eduKate Learning Manual · Climate Science × Atmospheric Chemistry × Measurement Science · Secondary → JC · Calibrate → Filter Background Air → Measure → Separate Season → Track Trend
Wait, What? The Jagged Up-and-Down Pattern Is Part of the Evidence That the Long-Term Rise Is Real
Atmospheric carbon dioxide does not rise in a perfectly smooth line. At Mauna Loa, the record climbs and falls every year in a seasonal sawtooth — while the baseline underneath keeps rising over decades.
That seasonal variation is not noise to be erased. It is a repeatable biological signal, driven largely by the seasonal cycle of photosynthesis and respiration across the Northern Hemisphere.
Charles David Keeling’s achievement was to build a measurement system precise and stable enough to see both patterns at once: the annual breathing of the biosphere and the persistent accumulation of CO₂ in the atmosphere.
calibrated reference gases → remote background-air sampling → reject local contamination → continuous CO₂ measurement → recurring seasonal cycle → long-term baseline rises → independent stations and carbon-budget evidence confirm a global atmospheric increase.
The Big Question
How can scientists distinguish a genuine global increase in atmospheric CO₂ from instrument drift, a nearby volcano, local pollution or an ordinary seasonal cycle?
Quick Answer
Keeling’s programme used precise nondispersive infrared measurements, repeatedly tied to reference gas standards, at remote sites chosen to sample well-mixed background air.
At Mauna Loa, individual intervals affected by local volcanic or upslope air could be recognised and excluded from background records. The remaining measurements showed a highly repeatable seasonal cycle superimposed on a persistent year-to-year increase.
The long-term curve itself establishes that atmospheric CO₂ concentration is rising. Attribution of most of the modern rise to fossil-fuel combustion and land-use change uses a broader evidence stack: emissions inventories, carbon isotopes, declining atmospheric O₂ relative to combustion expectations, ocean uptake and global carbon-budget closure.
What You Will Learn
- what atmospheric CO₂ concentration means
- why parts per million is used
- how nondispersive infrared analysers measure CO₂
- why reference gases and calibration matter
- why Mauna Loa is useful but not magically free of local contamination
- how contaminated intervals can be identified
- why the Northern Hemisphere seasonal cycle creates a sawtooth
- how a trend can coexist with periodic variation
- why multiple remote stations matter
- what the Keeling Curve proves directly and what fossil-fuel attribution requires additionally
Part 1 — What Does 315 ppm Mean?
Atmospheric CO₂ is commonly reported as a mole fraction in parts per million.
A value of 315 ppm means approximately 315 CO₂ molecules per million molecules of dry air, expressed as a molar ratio.
The early Mauna Loa record began near this scale in 1958–1959.
Because the signal changes by only a few ppm over a season and historically by fractions to a few ppm per year, the measurement system must distinguish very small relative changes reliably over decades.
Part 2 — Infrared Absorption Turns Gas Concentration Into a Signal
CO₂ absorbs infrared radiation at characteristic molecular vibration frequencies.
A nondispersive infrared analyser sends infrared radiation through sampled air and compares absorption in CO₂-sensitive spectral regions.
More CO₂ generally produces a larger absorption response, but the raw instrument output is not yet a trustworthy atmospheric concentration.
It must be calibrated against gases whose CO₂ mole fractions are known.
Part 3 — Calibration Is the Spine of a Decades-Long Record
An instrument can drift while appearing internally consistent.
If the analyser slowly changes sensitivity, a false long-term trend could appear even if the atmosphere did not change.
Keeling’s programme therefore relied on reference gases and calibration scales so measurements made at different times could be connected to one stable concentration framework.
raw detector response → reference standard → calibrated mole fraction → long-term comparability.
Scripps notes that historical data can be revised when reference-gas calibrations or scales are improved. That is not evidence that the trend is arbitrary; it is evidence that metrology remains traceable and correctable.
Part 4 — Why Mauna Loa?
