The Mauna Loa CO₂ record as a reference dataset

Summary

Background

The measurement programme

The Mauna Loa series is the longest continuous direct record of atmospheric CO₂, begun by Charles David Keeling in March 1958 (Keeling et al., 1976) and continued by NOAA since 1974 (Thoning et al., 1989). The current files are published by NOAA’s Global Monitoring Laboratory (Lan et al., 2026). Its longevity is the point: a measurement that has been made the same way, at the same place, for more than sixty years, with every change of instrument or procedure documented.

Why an accounting scaffold cares

For SEEA (System of Environmental-Economic Accounting)-style work the value of a series like this is not any single number but its lineage. An account that cites an atmospheric baseline needs to know which file, which version and which publisher’s citation applies (United Nations et al., 2021). This report follows the data from NOAA’s files, through a versioned bundle, to the figures below.

Data and methods

The bundle

The values on this page are read at build time from the noaa-mlo-co2 bundle described under Data: monthly means, annual means and annual growth rates, with NOAA’s flags for interpolated months carried through. Nothing in the text is typed by hand — change the bundle, rebuild, and the numbers update.

Treatment of gaps

The rules below say what each treatment does to the series a reader sees (Table 1). They are prose, not numbers, so they are written here rather than computed; the table is kept out of the presentation deck and printed as records rather than as a grid, because each row is a paragraph.

Results

The long-term trend

Figure 1 shows the annual mean since 1959. The curve is not a straight line: the increment added each year has grown from under one part per million in the early 1960s to two-to-three parts per million in most recent years (Figure 2).

The rate of increase

Decadal averages make the acceleration plain (Figure 3, Table 2): about 0.86 ppm per year in the 1960s against about 2.42 ppm per year in the 2010s.¹

The seasonal cycle

Within each year, CO₂ at Mauna Loa peaks in May and reaches its minimum in September or October (Figure 4), as Northern-Hemisphere plants draw down carbon through the growing season and release it in autumn and winter. The amplitude is a few parts per million — small against the trend, but a useful reminder that a single month’s reading is not a year’s mean.

The same five cycles as a table (Table 3): the May peak and the autumn trough are visible in every column, and reading down any month shows the trend running through the seasons. In print the table takes a landscape page of its own — six columns of tabular numbers are a comparison, not a list — with the report’s running furniture and the table’s title repeated on every page it occupies.

Implications for accounts

Three practices transfer directly from the NOAA record to accounting work. First, publish the file, not just the figure: a reader must be able to reach the rows a number came from. Second, version the release and say what changed — NOAA’s files carry creation dates and the bundle carries a changelog. Third, ask to be cited in a specific form, and make that form easy to copy; the citation block at the end of this report does the same for the report itself. Physical-science context for the trend is given by the IPCC’s Sixth Assessment Report (IPCC, 2021).

Data

References