Climate literacy lets you read Earth's climate the way the measurements do: as an energy budget that added greenhouse gases have pushed out of balance, with the ocean, ice, carbon and water cycles each answering on its own clock. Start with that budget and its carbon, then see how the change is measured and told apart from natural causes. Follow it into the ocean, ice, rain and extremes, and finish with models, impacts, cutting emissions, adapting, and checking a claim. Orientation for study, not policy advocacy or personal advice. Figures are dated, so check the current record before you rely on one.
Each chapter opens with the short version. Tap one to read the detail.
Weather, climate, and a system of linked parts
~2 min
Climate is what weather typically does over decades, so one day or one season is a single sample. Air, ocean, ice, land and living things trade heat, water and carbon, each on its own clock.
Weather is the state of the air at a given moment; climate is the typical weather over several decades. Its yardstick is the normal, the average of a weather value over three consecutive decades. The U.S. normals now in use cover 1991 to 2020, and most of the country ran warmer then than in 1981 to 2010, so when you hear "above normal" today, it means above a warmer bar than a decade ago.
The air is only one part. The ocean, ice, land surface and living things store and swap heat, water and carbon, so a change in one feeds the others. The tropics absorb more sunlight than the rest of the planet, and winds and ocean currents carry that surplus toward the poles.
Each part keeps its own time. Weather turns over in hours to weeks, pollution particles fall out within days to weeks, a big eruption's cooling fades in one to two years, and water in the deep overturning circulation takes about 1,000 years to finish its loop.
The trap is reading one season as a verdict on the climate. Say your town logs its coldest February in decades. That is one draw from a distribution; whether the distribution has moved is a question about decades of draws over a wide area.
Rule: before you call something climate change, ask what the long-run baseline is and how many years and places the comparison covers.
The energy budget, the greenhouse effect, and the gases that tip it
~2 min
Earth's temperature settles where the energy it sheds matches the sunlight it absorbs. Greenhouse gases slow the heat on its way out, and while added gases leave a surplus, about 90 percent of it piles up in the ocean.
In the National Oceanic and Atmospheric Administration (NOAA) energy budget, about 30 of every 100 units of incoming sunlight are reflected straight back, mostly by clouds. That reflected share is albedo, and bright snow and ice reflect far more than dark ground or open water. The rest warms the surface, which sheds energy as infrared. Greenhouse gases absorb most of that infrared and re-emit it in all directions, partly back down, so heat leaks to space more slowly. Without this natural effect Earth would average about -2 F instead of roughly 57 F.
Nitrogen and oxygen make up most of the air, yet the trapping is done by small amounts of carbon dioxide (CO2), methane, nitrous oxide, ozone and fluorinated gases, along with water vapor, which follows temperature rather than setting it.
Radiative forcing is the heat a driver adds to the budget, in watts per square meter. NOAA put the extra forcing from long-lived greenhouse gases in 2025 at about 3.6 W/m2, with CO2 supplying about 81 percent of the increase since 1990. Global warming potential (GWP) compares the energy a ton of each gas absorbs over a chosen period, usually 100 years. Methane traps far more energy per ton than CO2 but lasts only about a decade, so a 20-year horizon weights methane more heavily.
The trap is the saturation argument. Say you hear that CO2 already traps all it can, so adding more changes nothing. If that were true, forcing calculated from measured gas amounts would have flattened. NOAA's yearly index shows it still climbing as CO2 rises.
Rule: judge any driver by what it does to the energy budget, and check the time horizon before you compare two gases.
Where the carbon goes, and how emissions are counted
~2 min
Burning fossil fuel moves carbon from the slow rock cycle into the air far faster than nature returns it, and ocean and land absorb only about half. The accounting boundary decides what an emissions ranking says.
Carbon cycles quickly through living things: plants draw CO2 down by photosynthesis, and respiration and decay give it back. That is the yearly zigzag in the CO2 record, strongest in the Northern Hemisphere, which has more land. Fossil fuels hold carbon from organisms that lived millions of years ago, which normally returns very slowly. Burning them moves it into the atmosphere, a small reservoir whose level swings easily.
