The Line That Would Not Stop Rising: The Keeling Curve

One stubborn instrument, one clean mountaintop, one number measured the same way for sixty years. The result is the most consequential graph in science.

In March 1958, a young geochemist named Charles David Keeling started measuring the amount of carbon dioxide in the air at a weather station near the summit of Mauna Loa, the great volcano in Hawaii. He did not set out to document a planetary emergency. He had built an unusually precise instrument and wanted to know whether the concentration of carbon dioxide was the same everywhere and steady over time, a basic question nobody had answered with good data. The record he began that year is now one of the most important graphs in science, and it has been running, with barely a gap, ever since.

The choice of Mauna Loa was deliberate. Keeling needed air that represented the whole atmosphere rather than the exhaust of a nearby city or the breathing of a nearby forest. High on a remote Pacific volcano, thousands of kilometers from major land masses and well above the local vegetation, the air is thoroughly mixed and clean, a fair sample of the Northern Hemisphere as a whole. When the first numbers came in, the concentration was about 315 parts per million, meaning that for every million molecules of air, roughly 315 were carbon dioxide.

Almost at once the data showed something nobody had measured before: the air breathes. Over each year the concentration falls through the northern spring and summer, then climbs back through autumn and winter, tracing a regular sawtooth. The cause is the land plants of the Northern Hemisphere, which hold most of the world’s vegetation. As they leaf out and grow in summer they pull carbon dioxide out of the air, and as they die back and decay in winter they give it up again. The whole hemisphere inhales and exhales once a year, and Keeling’s instrument was sensitive enough to watch it happen.

Underneath that seasonal wobble was the finding that mattered. Year after year, each winter peak sat a little higher than the last, and each summer trough did too. The line was not flat, and it was not noise. Carbon dioxide was rising, steadily, and the longer Keeling watched the clearer the trend became. By the time he had a couple of decades of measurements, the upward march was undeniable. That rising line, the seasonal zigzag riding on a relentless climb, is what we now call the Keeling Curve.

There was never much doubt about where the extra carbon was coming from, and the air carries its own evidence. Carbon comes in slightly different forms, or isotopes, and the carbon released by burning coal, oil, and gas has a distinct isotopic signature, because it comes from plants that lived millions of years ago. As fossil carbon dioxide accumulates, it shifts the isotopic mix of the atmosphere in exactly the way measurements show it shifting. The fingerprints match. The rise is the product of burning fossil fuels, with deforestation adding to it.

The numbers from recent years make the early measurements look almost quaint. The preindustrial level, before large-scale fossil fuel use, was around 278 parts per million. Keeling found 315 in 1958. In 2024 the annual average stood near 422 parts per million, with the seasonal peak in May reaching just under 427, and the jump from 2023 to 2024 was the largest single-year increase in the entire record. The line is not only rising. The rate at which it rises has itself been speeding up. Carbon dioxide is now more than fifty percent above its preindustrial value, a change driven entirely within a span of time shorter than a long human life.

Keeling fought for years to keep his measurements funded, since long-term monitoring is the kind of patient, unglamorous work that grant committees are forever tempted to cut. He won those fights often enough to keep the record unbroken, and after his death in 2005 his son Ralph took over the program at Scripps. The result is a continuous, internally consistent measurement of the planet’s atmosphere going back more than sixty years, the backbone against which all other carbon dioxide data are checked.

What gives the Keeling Curve its authority is its sheer simplicity. It is not a model or a projection or a contested reconstruction. It is a direct measurement of the air, taken the same careful way in the same clean place, decade after decade, by an instrument doing the most basic thing an instrument can do: counting. The curve makes no argument. It just records, year after year, a number that keeps going up.

The curve makes no argument and needs none. It is the rare piece of science whose authority rests entirely on the patience of the measuring.