Mauna Loa Observatory sits high above much of the local vegetation and boundary-layer pollution, in the central Pacific far from major continental industrial sources.
At suitable times and wind conditions it samples air that has travelled long distances and is comparatively well mixed.
This makes it valuable for measuring background atmospheric composition.
But the location is on an active volcano. Local volcanic CO₂ can contaminate some observations.
The correct claim is not “Mauna Loa has no local CO₂.” It is:
background-air conditions can be identified, while locally contaminated intervals are flagged or excluded from the background record.
Part 5 — Local Contamination Has a Pattern
Volcanic or local upslope air does not usually behave like stable remote background air.
Contaminated intervals can show rapid variability, unusual concentration excursions and meteorological associations.
Historical Mauna Loa processing used the stability of the analyser trace and atmospheric context to distinguish baseline conditions from locally disturbed periods.
A measurement is therefore not made “global” just because it came from a mountaintop. It becomes useful as background evidence through selection criteria, calibration and comparison with other sites.
Part 6 — The Seasonal Sawtooth
Most of Earth’s land vegetation lies in the Northern Hemisphere.
During Northern Hemisphere spring and summer, growing plants remove large amounts of CO₂ through photosynthesis.
During autumn and winter, photosynthetic uptake weakens while respiration and decomposition continue returning CO₂.
The atmospheric record therefore tends to:
- rise through much of autumn and winter;
- reach a seasonal maximum around Northern Hemisphere spring;
- fall during strong growing-season uptake;
- reach a minimum around late summer or early autumn.
The exact phase and amplitude vary with latitude and atmospheric transport.
A Signal-Decomposition Window
A useful conceptual model is:
C(t) = T(t) + S(t) + ε(t)
where:
- T(t) = long-term trend;
- S(t) = recurring seasonal component;
- ε(t) = short-term weather, transport and measurement variation.
A seasonal oscillation can be large enough to make CO₂ fall for several months even while T(t) continues rising over years.
Part 7 — Why the Seasonal Cycle Does Not Explain Away the Trend
If CO₂ simply followed a closed annual cycle, each year’s peaks and troughs would return to approximately the same baseline.
They do not.
Successive seasonal minima rise. Successive seasonal maxima rise. Smooth the repeating cycle and the long-term baseline rises.
This is the key discriminator:
seasonal cycle repeats around a moving baseline → periodic biosphere exchange and long-term atmospheric accumulation are separate signals.
Part 8 — Why One Mountain Is Not Enough
A global atmospheric conclusion should not depend on one instrument at one volcano.
Scripps measurements began at additional locations including the South Pole and later expanded across a network from Arctic to Antarctic regions. NOAA and other programmes built further independent monitoring networks.
Different stations show different seasonal amplitudes and local details, but the long-term global rise appears across the network.
That spatial replication strongly weakens explanations based on one mountain’s local source or one analyser’s drift.
The Historical Carrier — Charles David Keeling
Keeling developed highly precise CO₂ measurement methods in the 1950s and began atmospheric measurements at La Jolla and the South Pole in 1957. Continuous measurements at Mauna Loa began in 1958.
Within the early years, the record clearly displayed both a regular seasonal oscillation and a year-to-year rise.
The power of the Keeling Curve came from continuity. A programme designed to maintain calibration and measurement discipline over decades turned individual concentrations into a planetary time series.
Part 9 — What the Curve Directly Shows
The direct measured claim is:
background atmospheric CO₂ concentration has increased persistently over the instrumental record while undergoing a repeatable seasonal cycle.
The curve does not, by itself, label each added CO₂ molecule “fossil fuel.” Source attribution requires other evidence.
Part 10 — How Fossil-Fuel Attribution Is Tested
The modern attribution rests on convergent observations.
