Deep Time The three numbers that decide it

The curves

Oxygen · carbon dioxide · temperature  ·  4,540 Myr on one axis

Every interval page on this site quotes these three for its own slice of time. Put them together and the argument stops needing words: oxygen crosses the line you can stay conscious above at roughly 550 Ma, and does not reach the line fire needs until roughly 420 Ma. The gap between those two numbers is why the middle of this site reads the way it does.

Four and a half billion years, three measures

Three panels rather than one, because these quantities share no common scale and stacking them on a single axis would invent a relationship that is not in the data. They share the time axis only.

Hover anywhere on the chart to read all three at that moment, or focus it and use ← → (hold shift to move faster). The axis has a deliberate break at 539 Ma: on a straight linear scale the whole Phanerozoic, which is every interval you could last more than a few minutes in, is squeezed into the last 12% of the width.

Table view: the same values as numbers
IntervalStartsOxygenCO₂Mean tempYou last

What to look at

01: The two oxygen lines

The top panel carries two rules. The lower, near 10%, is where a human stops staying conscious, and the curve spends four billion years underneath it. The upper, near 16%, is where fire starts to burn. They are crossed 130 million years apart, and treating them as one number, which this site did until recently, hides the more interesting half of the story.

02: The Carboniferous spike

Oxygen peaks near 35% around 305 Ma while CO₂ simultaneously bottoms out near modern values. Both are the same event: forests burying carbon faster than anything could rot it, which is the coal.

03: The Permian and Triassic dip

Oxygen collapses from 35% to about 15% while temperature spikes 10 °C. The two worst post-Cambrian survival verdicts on this site sit precisely in that trough.

04: The temperature notches

Four deep negative spikes: the Huronian, the two Snowball Earths, and the Hirnantian glaciation. Two sharp positive ones: the end-Permian at 252 Ma and the PETM at 56 Ma.

Where the numbers are soft

These curves are a synthesis, not a dataset. They are hand-built from published reconstructions and resampled onto a common grid so three quantities on very different scales can share a time axis. The shapes, the crossings and the order of events are the claim. Individual values are not precise and should not be quoted as measurements.
The oxygen curve before the Devonian is model output. It comes from geochemical carbon–sulfur modelling, and independent groups disagree by roughly a factor of two on the Cambrian and Carboniferous values. The Carboniferous peak in particular is quoted anywhere between 25% and 35% depending on method.
The Hadean is off the CO₂ scale. Early CO₂ may have reached tens of bar (millions of ppm), which no axis that also has to show 428 ppm can display. The CO₂ panel therefore starts at 4,000 Ma rather than 4,540.
A mean global temperature hides the distribution. Hothouse worlds are not uniformly hotter; they are flatter, warming the poles far more than the tropics. A single mean is the wrong number for asking whether a specific latitude was survivable, which is why the interval pages carry that detail and this chart does not.