In the summer of 2021, a heat dome settled over the Pacific Northwest of North America. The town of Lytton, British Columbia, recorded a temperature of 49.6 degrees Celsius — nearly 5 degrees above the previous Canadian record — on June 29. The following day, Lytton burned to the ground in a wildfire. That same week, approximately 600 people died in British Columbia alone from heat-related causes, a toll more than ten times the annual average. Friederike Otto and the World Weather Attribution initiative published a rapid attribution analysis within weeks: a heat event of that magnitude was "virtually impossible" without the influence of human-caused climate change. The probability of such a heat dome occurring was increased by a factor of at least 150 by the warming that has already taken place.
For many people, the gap between "climate change is real" and understanding how it actually works remains wide. The headlines cycle through floods, fires, droughts, and diplomatic summits, but the underlying physics — why adding carbon dioxide to the atmosphere warms the planet, how fast things are changing, what the critical thresholds are — often gets lost in the political noise. This is a problem, because the details matter enormously for understanding which claims about climate change are well-supported, which are uncertain, and which are misleading in either direction.
The scientific understanding of climate change is not the product of a single model or a single research group. It is built on multiple independent lines of evidence — from nineteenth-century laboratory physics to ice cores to satellite measurements to isotopic chemistry — that converge on the same conclusion. Understanding that convergence is the beginning of understanding what is actually known, with what confidence, and where genuine scientific uncertainty remains.
"It is certain beyond reasonable doubt that the world has warmed since the late nineteenth century, that human activities — especially the burning of fossil fuels — are the main reason, and that further warming will occur if greenhouse gas emissions continue to rise." — IPCC Sixth Assessment Report, Working Group I Summary for Policymakers, 2021
Key Definitions
Greenhouse effect: The process by which atmospheric gases — primarily water vapour, CO2, methane, and nitrous oxide — absorb outgoing infrared radiation from Earth's surface and re-emit it, trapping heat and warming the surface above what it would be in their absence.
Climate sensitivity: The equilibrium global average surface temperature increase resulting from a doubling of atmospheric CO2. The IPCC AR6 best estimate is 3 degrees Celsius, with a likely range of 2.5-4 degrees Celsius.
Keeling Curve: The continuous record of atmospheric CO2 concentration measured at Mauna Loa Observatory, Hawaii, since 1958 by Charles David Keeling. The record shows CO2 rising from 315 ppm in 1958 to over 420 ppm today, with an annual cycle from seasonal plant growth.
Tipping point: A threshold in the Earth system beyond which self-reinforcing feedbacks drive a component to shift to a qualitatively different state — potentially irreversibly — regardless of further human action.
Attribution science: The scientific discipline that quantifies the degree to which human-caused climate change altered the probability or magnitude of a specific weather event.
Radiative forcing: A measure of the energy imbalance imposed on the climate system by a change in atmospheric composition or other factor, expressed in watts per square metre. Positive forcing (as from increased CO2) warms the planet; negative forcing (as from aerosol pollution) cools it.
Mitigation: Actions that reduce greenhouse gas emissions or remove atmospheric carbon, limiting the degree of future warming.
Adaptation: Adjustments in human systems and infrastructure to reduce harm from warming that has already occurred or is projected.
