Analyze climate data across the geologic record, historical data and present-day measurements, and explain the mechanisms and feedbacks of climate change

Learning goal: Distinguish weather from climate; explain the greenhouse mechanism and the main feedbacks; interpret carbon dioxide and temperature records from the geologic record, human historical data and present-day measurements and explain their correlation; compare El Niño and La Niña effects at the same latitude; describe how climate models produce evidence-based forecasts; investigate local habitat change; and separate scientific findings from policy positions.

Weather, climate and what "change" means. Weather is the state of the atmosphere over hours or days; climate is the statistical pattern of weather over decades, conventionally 30 years. Climate change means a change in those statistics: the average, the extremes, the timing of seasons. One hot summer is weather; a run of decades in which each is warmer on average than the last is climate. This distinction is the first thing to apply to any data set: ask what period it covers and whether the period is long enough to show a trend rather than year-to-year noise.

The mechanism. Sunlight reaches Earth mostly as visible light. The surface absorbs it and re-emits energy as infrared radiation. Certain gases in the atmosphere, chiefly water vapor, carbon dioxide (CO₂), methane and nitrous oxide, absorb infrared radiation and re-emit it in all directions, including back toward the surface. This greenhouse effect is natural and keeps Earth's average surface temperature roughly 33 °C warmer than it would be with no greenhouse gases. Adding more of these gases increases the amount of outgoing infrared that is intercepted, so the surface and lower atmosphere warm until the energy leaving the planet again balances the energy arriving. Nitrogen and oxygen, which make up about 99% of dry air, do not absorb infrared significantly, which is why gases present in parts per million can matter so much.

The records. Three kinds of data cover three time scales. From the geologic record, air bubbles trapped in Antarctic ice cores give a direct record of atmospheric CO₂ over about the last 800,000 years, and the isotopes of the ice give temperature; in that record CO₂ moved between roughly 170 and 300 parts per million (ppm) as ice ages came and went, and the two curves rise and fall together. Older records from ocean sediments and rocks extend the picture further back. Human historical data include thermometer records since the mid-1800s, ships' logs, harvest dates, and the dates lakes froze and thawed. Present-day data include the continuous CO₂ measurements at Mauna Loa, Hawaii, begun by Charles Keeling in 1958, which started near 315 ppm and passed 420 ppm in the 2020s, together with satellite measurements of sea level, ice, and ocean temperature. Compared with a pre-industrial level of about 280 ppm, CO₂ has risen by about half; the extra carbon carries the isotopic signature of fossil fuels, which is how scientists know its source.

Correlation and cause. The correlation between historical CO₂ and temperature is strong, and a good explanation must handle a subtlety. In the natural ice-age cycles, small changes in Earth's orbit started the warming or cooling, and CO₂ rose or fell within centuries afterwards as the oceans and biosphere responded; CO₂ then amplified the change as a feedback. So CO₂ has been both a follower and a driver. Today the situation is different: CO₂ is rising first, because of emissions, and the physics that made it an amplifier in the past makes it a driver now. Correlation alone does not prove cause; the case rests on the measured absorption properties of the gases, the isotopic evidence of where the carbon came from, and the fact that models including human emissions reproduce the observed warming while models with natural factors alone do not.

Feedbacks. A feedback is a process by which the result of warming changes the amount of further warming. Positive (amplifying) feedbacks: warmer air holds more water vapor, itself a greenhouse gas; melting snow and sea ice expose darker ocean and land that absorb more sunlight (the ice-albedo feedback, shown in the diagram); thawing permafrost releases CO₂ and methane. Negative (damping) feedbacks: a warmer surface radiates more infrared to space, which is the main reason warming does not run away; more CO₂ can increase plant growth, which stores some carbon. Clouds can act either way and remain the largest source of uncertainty in how much warming a doubling of CO₂ will produce.

