March 1958: Charles Keeling begins long-term measurements of atmospheric CO₂ on Mauna Loa

One of the most important data collection projects on the planet is located on top of the world’s largest active volcano, Mauna Loa on Hawaii’s Big Island. Measurements of atmospheric carbon dioxide concentrations were started there by Charles Keeling and have been interrupted only twice, once due to funding issues, and once due to natural causes: In November 2022, lava blocked off access and power to the Mauna Loa Observatory.

But the active volcano was the least of Keeling’s problems. Over its 67-year history, the Keeling curve has been vulnerable to changes in policy at funding agencies and shifting political winds. Keeling, who died in 2005, spent his entire career measuring atmospheric CO₂, and plenty of time and energy fighting to keep the project afloat.

Keeling’s impact has been profound. “These measurements have revolutionized our thinking about how the Earth functions,” says Rob Jackson, an earth systems scientist at Stanford University.

The Keeling curve revealed that the biosphere breathes. In the Northern Hemisphere, atmospheric CO₂ rises in winter, when photosynthesis is slowed, peaking at the start of the spring. When leaves emerge, photosynthesis accelerates, drawing down CO₂. The Keeling curve shows this annual inhalation and exhalation of CO₂, a seasonal change in the planet’s atmosphere that was previously invisible. Keeling’s data also help show that this effect is less pronounced in the Southern Hemisphere — and that the planet’s rotation contributes to relatively quick atmospheric mixing from east to west, and slower mixing along the north-south axis.

Keeling curve data also helped establish that fossil fuel emissions are increasing atmospheric CO₂, and that the planet’s land and oceans are not able to take up enough of the gas to counteract human emissions.

“We use these data all the time — they are the foundation of our analysis of emissions,” says Jackson, who chairs the Global Carbon Project, an international scientific team that monitors greenhouse gas levels and their sources. The latest Keeling curve data show record growth in CO₂ emissions in 2024. Last year, atmospheric CO₂ concentrations reached 422.5 ppm, 52% above the preindustrial level of about 278 ppm in 1750.

Though these data are central to earth and climate science, having a long-term, continuous CO₂ record was not inevitable.

In the early 1950s, there wasn’t much interest in atmospheric CO₂. At the time, scientists thought concentrations of the gas were variable. Measurements were sparse, and sometimes unreliable. Keeling became interested in the problem when he was working as a postdoc in geochemistry at Caltech in Pasadena. To calibrate measurements for a study of carbonate rocks in Big Sur, he wanted to make sure he wasn’t making assumptions about ambient CO₂ concentrations.

“Published values of atmospheric CO₂ concentration varied widely,” Keeling recalls in an essay published in 1998. He decided to see for himself. His advisor supported his interest in CO₂, so Keeling built a pressure-based instrument called a manometer for measuring the gas. He also collected air and water samples at his study site every few hours throughout the day and night —though he had no reason to do so.

“The reason was simply that I was having fun,” he writes. He was 27, and he was enjoying being in Big Sur State Park, a beautiful area where redwood forests meet cliffsides above the foamy blue-green waters of the Pacific Ocean. At the suggestion of a colleague, Keeling saved samples for isotopic analysis. “I did not anticipate that the procedures established in this first experiment would be the basis for much of the research that I would pursue over the next forty-odd years,” he writes.

His early measurements hinted at what were then unexpected patterns. Afternoon CO₂ concentrations were relatively constant, while previous research suggested they would vary widely. Keeling also saw a diurnal pattern. Concentrations were higher at night, and his isotopic measurements suggested this was due to CO₂ release from soil and plants.

Keeling’s findings came to the attention of Harry Wexler, the head of research at the U.S. Weather Bureau (now the National Weather Service). Wexler had just overseen the construction of an observatory on Mauna Loa in 1956 and was looking for projects to house there. Despite the risk, the volcano is a good place to measure CO₂. Mauna Loa’s elevation and remote location provide a relatively clean, well-mixed sample of the atmosphere, with less contamination from industrial activity and traffic than would be found in the continental U.S. And the harsh volcanic landscape means the observatory is not influenced by emissions from local plant life.

Wexler suggested to Keeling that he measure CO₂ at the new Mauna Loa Observatory, and in other locations around the world, as part of the International Geophysical Year, a collaborative scientific interchange in 1957 and 1958. (The United Nations has declared 2025 the International Year of Quantum Science and Technology.) Keeling got a job at the Scripps Institute for Oceanography at the University of San Diego, where the CO₂ program is based to this day, and rushed to design and kick off the program.

Keeling pushed for the project to use infrared spectrometers, which became commercially available after World War II, because the instruments provided accurate, continuous readouts of gas concentrations. “Most of Keeling’s seniors thought that such instruments were more costly than anyone needed to measure something that fluctuated so widely as atmospheric CO₂ levels,” writes Spencer Weart in the website accompanying his book The Discovery of Global Warming. “Yet the IGY money pot was big enough, and Keeling persuasive enough, to get Wexler to dig up funds to buy the spectrophotometers.” They were installed in Hawaii and at a station in the Antarctic. On March 29, 1958, Keeling’s first Mauna Loa reading came in, measuring atmospheric CO₂ concentration at 313 ppm.

After just one year of data collection, both stations showed a rise in CO₂ — a surprise. Keeling and his colleagues kept collecting data, and began publishing papers detailing and interpreting it. But as he notes in his 1998 essay (appropriately titled “Rewards and Penalties of Monitoring the Earth”), the CO₂ monitoring program came in and out of fashion with various funding agencies.

There’s a small gap in the Keeling curve in 1964, when instruments broke down after funding cuts. Keeling’s measurements in Antarctica ceased. In the 1970s, interest in global warming attracted government attention to the project, and proposals that it be transferred to a government agency — moves Keeling successfully resisted. “One more year,” funders frequently warned.

Weart estimates that all this attention and government and scientific labor was expended over a research project that cost only about $200,000 a year — a drop in the bucket of the U.S.’s federal budget.

“Why Go On?” asks a heading in Keeling’s biographical essay. He answers, “the data gathered in my program became more fascinating as the records lengthened.” And he wanted to stay involved to make sure those data were high-quality, having found issues with the methods being pitched and developed by the federal agencies that proposed to take over CO₂ monitoring.

The Scripps CO₂ Program continues to this day, under the auspices of Ralph Keeling, Charles’ son. The curve is bolstered by measurements performed by U.S. government agencies and other scientific teams around the world. But the Mauna Loa Observatory and the Keeling curve have unique scientific importance, in part due to the long-term continuity of the data, says Jackson.

“This is the most unique and perhaps the most important dataset in earth science,” he says.