Key Climate Indicators and Their Current Trajectory
| Indicator | Pre-Industrial Baseline | Current Value (2024) | Trend |
|---|---|---|---|
| Atmospheric CO2 | ~280 ppm (Holocene) | ~422 ppm | +2.5 ppm/year |
| Global average temperature | 1850-1900 average | +1.2 degrees C above baseline | +0.2 degrees C per decade |
| Arctic sea ice extent (summer) | Pre-1980 average | ~40% below 1980 average | Declining |
| Global mean sea level | 1900 baseline | +20 cm above 1900 | +3.7 mm/year (accelerating) |
| Ocean heat content | Pre-industrial baseline | Record highs every year since 2017 | Rising |
| Frequency of extreme heat events | Historical baseline | 4-5x more likely globally | Increasing |
Sources: IPCC AR6, NASA, NOAA, Copernicus Climate Change Service
The Physics: From Arrhenius to the IPCC
The greenhouse effect is among the best-established phenomena in atmospheric science. Its basic mechanism — that certain gases absorb and re-emit infrared radiation, trapping heat in the lower atmosphere — was demonstrated experimentally by John Tyndall in a series of laboratory experiments published in 1859. Tyndall showed that water vapour and CO2 were the primary heat-trapping gases in Earth's atmosphere, while nitrogen and oxygen (which make up 99% of the atmosphere) are essentially transparent to infrared radiation.
In 1896, Swedish chemist Svante Arrhenius published "On the Influence of Carbonic Acid in the Air upon the Temperature of the Ground," a paper that made the first quantitative estimate of how much warming would result from doubling atmospheric CO2. Arrhenius calculated approximately 5-6 degrees Celsius — achieved through a remarkable feat of manual computation that took him nearly a year. Modern estimates, using far more sophisticated models and observations, put the best estimate at 3 degrees Celsius, with a likely range of 2.5-4 degrees Celsius. That Arrhenius's back-of-envelope calculation in 1896 was roughly in the right order of magnitude is itself testimony to the robustness of the underlying physics.
The Keeling Curve provides the most direct evidence of what humans have done to the atmosphere. Charles David Keeling began measuring CO2 at Mauna Loa Observatory in Hawaii in 1958, establishing the first continuous, precise atmospheric record. When he started, CO2 was 315 parts per million (ppm). By the time of his death in 2005, it had risen to 380 ppm. Today it exceeds 420 ppm — a 50% increase above the pre-industrial level of approximately 280 ppm that prevailed for the approximately 10,000 years of the Holocene. Ice cores from Antarctica and Greenland, which trap bubbles of ancient air, confirm that current CO2 concentrations are the highest in at least 800,000 years, covering multiple ice age cycles.
The Evidence for Human Causation
The attribution of observed warming to human activities rests on multiple independent lines of evidence whose convergence is the scientific foundation of climate change research.
The isotopic fingerprint is one of the most elegant. Fossil fuels are ancient organic carbon — they are depleted in the heavy carbon isotopes carbon-13 and carbon-14 compared with atmospheric carbon. When fossil fuels are burned, they release carbon with this distinctive isotopic signature. Atmospheric CO2 has shown exactly this isotopic shift since industrialization, confirming that the rising CO2 comes from fossil fuel combustion rather than from volcanic outgassing or ocean warming releasing dissolved CO2.
The spatial pattern of warming provides another independent line of evidence. The observed warming pattern — greatest at the poles (Arctic amplification), larger over land than ocean, greater at night than day, and accompanied by stratospheric cooling (as the troposphere warms and traps more heat below) — matches the theoretical fingerprint of greenhouse gas forcing and differs from the patterns expected from solar variability, volcanic activity, or internal climate oscillations. The stratospheric cooling is particularly diagnostic: a brighter sun would warm both troposphere and stratosphere; greenhouse gas forcing warms the lower atmosphere while the upper atmosphere cools.
Peter Stott, Myles Allen, Simon Tett, and colleagues developed the formal "optimal fingerprinting" method for climate attribution in the late 1990s, providing a statistical framework for separating human and natural influences on climate records. Subsequent decades of work have strengthened the case: natural factors alone cannot account for the observed warming since 1950. When human forcings — primarily CO2 and other greenhouse gases, partially offset by aerosol cooling — are included, models reproduce the observed temperature record well.
How Much Warming, How Fast: The IPCC Assessment
The Intergovernmental Panel on Climate Change, established in 1988, synthesizes peer-reviewed climate research in periodic assessment reports. Its Sixth Assessment Report (AR6), released between 2021 and 2022 in three volumes covering physical science, impacts and adaptation, and mitigation, represents the most comprehensive review of climate science available.