Natural sinks take part of each year's emissions. NOAA's estimates of the ocean's share run from roughly a quarter to 30 percent, and land and ocean together took about 60 percent in 1959 against about 55 percent now. Only about half of fossil CO2 has stayed airborne, which is why concentration rises more slowly than total emissions. Isotopes name the source: fossil carbon is low in carbon-13 and holds no carbon-14, and the carbon-14 decline is what rules out plants and soils.
Before you trust an emissions figure, know how it was counted. The Environmental Protection Agency (EPA) national inventory is territorial: it counts emissions where they physically occur inside the country and subtracts carbon taken up by U.S. forests and soils. Per million British thermal units (Btu), bituminous coal releases about 206 pounds of CO2 and natural gas about 117, so switching fuels cuts emissions without ending them.
The trap is treating one ranking as the ranking. A total shows where the most emissions happen but ignores population. Two countries with equal totals, one with ten times the people, differ tenfold per person, so the two rankings answer different questions.
Rule: before you compare two emissions figures, check the boundary, the gases included, and whether each is a total or per person.
Measuring the change, from thermometers back to ice cores
~2 min
The global temperature record starts around 1880, and independent teams agree on its trend. Proxies such as ice cores and tree rings reach back across the ice ages and show how unusual today's pace is.
Stations, ships, buoys, balloons and satellites feed the record, but before about 1880 there are too few readings for a global average. Analysts work in anomalies, each reading minus its own site's baseline: a mountaintop and a nearby valley differ in temperature yet share a similar anomaly. Station moves and new instruments are corrected by comparing each station with several neighbors. The National Aeronautics and Space Administration (NASA) checks city records, which really do run warmer, against nearby rural ones; once that is done, the urban effect plays no substantial role in global warming. NASA, NOAA and the UK Meteorological Office build separate records with different methods, and they agree closely on the trend. Measurements that need no thermometer agree as well: sea level rising, mountain glaciers losing mass, Arctic sea ice shrinking in extent and thickness, spring snow melting earlier.
Before instruments, proxies stand in. Bubbles in ice cores hold samples of the old atmosphere, and the ice's oxygen isotopes track temperature. Together they show ice ages and warm interglacials alternating over periods of about 100,000 years, with CO2 and temperature rising and falling together.
The trap is reading past warm climates as reassurance. You may hear that Earth has been warmer before, and it has: the early Eocene, about 54 to 48 million years ago, ran 9 to 14 C warmer than today, with CO2 between 1,000 and 2,000 ppm. The difference is speed: leaving the last ice age, CO2 rose about 1 ppm every 125 years, and since 1850 it has risen about 1 ppm every 23 weeks.
Rule: trust a trend that independent records and different kinds of measurement agree on, and ask how fast as well as how much.
The Sun, orbits, volcanoes and particles
~2 min
Orbital cycles paced the ice ages, the Sun varies slightly on an 11-year cycle, and volcanic and pollution particles cool for days to years. Each cause leaves a different pattern, and the measured pattern fits added greenhouse gases.
Milankovitch cycles are slow changes in the shape of Earth's orbit, the tilt of its axis and the axis's wobble. They shift sunlight between seasons and latitudes far more than they change the yearly total. They set off each exit from an ice age; warmer oceans then absorbed less carbon and warmer soils released more, so CO2 rose and amplified the warming. The same gas can respond to warming and also cause it.
In more than 40 years of satellite records, the Sun's output has varied by less than 0.1 percent, with no net increase in sunlight reaching Earth since the 1950s. NASA judges it extremely unlikely the Sun caused the recent warming. Volcanoes emit CO2, but human activity emits more than 100 times as much, and an eruption's main climate effect is cooling: sulfate particles in the stratosphere block sunlight for one to two years.
