want to fix.

What's solid, and what's still argued

• Solid: the ~4.54-billion-year age, the radiometric method, and the meteorite logic.
• Still active research: the fine timeline of Earth's first ~100 million years — exactly when the planet finished accreting, and precisely when the Moon-forming giant impact happened. Estimates for the impact cluster around 4.5 billion years ago, but the precise date and dynamics are genuinely debated.

Next on the bench

1. Write the isochron section with a worked example (top priority — it's the biggest gap).
2. Add the Jack Hills zircon story as its own short aside.
3. Add references for Patterson (1956) and a current best-estimate review.

Changelog

• 2026-06-15 — Added the Moon-rock cross-check and the "three independent lines" framing.
• 2026-03-03 — Started; isochron section left as a deliberate stub.