The AR6 Working Group I report states that global average surface temperature has risen by approximately 1.1 degrees Celsius above pre-industrial (1850-1900) levels as of the period 2011-2020, and that it is "unequivocal" that human influence is the dominant cause. Under the highest-emission scenario (SSP5-8.5, roughly consistent with current policies without additional action), the AR6 projects warming of approximately 3.3-5.7 degrees Celsius by 2100. Under the lowest-emission scenario (SSP1-1.9, requiring very rapid and deep emissions cuts), warming would likely be kept below 1.5 degrees Celsius by the end of the century, though it would overshoot 1.5 degrees temporarily during the mid-century.
The 1.5-degree and 2-degree targets referenced in the Paris Agreement are not arbitrary. They reflect an assessment by the IPCC's Special Report on 1.5 degrees (2018) of meaningfully different outcomes. At 1.5 degrees, coral reef systems face severe stress; at 2 degrees, virtually all tropical reefs are projected to bleach annually. At 1.5 degrees, Arctic summers remain ice-covered in one-in-ten years; at 2 degrees, summer ice disappears roughly once per decade. Sea level rise projections, extreme heat exposure, and agricultural impacts all worsen substantially between 1.5 and 2 degrees. The half-degree difference matters at the scale of hundreds of millions of affected people.
Climate Sensitivity: The Central Uncertainty
Climate sensitivity — the equilibrium warming from a doubling of CO2 — is one of the most important quantities in climate science and has been a source of persistent uncertainty. The 1979 Charney Report, the first major scientific assessment of the problem, estimated a likely range of 1.5-4.5 degrees Celsius. That range has remained essentially unchanged for forty years, though the AR6 has slightly narrowed and shifted it to 2.5-4 degrees Celsius (likely range) with a best estimate of 3 degrees.
Why hasn't the range narrowed more dramatically? Climate sensitivity is difficult to measure directly because the equilibrating processes operate on timescales of decades to centuries. Estimates come from three independent approaches: climate models (where sensitivity emerges from simulated physical processes), paleoclimate data (using past climate states where CO2 was higher or lower), and the instrumental record of observed warming relative to estimated forcing since 1850. Each approach has its own uncertainties, and they have not converged as tightly as researchers hoped.
The remaining uncertainty range from 2.5 to 4 degrees has substantial practical significance: at the low end, ambitious mitigation buys considerably more time and the worst outcomes are more avoidable; at the high end, even aggressive action may not prevent severe impacts. However, this uncertainty does not imply that doing nothing is rational — a wider uncertainty range does not mean the expected damage is lower; if anything, greater uncertainty about high-end outcomes increases the expected damage when the distribution is skewed toward catastrophic outcomes.
Feedbacks and Amplification
The direct warming effect of CO2 doubling — ignoring feedbacks — is approximately 1.2 degrees Celsius. The reason best estimates run to 3 degrees is that initial warming triggers feedbacks that amplify the effect.
The water vapour feedback is the most important. A warmer atmosphere holds more water vapour, which is itself a powerful greenhouse gas. As CO2 warms the surface, more water evaporates, adding water vapour to the atmosphere, which causes additional warming — roughly doubling the direct CO2 effect. This feedback is well-understood physically and is among the most robustly simulated processes in climate models.
The ice-albedo feedback amplifies warming at high latitudes. Ice and snow are highly reflective; when they melt, they expose darker ocean or land surface, which absorbs more solar radiation, warming the surface further and causing more melt. This is the primary mechanism behind Arctic amplification — the Arctic has warmed approximately three to four times as fast as the global average over recent decades.