Pollution particles act near the ground. Light-colored sulfate reflects sunlight and has offset part of greenhouse warming, though it falls out within days. Soot does the opposite, absorbing sunlight and darkening snow. As energy use shifts away from the sources of reflective particles, losing their cooling could bring some temporary warming, which NASA judges unlikely to be large because the change is gradual.
The trap is ranking causes only by whether each could warm the planet. A brighter Sun would warm every layer up through the stratosphere, but the surface has warmed while the stratosphere cooled. NASA reads that pattern as consistent with heat-trapping gases building up near the surface, not a hotter Sun.
Rule: when two causes could explain a trend, work out the pattern each would leave, then check which one you actually see in the data.
The ocean: heat store, slow conveyor, and changing chemistry
~2 min
The ocean has taken up about 90 percent of the excess heat and a large share of the CO2, and each changes the water in its own way. Acidification comes from dissolved CO2 itself, so it does not need the water to warm.
Water stores far more heat than air, so the ocean holds about 90 percent of the energy surplus, most of it in the top 700 meters. Warming water expands, which caused about a third of global sea-level rise over the past century. The ocean is layered: a wind-mixed surface layer sits on top of colder water, and surface water reaches the depths only slowly and in a few places. Near the poles, forming sea ice leaves its salt behind, so cold, salty water sinks and drives a deep conveyor moving a few centimeters per second; one loop takes about 1,000 years.
Absorbed CO2 forms carbonic acid, which releases hydrogen ions. Those ions bond with carbonate, the building block oysters, corals and other shell builders need. Seawater is still basic, around pH 8.1, but pH is logarithmic, so the 0.1-unit drop in surface pH since the industrial revolution is sizable: NOAA's education pages call it about a 30 percent rise in acidity, and NOAA's Ocean Acidification Program puts the rise over 250 years at about 26 percent.
The trap is treating acidification as a side effect of warming. It is chemistry: ocean CO2 rises with atmospheric CO2, so pH keeps falling as more dissolves, a change that needs no warming. Off the U.S. West Coast, upwelling lifts deep water that last touched the air about 1,000 years ago, and absorbed human CO2 has widened the area this acidic water reaches.
Rule: track the ocean's heat and its carbon as two separate changes, each with its own consequences.
Ice, and why sea level rises unevenly
~2 min
Only ice resting on land raises sea level when it melts, and land ice now supplies about two-thirds of the rise. Local sea level also depends on whether the land is rising or sinking.
Floating sea ice already displaces its own volume, so melting it does not change sea level. Glaciers and ice sheets rest on land, and their meltwater now supplies about two-thirds of global sea-level rise; warming seawater's expansion is the other main cause. Bright sea ice reflects sunlight, so as it melts the darker water absorbs more energy, delaying the fall freeze and speeding spring melt, one driver of an Arctic warming far faster than the rest of the planet. Satellites count sea-ice extent as any area at least 15 percent ice-covered, which says nothing about thickness, and Arctic ice has lost both; its summer minimum extent is shrinking by about 12 percent per decade.
A glacier's mass balance is the snow it gains minus what it loses to melt, runoff, evaporation and calving. Most mountain glaciers have run decades of net loss, so a retreat is a lasting imbalance, not one warm summer.
A tide gauge measures water against nearby land, so it records relative sea level. Satellites, since the early 1990s, measure the ocean surface's height above Earth's center.
The trap is assuming your coast follows the global average. In EPA's record from 1960 to 2023, relative sea level fell at some Alaska and Pacific Northwest gauges because the land there rose faster than the ocean. Where land sinks, the opposite happens, and high tide flooding is now twice as frequent in U.S. coastal communities as it was 20 years ago.
Rule: ask whether melting ice is floating or on land, and whether a sea-level figure is relative to local land or absolute.
A wetter atmosphere, and how extremes shift
~2 min
Warmer air holds about 7 percent more water vapor per degree Celsius, loading heavier downpours, while faster evaporation dries soils. A warmer average makes old hot extremes more common.