- emissions inventories: humans emit large, quantified amounts of carbon from fossil-fuel combustion and cement production;
- carbon budget: atmospheric increase is smaller than total human emissions because land and ocean absorb part of the emitted carbon;
- carbon isotopes: the isotopic composition of atmospheric carbon shifts in the direction expected from addition of fossil carbon depleted in ¹³C and lacking modern ¹⁴C;
- oxygen: atmospheric O₂ declines consistently with combustion consuming oxygen;
- ocean chemistry: oceans absorb anthropogenic CO₂, changing dissolved inorganic carbon and pH.
These independent ledgers make the attribution much stronger than simple temporal correlation.
Part 11 — Why Volcanoes Cannot Explain the Global Trend
Volcanoes emit CO₂, including Mauna Loa itself.
But local volcanic contamination can be identified in high-frequency data, and the long-term background increase appears at widely separated stations.
At the global budget scale, annual human fossil-fuel and industrial CO₂ emissions greatly exceed volcanic CO₂ emissions.
The volcanic alternative therefore fails both the local-data structure and the global source ledger.
Part 12 — A Pause in Measurement Is Not a Pause in the Atmosphere
Long scientific records can face outages, funding interruptions, instrument changes or site disruptions.
A robust monitoring system therefore needs overlapping instruments, flask samples, reference standards and multiple stations.
The science does not depend on pretending one physical analyser has run untouched forever. It depends on traceable continuity across measurement systems.
RFE Stress Test — Global CO₂ Rise or Instrument/Local Artefact?
- reference-gas test: do repeated standards reveal analyser drift?
- local-variability test: are volcanic or upslope episodes flagged by unstable concentrations and meteorology?
- seasonal decomposition: do successive annual peaks and troughs rise despite the repeating cycle?
- multi-station test: do remote stations in different hemispheres show the long-term increase?
- independent sampling: do flask and in-situ measurements agree within calibration limits?
- source-budget test: can human emissions, sinks, isotopes and O₂ changes account for the atmospheric increase?
- volcanic alternative: can local or global volcanism reproduce the spatial, temporal and isotopic evidence?
The modern atmospheric-CO₂ conclusion survives because calibration, seasonal structure, global replication and carbon-source accounting all tell compatible stories.
Observation vs Inference
Observation: calibrated background-air measurements show a repeating seasonal CO₂ cycle superimposed on a persistent long-term rise.
Atmospheric inference: the global atmosphere is accumulating CO₂ over time.
Source inference: convergent emissions, isotope, oxygen and carbon-budget evidence identifies human fossil-fuel use and land-use change as the dominant cause of the modern increase.
Common Misconceptions and How to Repair Them
- “CO₂ falls every summer, so it is not rising.” Repair: seasonal decline occurs around a baseline that rises across years.
- “Mauna Loa measures volcano gas.” Repair: local volcanic intervals are identified; background conditions and other stations reproduce the trend.
- “One instrument running for decades proves the curve.” Repair: reference standards, overlapping methods and network measurements preserve traceability.
- “The Keeling Curve alone proves fossil fuels caused every extra molecule.” Repair: the curve measures concentration; source attribution comes from a convergent evidence stack.
- “Photosynthesis should remove all emitted CO₂.” Repair: land and ocean absorb only part of human emissions; the remainder accumulates in the atmosphere.
- “ppm is a mass percentage.” Repair: atmospheric CO₂ ppm is commonly reported as dry-air mole fraction.
Checkpoint Questions
- What does atmospheric CO₂ ppm represent?
- Why are reference gases essential?
- Why is Mauna Loa useful for background air?
- Why must local volcanic influence still be checked?
- What causes much of the Northern Hemisphere seasonal sawtooth?
- How can a seasonal cycle coexist with a long-term trend?
- What additional evidence attributes the rise to fossil carbon?
Apply It — A Year With a Lower June Value
Suppose June CO₂ is lower than May because Northern Hemisphere photosynthesis has intensified. That does not imply the long-term rise stopped. Compare June with June in earlier years, or remove the seasonal component before evaluating the long-term trend.