Cloud feedbacks are the largest source of uncertainty in climate sensitivity. Low-level marine clouds cool the surface by reflecting sunlight. Whether these clouds increase or decrease in a warmer world — and by how much — depends on complex small-scale processes that climate models struggle to simulate accurately. Recent research, including analysis of observed cloud responses to volcanic eruptions and satellite data, has modestly constrained the cloud feedback uncertainty, contributing to the slight narrowing of the sensitivity range in AR6.
Carbon cycle feedbacks add further complexity. Warmer temperatures accelerate the decomposition of soil organic matter by microbial activity, releasing CO2 and methane that would otherwise remain locked in the ground. Permafrost in the Arctic holds vast stores of frozen organic carbon — estimated at roughly twice the current atmospheric CO2 load — and its thaw is a carbon feedback of potentially enormous consequence. The exact magnitude and timing of permafrost carbon release remains uncertain, but it is not included in most mitigation scenario calculations, meaning current projections may underestimate warming.
Tipping Points: The Cascade Risk
The tipping point concept has moved from the margins to the centre of climate risk assessment. Timothy Lenton (University of Exeter), Johan Rockstrom (Potsdam Institute for Climate Impact Research), Will Steffen, and colleagues outlined the framework in a 2018 paper in the Proceedings of the National Academy of Sciences and a 2019 commentary in Nature. Their analysis identifies sixteen potential tipping elements — components of the Earth system that could shift to qualitatively different states — and argues that several may be approaching critical thresholds at current warming levels.
The Atlantic Meridional Overturning Circulation (AMOC) is the ocean circulation system that carries warm surface water northward into the North Atlantic and returns cold, dense water southward at depth. It is the reason northern Europe has a much milder climate than its latitude would suggest — London, at 51 degrees north, is closer to the latitude of Calgary than Miami, yet has a far warmer winter. Paleoclimate records show AMOC has collapsed suddenly in the past (during Heinrich events and the Younger Dryas), with dramatic effects on European climate. A 2021 study by Boers in Nature Climate Change, using statistical methods to detect early warning signals in observational data, found patterns consistent with AMOC losing resilience. A collapse would substantially cool northern Europe while disrupting monsoon patterns globally.
The West Antarctic Ice Sheet contains enough ice to raise global sea levels by approximately 3.3 metres if it fully collapsed. Its base sits below sea level — meaning that as warm ocean water intrudes, the grounding line retreats further below sea level, potentially triggering self-sustaining instability through the "marine ice sheet instability" mechanism identified by Johannes Weertman in 1974 and extensively studied since. Multiple studies have found that Thwaites Glacier, a major contributor, may already be committed to significant retreat.
Amazon dieback represents a tipping point driven by both climate change and deforestation. The Amazon generates approximately half of its own rainfall through transpiration; deforestation reduces this moisture recycling, reducing rainfall, which stresses remaining forests, which reduces transpiration further — a savannification feedback. Research by Carlos Nobre, Thomas Lovejoy, and colleagues suggests a tipping point at 20-25% deforestation, above which large-scale Amazon dieback could occur regardless of climate change. Current deforestation has reached approximately 17-20%, depending on methodology.
The 2022 paper by Lenton and colleagues in Science estimated that several tipping points may be crossed at warming levels between 1.5 and 2 degrees Celsius, raising the stakes of the current policy trajectory considerably beyond what was assumed in earlier assessments.
Attribution Science: Connecting Emissions to Events
One of the most significant developments in climate science over the past two decades has been the emergence of event attribution — the ability to quantify how human-caused climate change altered the probability or intensity of specific weather events.
Friederike Otto, now at Imperial College London, and Myles Allen at Oxford developed the probabilistic attribution framework and co-founded the World Weather Attribution (WWA) initiative. The approach compares the probability of an event in the observed world (with human-caused warming) to its probability in a counterfactual world (as if industrialization had not occurred), using ensembles of climate models. The method has been applied to dozens of events: the 2021 Pacific Northwest heat dome, the 2019-2020 Australian fires, the 2022 Pakistan floods, European heat waves, and more.
