
build a quantitative model of carbon moving among the atmosphere, hydrosphere, geosphere and biosphere, use NOAA CO2 data to show that carbon is conserved, and name the limits of my model.
Phase 3: Make Sense
Do Now. Reweigh your dried shell chips from 3.04 and compute each percent mass loss. Compare with your predictions: which was furthest off, and why?
The atmosphere's CO2 rises and falls by a few ppm every year, lowest in early autumn and highest in late spring in the Northern Hemisphere. What causes this yearly swing?
Reservoirs and fluxes. A carbon-cycle model has two kinds of numbers. Reservoirs (stocks) are amounts stored, in GtC. Fluxes are rates of flow between reservoirs, in GtC per year. Our reservoir table from 3.02 tells us where carbon is; today we add how fast it moves. Rounded values for the 2010s (IPCC Sixth Assessment Report, 2021; Global Carbon Project, Global Carbon Budget):
| Flux | From → to | About (GtC/yr) |
|---|---|---|
| Photosynthesis on land | atmosphere → biosphere | 120 |
| Respiration, decay and fire on land | biosphere → atmosphere | slightly less than 120 |
| Gas exchange at the sea surface | both ways | about 80 each way |
| Fossil fuels and cement | geosphere → atmosphere | about 10 |
| Land-use change (clearing forests) | biosphere → atmosphere | about 1 |
| Volcanoes | geosphere → atmosphere | about 0.1 |
| Rock weathering, burial of shells and sediment | atmosphere and ocean → geosphere | a few tenths |
The Keeling curve. Since 1958, NOAA and partner scientists have measured CO2 in clean air at a high mountain observatory on a Pacific island and, later, at stations around the world. Annual means (NOAA Global Monitoring Laboratory, rounded): 1960, 317 ppm; 1970, 326; 1980, 339; 1990, 354; 2000, 370; 2010, 390; 2020, 414; 2023, 419. Air trapped in ice cores shows about 280 ppm before industrialization. The saw-tooth inside each year is the seasonal swing of the biosphere; the steady climb is the human flux.
Worked example 1: from ppm to GtC. One ppm of CO2 in the whole atmosphere holds about 2.12 GtC. Tomas computed the rise since pre-industrial times: (419 − 280) ppm × 2.12 GtC/ppm = 295 GtC, consistent with the reservoir growing from about 590 to about 870 GtC. For 2010 to 2020: (414 − 390) ÷ 10 = 2.4 ppm per year, or 2.4 × 2.12 = 5.1 GtC added to the air each year.
Worked example 2: does the carbon add up? People release about 11 GtC per year (10 from fossil fuels and cement, 1 from land-use change). Hana checked conservation with a rounded budget: the atmosphere gains about 5, the ocean takes up a net of about 3, and land plants and soils take up a net of about 3. 5 + 3 + 3 = 11. Every tonne released ends up in a reservoir; none disappears. The fraction left in the air, 5 ÷ 11 ≈ 0.45, is the airborne fraction. The ocean’s share, 3 ÷ 11 ≈ 27%, is the carbon that is lowering its pH and carbonate.
Big stocks, small imbalances. The natural fluxes, 120 and 80 GtC per year, dwarf the human 11. Before industrialization they were almost exactly balanced, so the atmosphere held steady near 280 ppm for thousands of years. The human flux is small compared with the natural ones but it is one-way, so it accumulates. The fast reservoirs (air, surface ocean, plants) share the extra carbon within years to decades; moving it into the deep ocean takes centuries, and returning it to rock by weathering and burial takes tens of thousands of years.
What every model leaves out. Our box model omits methane, treats the whole ocean as one box when the surface and deep ocean behave very differently, hides the uncertainty ranges behind rounded numbers (the IPCC gives soils as 1,500–2,400 GtC), and ignores that fluxes change as conditions change: warmer water holds less CO2, and the land sink depends on rainfall, fire and CO2 itself. Naming these limits is part of using a model honestly.
Reading. NOAA Global Monitoring Laboratory, “Trends in Atmospheric Carbon Dioxide”; IPCC AR6 Working Group I, chapter 5 (global carbon cycle figure); OpenStax Biology 2e, chapter 46.3 (biogeochemical cycles), CC BY 4.0.
Use the reservoir and flux tables and the NOAA annual means.
Model check, in your notebook.
Use your model to answer: if human emissions stopped tomorrow, which flux in your diagram would start lowering atmospheric CO2, and would it be fast or slow? Explain with the numbers.
Name two limitations of your model and, for each, a way the real Earth could behave differently from what the model predicts.
The carbon adds up, and some of it is ending up in the hatchery’s bay. Next phase, we turn this into a computational model. Revisit the Wonder Board: which questions can we now answer?
Exit ticket: connect your shell-chip data to the carbon budget in two sentences. Where did the chips' carbon go, and where is the ocean's extra carbon coming from?
Materials: dried shell chips and balances, the reservoir and flux tables, the NOAA annual-mean table or the NOAA GML Keeling-curve graph projected, colored pencils, large paper for the quantitative model.
Data notes: values are rounded from IPCC AR6 and the Global Carbon Budget; land and ocean sinks vary from year to year, and the published budget carries a small imbalance. Students should state values as “about” and cite the source.
Listen for: “the ocean is a carbon source” (today it is a net sink); “volcanoes cause the rise” (about 0.1 GtC/yr, roughly one percent of the human flux); confusing reservoirs with fluxes.
Support / stretch: provide a partly labeled four-box diagram. Stretch: estimate how long an average carbon atom stays in the atmosphere before it is exchanged (about 870 ÷ (120 + 80) ≈ 4 years) and explain why that is very different from how long an extra ppm of CO2 lasts.