Keeling Curve Explained: Reading the CO2 Record
- What does the Keeling Curve show?
- Who started the record, and which dataset are you reading?
- What does ppm mean?
- Why is Mauna Loa useful, and how is air measured?
- Why does the line rise and fall each year?
- How can you compare two published values correctly?
- Is Mauna Loa the global average?
- Does the upward curve alone establish the cause?
- How does an observed curve connect to future projections?
- Sources
What does the Keeling Curve show?
The Keeling Curve records atmospheric carbon dioxide measured at Mauna Loa, Hawaii, beginning in 1958. Its repeating seasonal rise and fall sits on a long-term increase. It measures carbon dioxide abundance, not temperature or annual emissions. Understanding the graph means separating those quantities, identifying the observing programme and reading the data notes. The evidence for human causation extends beyond the curve's shape to carbon accounting and chemical measurements.
A familiar graph can conceal several different questions. Why does the line dip? Why measure near a volcano? Does the number describe the whole planet? Answering each requires knowing what was measured before deciding what the result means.
Who started the record, and which dataset are you reading?
Charles David Keeling began the Mauna Loa measurements in March 1958. Scripps' history of the record describes the early discovery of both the seasonal cycle and an increase between years. Earlier measurements at other locations had helped him distinguish broadly representative air from local influences.
Scripps and the National Oceanic and Atmospheric Administration, or NOAA, operate separate measurement programmes. NOAA's Mauna Loa trends page says its measurements began in May 1974, running alongside Scripps' observations. Its full-record graph combines Scripps and NOAA data.
Therefore, record the provider as well as the place. “Mauna Loa” identifies a location; it does not, by itself, identify the data series, processing or release. Do not silently attach a Scripps caption to values taken from NOAA.
What does ppm mean?
Parts per million, abbreviated ppm, expresses a proportion. For atmospheric carbon dioxide reporting, dry-air mole fraction means the number of carbon dioxide molecules divided by the total number of air molecules after excluding water vapour.
An illustrative value of 400 ppm means 400 carbon dioxide molecules per million dry-air molecules, or 0.04%. It is an explanation of units, not a current reading. NOAA's global data page defines this convention.
A ppm value is not tonnes emitted that year. Emissions describe a flow into the atmosphere; atmospheric abundance reflects additions and removals over time. Nor is ppm a temperature unit. The greenhouse-effect explainer supplies the physical connection between added greenhouse gases and Earth's energy balance.
Why is Mauna Loa useful, and how is air measured?
NOAA's measurement-method description, updated in September 2020, explains that the high-elevation site samples air representative of large areas. Researchers seek background air, meaning air whose composition is not dominated by nearby additions or removals.
The account describes an analyser installed in April 2019 using cavity ring-down spectroscopy: carbon dioxide is quantified through its effect on how quickly light decays inside an optical cavity. Reference gases of known composition calibrate the measurements.
Local influences still matter. The documented selection procedure identifies periods affected by variability and upslope air influenced by vegetation. Selection flags distinguish background observations; the underlying hourly data are retained. This is a description of a professional monitoring system, not a home sampling procedure.
The useful question is consequently not whether the location has any local influences. It is how the programme identifies those influences, checks instruments and documents which observations enter its background record.
Why does the line rise and fall each year?
The seasonal pattern reflects changing exchange with land vegetation, especially in the Northern Hemisphere. Photosynthesis removes carbon dioxide from air during the growing season; respiration and decomposition return carbon. The larger extent of northern land and vegetation helps explain the strong northern seasonal signal. NOAA's seasonal explanation
A seasonal decrease is therefore not evidence that the long-term human contribution has disappeared. Likewise, a seasonal increase should not be described as a direct measurement of that month's industrial emissions.
The NOAA graph distinguishes monthly means from values adjusted for seasonality. Neither is a forecast; check the legend before comparing.
How can you compare two published values correctly?
On the NOAA Mauna Loa page checked on 8 September 2026, the displayed update date was 5 August 2026. It listed July 2026: 429.12 ppm and July 2025: 427.87 ppm.
The difference is:
429.12 − 427.87 = 1.25 ppm.
That compares two displayed monthly means for the same calendar month. It is not today's reading, the difference between annual means, or a conversion into annual emissions. NOAA notes that its latest year remains preliminary.
Use this original comparison record before doing arithmetic:
| Check | Record for this example | Why retain it? |
|---|---|---|
| Provider and series | NOAA Mauna Loa monthly means | Prevents switching programmes unnoticed |
| Periods | July 2026 and July 2025 | Keeps months distinct from annual averages |
| Unit | ppm | Prevents treating abundance as temperature |
| Release context | Page dated 5 August 2026; checked 8 September | Shows which displayed values were used |
| Calculation | Later value minus earlier value | Makes the 1.25 ppm difference reproducible |
This subtraction is not a named annual-growth-rate statistic.
Is Mauna Loa the global average?
No. NOAA's global monthly series averages observations from a network of marine surface sites. Its construction differs from the single high-elevation Mauna Loa record. Shared units do not make the geographic sampling identical.
Keep location and averaging method attached to a number. If one chart shows Mauna Loa and another a global marine-surface average, a difference between them does not automatically indicate an error.
The same principle applies through time. A chart extending before 1958 cannot consist entirely of direct Mauna Loa measurements. Identify the additional record and method in its documentation rather than assuming the instrument observed every year shown.
Does the upward curve alone establish the cause?
The curve establishes an increase in the measured abundance. Identifying its cause requires additional evidence.
Scripps' explanation of human causation connects fossil-fuel emissions with carbon accumulating in the atmosphere, oceans and land. The accounting includes carbon taken up outside the atmosphere; emitted carbon does not all remain airborne.
Chemical evidence adds another test. Isotopes are forms of an element with different numbers of neutrons. Scripps' flask-measurement explanation describes carbon-isotope analysis, including the absence of carbon-14 in fossil-fuel carbon after its radioactive decay over geological time. Such measurements provide source information that a concentration line alone cannot.
This strengthens, rather than weakens, the explanation: human causation is supported by converging evidence, not simply asserted from a rising line.
How does an observed curve connect to future projections?
Observations and projections answer different questions. The curve records measured atmospheric change; a projection examines outcomes under specified future conditions. Our climate-models guide explains that distinction, and the SSP scenario guide decodes the conditional labels.
Extending the curve with a ruler is not a climate assessment. The practical reading sequence is simpler: identify the series, read its units and processing, compare matching periods, and keep the causal evidence separate from the arithmetic.
Sources
Primary pages linked above were opened and read on 8 September 2026: NOAA Global Monitoring Laboratory's Mauna Loa trends, global trends, measurement-method and seasonal explanations; Scripps Institution of Oceanography's history, human-causation and flask-analysis articles. Historical methods and dated observations are identified as such. No monitoring instrument or climate model was operated for this guide.