Atmospheric water vapor rises about 7 percent per degree Celsius of warming, so more water is available when air rises. In EPA's U.S. record through 2023, a growing share of precipitation fell in intense single-day events, and nine of the top ten years for them came after 1995. Warmth also speeds evaporation from soil and plants, so drought can deepen without less rain; NASA notes droughts are intensifying mainly through drier soils. From 1949 to 2024, more than 80 percent of stations in the lower 48 states saw a falling share of winter precipitation arrive as snow.
Extremes sit in the tails of the temperature range, so a warmer average pushes more days past old hot thresholds and fewer past cold ones. Attribution studies use models and observations to ask how much human warming changed an event's likelihood or intensity. The links are strongest for heat-driven events and weaker where atmospheric motion dominates. In EPA's Atlantic record through the early 2020s, hurricane counts show no clear trend since 1878 once observing changes are allowed for, yet the storms' combined intensity rose noticeably over the last 30 years, though the 1950s and 1960s were also active.
The trap is reading a steady yearly total as no change. Flood damage turns on thresholds, such as the water that overwhelms a storm drain. Say your town's yearly rainfall barely moves, but more of it arrives in heavy bursts: against unchanged drainage, the storm that floods your street returns more often.
Rule: separate total from intensity, and frequency from strength, before deciding whether a hazard has changed.
Feedbacks, sensitivity, and what models do
~2 min
Feedbacks such as water vapor and melting ice amplify a warming that something else starts, and climate sensitivity sums their effect. Models are tested against the observed record, and every projection is an if-then statement tied to an emissions scenario.
A feedback is a change set off by a first change that then amplifies or damps it. Water vapor responds to temperature, so it amplifies whatever started the warming. Melting sea ice exposes dark water that absorbs more sunlight, which warms the water and melts more ice. Thawing permafrost, releasing carbon because of the warming, is another amplifier. The basic brake is the Planck response: a warmer surface and atmosphere send more infrared out to space.
Equilibrium climate sensitivity is the global warming once climate settles after CO2 doubles from preindustrial levels. Recent community assessments narrowed its range for the first time in decades, and records of past climates such as the last ice age gave the strongest evidence against very high values.
Models are judged against the record they must explain. NASA notes that models including changes in the Sun's output cannot reproduce the past century's temperature trend without the rise in greenhouse gases. Processes too small for a global model to resolve, such as how plumes of injected aerosol spread, are represented through parameterizations. Given the same greenhouse increase, models still warm by different amounts; cloud feedback was long argued to be the main cause, but one 2020 study found much of that spread offset by other feedbacks.
The trap is reading a projection as a forecast. Suppose you see two warming figures for 2100 that differ widely. Before you call the models inconsistent, check whether they assume the same emissions. A projection estimates how long-term averages shift under an assumed emissions path; NASA's summary, for instance, gives its upper 2100 sea-level figure for a high-emissions scenario. Judging a model against a scenario that did not happen tests the scenario, not the physics.
Rule: for every projected number, name its scenario before comparing it with anything.
Who is affected, and how adaptation works
~2 min
How hard a hazard hits depends on exposure, sensitivity and the ability to recover, not only on the hazard. Adaptation adjusts to changes already underway; it works alongside limiting emissions rather than replacing it.
EPA names three factors behind how hard a hazard hits a person: exposure, being where it strikes; sensitivity, as with children or people with asthma; and adaptability, the means to prepare and recover. That is why one storm is a nuisance in one place and a disaster in another, and why the communities that contributed least to warming are often among the most vulnerable.
Impacts travel by many routes. Extreme heat is the leading cause of weather-related deaths in the U.S. About three in five North American bird species studied shifted their wintering grounds north since the 1960s, though about a quarter moved south. Corals bleach under heat stress; a bleached coral can recover if the stress eases, while prolonged stress kills it.