Unfamiliar Transfer — How to Read Any Long Environmental Time Series
The transferable method is:
calibrate instrument → identify local contamination → preserve raw variation → separate recurring cycles from trend → replicate spatially → connect trend to independent causal evidence.
The same architecture applies to temperature records, sea level, ocean chemistry and atmospheric pollutants.
Answer Key
1. CO₂ molecules per million molecules of dry air on a molar basis. 2. They convert instrument response into a stable traceable concentration scale. 3. High elevation and remote Pacific location often provide well-mixed background air. 4. It is an active volcano and local air can contain extra CO₂. 5. Seasonal balance of terrestrial photosynthesis, respiration and decomposition, dominated globally by Northern Hemisphere land. 6. The annual oscillation repeats around a rising baseline. 7. Emissions inventories, isotope changes, O₂ decline, ocean uptake and carbon-budget closure.
Can You Explain WHY?
Explain why the seasonal sawtooth strengthens rather than weakens the measurement story. A strong answer should connect repeatable annual biology → expected up/down cycle → calibrated recovery each year → successive peaks and troughs shift upward → separate seasonal signal from persistent accumulation.
Singapore Secondary and JC Science Bridge
Secondary Science introduces photosynthesis, respiration and the carbon cycle. JC Chemistry and Biology add spectroscopy, equilibria and ecosystem fluxes. The Keeling Curve turns those ideas into one of science’s great monitoring lessons: reliable planetary knowledge comes from calibration, continuity, disciplined exclusion and independent causal ledgers.
Deep Science Windows
- δ¹³C: atmospheric carbon-isotope trends help identify changing carbon sources.
- O₂/N₂ measurements: combustion and ocean/land uptake alter atmospheric oxygen in diagnostic ways.
- Airborne fraction: only part of annual anthropogenic CO₂ emissions remains in the atmosphere because land and oceans take up the rest.
- Carbon-cycle feedbacks: climate affects land and ocean sink strength.
- Global observing networks: independent laboratories maintain calibrated greenhouse-gas records across many stations.
Evidence Boundaries
The Mauna Loa record is a high-quality background record, not a claim that every air parcel at the observatory is pristine. Historical values can be recalibrated as standards improve. The curve directly measures atmospheric CO₂ concentration and its seasonal/trend structure; attribution of the modern rise to human emissions is a broader convergent-evidence conclusion. This article complements rather than replaces the existing system-level How Climate Works owner.
Manual Summary — KNOW → CONNECT → EXPLAIN → APPLY → CHECK
- KNOW: calibrated CO₂ is reported as dry-air mole fraction in ppm.
- CONNECT: remote background-air selection separates global signal from local contamination.
- EXPLAIN: seasonal biosphere exchange creates a sawtooth around a rising baseline.
- APPLY: compare like seasons or remove seasonal structure before judging trend.
- CHECK: calibration, local source flags, multiple stations, independent sampling and source attribution evidence.
Teaching Guide for Parents, Tutors and Teachers
Why this opening works: students often mistake any downward segment for reversal of the trend. The sawtooth forces them to separate timescales before interpreting the graph.
- Central reasoning model: calibrated measurement → background filter → seasonal component + trend → independent causal evidence.
- Teaching sequence: ppm → infrared analyser → reference gas → Mauna Loa site → contamination rejection → seasonal cycle → rising baseline → global replication → fossil-fuel attribution.
- Diagnostic question: “If CO₂ falls every northern summer, what comparison would tell you whether the long-term trend is still rising?”
- If stuck: compare the same month across successive years instead of adjacent months.
- Ready for more: introduce carbon isotopes, O₂/N₂, airborne fraction and carbon-cycle feedbacks.
Quiet Teaching Standard: do not teach the Keeling Curve as “a graph showing climate change.” Require the learner to distinguish the measured quantity, the seasonal mechanism, the long-term atmospheric trend and the separate evidence used for source attribution and climate consequences.