The results are often striking. The 2021 Pacific Northwest event was found virtually impossible without climate change. The 2022 Pakistan floods, which submerged one-third of the country, were made approximately 75% more likely by the warming that has occurred. European heat waves in 2003 (which killed approximately 70,000 people) were made roughly twice as likely by climate change, and the 2019 European heat wave was made five times more likely. Attribution science has transformed the conversation about individual extreme events from "you can't link any single event to climate change" to precise probabilistic statements — a major advance for public understanding and increasingly for climate litigation.
Economic Costs: Nordhaus versus Stern
Estimating the economic cost of climate change requires combining physical projections with economic models in ways that involve substantial uncertainties and value judgments. The two most influential frameworks have been associated with William Nordhaus at Yale and Nicholas Stern at the London School of Economics.
Nordhaus developed the Dynamic Integrated Climate-Economy (DICE) model, for which he received the Nobel Prize in Economics in 2018. His model uses a discount rate that weights present welfare heavily relative to future welfare — reflecting the reasoning that future generations will likely be wealthier and better able to address impacts — and arrives at optimal carbon prices in the range of $30-50 per tonne in the near term, rising gradually. Critics, led by Stern, argue that the discount rate choice in Nordhaus's framework essentially discounts the welfare of future generations to near-zero, which is ethically indefensible. Stern's 2006 review of climate economics argued for near-zero pure time preference and arrived at carbon prices in the range of $85-100 per tonne (in 2006 dollars), implying much more aggressive near-term action.
This discount rate debate is not purely technical — it embeds fundamental ethical questions about intergenerational justice and the appropriate weighting of current versus future welfare. More recent damage function research suggests that both Nordhaus and Stern may have underestimated economic damages: Solomon Hsiang, Marshall Burke, and Edward Miguel's 2017 paper found that climate impacts on economic growth are substantially larger than integrated assessment models assumed, particularly for developing countries already near the warm end of the productivity-temperature relationship.
Solutions: The Evidence Base
The IPCC AR6 Working Group III report on mitigation, published 2022, synthesizes evidence on the cost and potential of different decarbonisation strategies. Several findings stand out for their robustness.
The cost of solar photovoltaics fell by approximately 90% between 2010 and 2022, driven by learning curves, manufacturing scale, and policy support. Wind power costs fell approximately 70% over the same period. Both are now typically the cheapest new electricity source in most of the world, according to IEA data. The speed of this cost decline was faster than most projections anticipated as recently as 2015, and it has dramatically changed the economics of decarbonising the power sector.
Electrification — of transport, heating, and industrial processes — combined with power sector decarbonisation, is the core strategy in most deep-decarbonisation pathways. Electric vehicles now have lower lifetime costs than internal combustion equivalents in most markets. Heat pumps deliver three to four units of heat for each unit of electricity consumed, making them dramatically more efficient than gas heating even when the grid is not fully decarbonised. Industrial processes (cement, steel, chemicals) are harder to electrify and represent the most challenging decarbonisation challenge, requiring process innovation, hydrogen, or carbon capture.
Carbon pricing is the most broadly endorsed policy mechanism among economists. By putting a price on emissions, it creates incentives throughout the economy for least-cost abatement without requiring governments to pick specific technologies. The EU Emissions Trading System (EU ETS) has reduced power sector emissions significantly, though its early periods of over-allocation of permits limited its effectiveness. British Columbia's carbon tax, in place since 2008, has been extensively studied: Kathryn Harrison and others have found it reduced gasoline consumption by approximately 7% relative to the rest of Canada without significant economic harm.
Katharine Hayhoe, a climate scientist at Texas Tech and author of "Saving Us" (2021), has written extensively on climate communication, arguing that most people — including many who express scepticism — hold underlying values (love of place, concern for community, care for children) that are entirely compatible with supporting climate action when engaged on those terms rather than through political framing. Hayhoe's research and practice emphasize that the political polarisation around climate is largely a consequence of cultural and partisan identity dynamics, not genuine scientific uncertainty.