What the observations show. Multiple independent lines of evidence agree. The global average surface temperature over the decade 2011–2020 was about 1.1 °C warmer than the 1850–1900 average, according to the Intergovernmental Panel on Climate Change (IPCC), the body that periodically assesses the published research. Global mean sea level rose about 20 centimeters between 1901 and 2018, from warming (expanding) seawater and melting land ice. Most mountain glaciers have retreated; Arctic summer sea-ice extent has declined since satellite records began in 1979; the ocean has absorbed most of the added heat, which is why ocean heat content is one of the clearest indicators. The IPCC's 2021 assessment concluded that it is unequivocal that human influence has warmed the atmosphere, ocean and land. That is the scientific consensus, stated as such: a conclusion of the large majority of researchers based on the weight of evidence, open to revision by new evidence as any scientific conclusion is, but not seriously disputed in the peer-reviewed literature.

Models and forecasts. A global climate model is a computer program that divides the atmosphere, ocean and land into grid cells and solves the equations for energy, motion and moisture in each. Models are tested by running them over the past and comparing with observations. Forecasts are made by running them forward under different scenarios of future emissions; the result is a range, not a single number, because emissions depend on human choices and because feedbacks are uncertain. An evidence-based forecast therefore states the scenario, the central estimate and the range: for example, the IPCC's 2021 assessment gives, for 2081–2100 relative to 1850–1900, a best estimate of about 4.4 °C (very likely 3.3 to 5.7 °C) under a very high emissions scenario and about 1.4 °C (very likely 1.0 to 1.8 °C) under a very low emissions scenario. Regional forecasts for the Midwest generally include warmer winters, more intense heavy-rain events and more variable precipitation, which is why monthly precipitation records are worth analyzing: a change in how rain arrives (more in bursts, longer dry spells between) matters even when the annual total barely changes.

El Niño and La Niña. Every few years the tropical Pacific shifts between two states. In El Niño, sea surface temperatures in the central and eastern equatorial Pacific run warmer than normal and the trade winds weaken; the rainfall that normally sits over Indonesia moves east, bringing drought to Indonesia and northern Australia and heavy rain to coastal Peru and Ecuador. La Niña is the reverse: cooler eastern Pacific, stronger trade winds, wetter west and drier east. Comparing two places at the same latitude but different longitudes, say the coast of Peru and the islands of Indonesia near 5° south, shows opposite precipitation responses to the same event, which is why satellite sea-surface-temperature and rainfall maps are the right tools. These are natural oscillations superimposed on the long-term trend; El Niño years tend to be warmer globally and La Niña years cooler, and several record-warm years have coincided with or followed strong El Niño events, without changing the underlying trend.

Local habitats. The Great Lakes are a natural case study: long-term records show a general decline in average winter ice cover since the 1970s with large year-to-year swings, warmer summer surface water, and longer growing seasons for algae, which interacts with nutrient pollution. A local study asks what habitat is changing, what is measured, over what period, and what other causes (land use, invasive species) could explain the same observation.

Science and policy. Whether to increase federal spending on climate research is a policy question. Science can say what is happening and what is likely under each choice; it cannot say what a society should value. A fair position paper for more spending argues that better regional forecasts reduce the cost of adaptation; a fair paper against argues that money is better spent on adaptation or on other needs now. Both can accept the same scientific findings. Confusing "the science is uncertain" with "the science is unknown" is the most common error on both sides.

A brief timeline. Joseph Fourier proposed in the 1820s that the atmosphere warms the surface; John Tyndall measured in 1859 which gases absorb infrared; Svante Arrhenius estimated in 1896 how much doubling CO₂ would warm Earth; Keeling's measurements began in 1958; the IPCC was created in 1988; the United Nations Framework Convention on Climate Change was signed in 1992, the Kyoto Protocol adopted in 1997 and the Paris Agreement in 2015. Developed countries produced most of the historical emissions, while many developing countries now have faster-growing emissions and less capacity to adapt, which is the central tension in every international negotiation.

