Plate I

Fig. 1958·01

The Series

Since March 1958 the observatory has reported one number: the dry-air mole fraction of carbon dioxide, in parts per million. Read end to end, the record does one obvious thing. It climbs.

The trend line has moved from 314.44 ppm to 429.06 ppm, a rise of 114.62 ppm, or 36.5 percent, across 68.3 years. No smoothing hides it and no single year explains it. The climb is the whole shape of the page.

Figure 1. Monthly mean CO2, 1958 to 2026. Every point is one month. The rise is unbroken; the fine tooth on the line is the subject of Plate II.

Plate II

Fig. 1958·02

The Breath

Lay a straightedge along the climb and a second motion appears beneath it. Every year the number rises into a May peak, falls to an October low, then rises again.

Subtract the long trend from each monthly reading and only the yearly cycle is left. Averaged across all sixty-eight years the swing measures 6.33 ppm from trough to peak. It has kept nearly perfect time while the trend beneath it moved more than a hundred parts per million. This is one figure drawn at two scales: the same tooth, repeated sixty-eight times, riding a line it never interrupts.

Figure 2. The mean yearly cycle (bold) with four sampled years behind it: 1961, 1985, 2010, 2024. Sixty years apart, they trace almost the same curve.

Plate III

Fig. 1958·03

The Rate

The climb is not a straight line, and the difference matters more than the height.

In the 1960s the trend rose by 0.86 ppm each year. In the 2010s it rose by 2.40. The open 2020s are running faster still. The record is not only high. It is steepening, and the steepening has itself been steady, decade over decade.

Figure 3. Mean yearly change in the CO2 trend, by decade. Derived by differencing annual trend means and averaging within each decade.
Table 1. Decadal record, computed from the NOAA trend series.
DecadeMean ppmRate ppm/yrYears

Plate IV

A Note from the Desk

There is a discipline in this record that has nothing to do with what it measures. One instrument, one method, one number, reported the same way through eight hundred and twenty months. Directors changed. Funding wavered. A volcano closed the station for the better part of a year and the readings picked up on a neighbouring peak without a seam.

We spend our days on other people's series: revenues, flows, failures, the slow numbers that do not trend on anyone's screen. The Mauna Loa record is our reminder of what the good ones have in common. They are not the loudest. They are the ones kept honestly, in the same units, for longer than anyone thought to ask.

The measurement was always going to tell the truth. The only question was whether someone would keep it long enough to be believed.

Appendix

Plates, Method & Source

Index of plates

  • Fig. 1 Monthly mean CO2, 1958 to 2026. Raw monthly series.
  • Fig. 2 Mean yearly cycle with four sampled years. Trend removed.
  • Fig. 3 Decadal growth rate of the trend.
  • Tab. 1 Decadal record: mean, rate, years counted.

How the figures were made

  • Raw Figure 1 plots NOAA's reported monthly mean, unaltered.
  • Cycle The yearly cycle is each monthly reading minus NOAA's seasonally adjusted trend, averaged by calendar month across all years.
  • Rate Decadal rate is the year-over-year change in the annual mean of the trend, averaged within each decade. Whole years only.
  • Data Fetched once and baked to a local file. Nothing is modelled or forecast; every number on this page is measured or arithmetically derived from measurement.

Source & citation

Lan, X., J.W. Mund, A.M. Crotwell, et al. Atmospheric Carbon Dioxide Dry Air Mole Fractions from the NOAA GML Carbon Cycle Cooperative Global Air Sampling Network. Data 1958 to 1974 from C.D. Keeling, Scripps Institution of Oceanography.

Retrieved 2026-07-17 from gml.noaa.gov/ccgg/trends. Freely available public data, U.S. Government (NOAA GML).