Mitigation limits how much the climate changes; adaptation adjusts to what is already underway. Neither replaces the other: some warming is already committed, and adaptation gets harder the faster climate changes. On coasts, responses fall into three families: protect with levees and seawalls, accommodate by floodproofing, or retreat to higher ground. Natural features buffer too; 15 feet of marsh absorbs half of incoming wave energy.
The trap is designing to the past. A culvert is sized for a storm with a set chance of being equaled or exceeded each year, its return period. When heavy rain grows more common, that design standard needs re-checking.
Rule: judge a climate risk by the hazard, who is exposed and how well they can recover, and re-check any design built on past extremes.
Cutting emissions, the energy system, and carbon removal
~2 min
The CO2 level keeps rising as long as emissions outpace natural sinks, even while emissions fall, and warming tracks the cumulative total. Avoiding an emission and removing CO2 from the air are different claims with different accounting.
Atmospheric CO2 works like a bathtub: emissions pour in, natural sinks drain about 55 percent of each year's human emissions, and while inflow beats the drain the level rises even as yearly emissions shrink. Warming has grown nearly in proportion to cumulative CO2, so any temperature limit implies a finite remaining total, whose size is uncertain. If emissions stopped entirely, NASA says, the temperature rise would begin to flatten within a few years, then plateau at an elevated level for centuries. Methane breaks down in about a decade, so cutting it slows warming quickly; added CO2 lasts thousands of years, so each ton keeps acting long after methane would have broken down.
Capacity is what a generator can produce; generation is what it actually delivers. Their ratio, the capacity factor, runs above about 70 percent for base-load plants and often below 15 percent for peaking units. Wind and solar generate only when their resource is available, so storage and grid operators match supply to demand at every moment. A source with no smokestack can still carry emissions from mining, construction or the grid that charges it, so honest comparisons count the whole life cycle.
The trap is blurring avoidance with removal. Capturing CO2 at a power plant keeps it out of the air; direct air capture pulls it back out and counts as removal only with lasting storage. Trees and soils store carbon too, but wildfire or a change in land management can release it, so the Department of Energy (DOE) lists durable, measurable storage as a key challenge.
Rule: when you meet a claim about cuts or removal, ask whether it slows the flow, ends net emissions, or takes carbon back out, and for how long it lasts.
Reading climate data and checking a claim
~2 min
Swings such as El Nino scatter single years around the long-term trend, so a short window can mislead. Trace a figure to its dataset, check its baseline and time window, and ask what physical pattern supports it.
In El Nino, the Pacific trade winds relax and warm water and rain shift east; La Nina is the cool phase. The two push global temperature and rainfall up and down from year to year. 1997 to 1998 brought one of the two largest El Nino events on record, so a trend line started in 1998 begins on a peak and can make warming look flat. Longer records show it kept going: EPA's indicators, as archived in January 2025, list 2023 as the warmest year and 2014 to 2023 as the warmest decade on record.
An anomaly is a difference from a chosen baseline, so moving from a 1951 to 1980 baseline to 1901 to 2000 shifts every value but not the shape; when you compare sources, compare slopes, not raw anomalies. A global mean hides local detail: in EPA's U.S. record through 2023, the North, the West and Alaska warmed most while parts of the Southeast changed little. Records fall even in a stable climate, but then highs and lows split about evenly; from 2000 to 2009, U.S. daily record highs outnumbered record lows two to one.
The trap is treating agreement as the argument. Climate scientists broadly agree that humans are causing global warming, but that agreement exists because measurements and multiple lines of evidence point the same way, and those are what you should check.
Rule: trace a number to its dataset, then check its baseline, its time window and the mechanism claimed before you weigh it.
Written by Keentune. We are not affiliated with or endorsed by the organizations whose documentation informs this guide, and any linked sources belong to their respective owners.
All exam, test, and product names and trademarks are the property of their respective owners and are used here for identification and reference only. Keentune is independent study practice — not affiliated with, authorized, or endorsed by any of these organizations.