Worked ice-core reading (hypothetical values): Suppose a core gives, for four times in the past, CO₂ of 185, 230, 275 and 290 ppm and temperature differences from today of −8, −5, −2 and 0 °C. Plotting temperature against CO₂ gives points that rise together: every increase in CO₂ pairs with an increase in temperature. The correlation is positive and strong. What it does not by itself show is which changed first or whether a third factor (orbital forcing) drove both; that requires the timing evidence and the physics. The correct statement is: "CO₂ and temperature are strongly correlated in the record; the known infrared absorption of CO₂ explains how a rise in CO₂ amplifies warming, and warmer oceans release CO₂, which is how warming raises CO₂."

Worked present-day trend: Mauna Loa CO₂ was about 315 ppm in 1958 and about 420 ppm in 2023, a rise of 105 ppm in 65 years, an average of about 1.6 ppm per year. The rise has accelerated: it was under 1 ppm per year in the 1960s and over 2 ppm per year in the 2010s. Compared with the pre-industrial 280 ppm, 420 ppm is a 50% increase. The seasonal wiggle on the curve (higher each northern spring, lower each autumn) is the Northern Hemisphere's plants taking up CO₂ in summer and releasing it in winter, and it shows the measurements are sensitive enough to see the biosphere breathe.

Worked precipitation analysis (hypothetical): A station's monthly rainfall for June over two decades: 1990s mean 95 mm with a standard deviation of 20 mm; 2010s mean 98 mm with a standard deviation of 45 mm. (The standard deviation measures how far individual values typically lie from the mean; for roughly bell-shaped data about two thirds of values fall within one standard deviation of it, so most 1990s Junes lay between about 75 and 115 mm.) The mean rose by 3 mm, which is small beside that year-to-year spread, so these figures give no evidence that average June rainfall changed; a test on the individual yearly values would be needed to say more. The spread more than doubled: June totals in the 2010s were much more widely scattered around the mean. The correct conclusion is "a small observed rise in the mean that cannot be told apart from year-to-year variation; a much larger spread of June totals", not "it got wetter" or "nothing changed". Whether that means more very wet and more very dry Junes, or more intense storms within months, needs the individual monthly and daily values, not just the mean and standard deviation.

Worked El Niño comparison: Two coastal stations at about 5° S: one in Peru (about 80° W), one in Indonesia (about 105° E). In a strong El Niño year (illustrative values, not the record of one particular event) sea surface temperature off Peru can run several degrees, for example about 3 °C, above normal and coastal rainfall several times its usual amount, while the Indonesian station shows normal or slightly cooler seas and rainfall far below normal. Same latitude, same event, opposite sign, because the warm water and the rising air that produces rain shifted eastward along the equator.

Representation task: Redraw the ice-albedo feedback loop from the diagram and add a second loop for water vapor (warming → more evaporation → more water vapor → more infrared absorbed → warming). Then draw one damping loop (warmer surface → more infrared emitted to space → less warming) and label each loop "amplifying" or "damping". Below the loops, make a three-column table: geologic record, human historical data, present-day data, with one measurement and its approximate time span in each column.

Data task, choose a United States location: For a location your teacher approves, collect (a) a pre-industrial or geologic indicator (for example, the region's climate during the last ice age from a museum or geological survey source), (b) a historical thermometer or lake-ice record, and (c) a recent 30-year record. Plot what can be plotted, state the period each covers, and write one paragraph interpreting the pattern and one paragraph on its limits (station moves, urban heat, gaps). Do not fill gaps with invented values.

Independent performance: Write two short position papers (about 250 words each), one for and one against increasing federal spending on climate research. Success criteria: (1) both papers accept the same stated scientific findings (cite at least three from this lesson with figures); (2) each paper's argument rests on values or priorities, clearly identified as such; (3) each paper names one thing it would change its mind about if evidence changed; (4) you close with a paragraph on how you would distinguish a scientific claim from a policy claim in a news article.