Mitigation, Adaptation, and Justice
Climate change is not equally distributed. The countries most vulnerable to its impacts — low-lying island nations, sub-Saharan Africa, South and Southeast Asia — have contributed the least to cumulative emissions. The concept of climate justice, elaborated by researchers including Henry Shue ("Basic Rights," 1980, and "Climate Justice," 2014) and Mary Robinson, holds that this asymmetry creates obligations on the part of high-emitting wealthy countries to both reduce emissions more rapidly and finance adaptation and loss-and-damage compensation for vulnerable nations.
The Loss and Damage fund, agreed at COP27 in Sharm el-Sheikh in 2022, represents the first formal acknowledgment that some impacts are beyond adaptation — that for communities losing their islands, their glaciers, or their traditional food systems, adaptation is insufficient and compensation is owed. The fund exists but remains woefully underfunded relative to projected needs.
Adaptation with the strongest evidence includes: sea wall construction and managed retreat from high-risk coastal zones; crop variety development for heat and drought tolerance (pursued through institutions including the CGIAR network); early warning systems and disaster preparedness; urban greening to reduce heat island effects; and building codes and urban planning standards adapted to projected climate conditions. The IPCC AR6 Working Group II report notes that adaptation options are already being implemented across sectors but that there is a substantial gap between current adaptation investment and what is needed.
The interaction between mitigation and adaptation deserves emphasis: every tonne of CO2 not emitted is adaptation investment — it reduces the severity of impacts that future generations will need to adapt to. The two strategies are complements, not alternatives. The challenge is that mitigation requires collective action and offers diffuse global benefits, while adaptation is often local and immediate. This creates political economy pressures that tend to favour adaptation investment over mitigation, even though the long-run logic runs in the opposite direction.
For related analysis of the policy frameworks for addressing climate change, see What Is Carbon Pricing. For the ethical dimensions of climate justice and intergenerational responsibility, see What Is Climate Justice. For how tipping points and feedback loops function more broadly as systems phenomena, see How Feedback Loops Work.
References
- IPCC. Sixth Assessment Report (AR6): Working Group I — The Physical Science Basis. Cambridge University Press, 2021. https://www.ipcc.ch/report/ar6/wg1/
- IPCC. Sixth Assessment Report (AR6): Working Group II — Impacts, Adaptation and Vulnerability. Cambridge University Press, 2022. https://www.ipcc.ch/report/ar6/wg2/
- IPCC. Sixth Assessment Report (AR6): Working Group III — Mitigation of Climate Change. Cambridge University Press, 2022. https://www.ipcc.ch/report/ar6/wg3/
- Arrhenius, Svante. "On the Influence of Carbonic Acid in the Air upon the Temperature of the Ground." Philosophical Magazine and Journal of Science 41(251): 237-276, 1896.
- Lenton, Timothy M., Johan Rockstrom, Owen Gaffney, et al. "Climate Tipping Points — Too Risky to Bet Against." Nature 575: 592-595, 2019. https://doi.org/10.1038/d41586-019-03595-0
- Lenton, Timothy M., David I. Armstrong McKay, Sina Loriani, et al. "The Global Tipping Points Report 2023." University of Exeter, 2023. https://global-tipping-points.org/
- Philip, Sjoukje, Sarah Kew, Geert Jan van Oldenborgh, et al. "Rapid Attribution Analysis of the Extraordinary Heat Wave on the Pacific Coast of the US and Canada, June 2021." Earth System Dynamics 13(4): 1689-1713, 2022. https://doi.org/10.5194/esd-13-1689-2022
- Nordhaus, William D. "Revisiting the Social Cost of Carbon." Proceedings of the National Academy of Sciences 114(7): 1518-1523, 2017.
- Stern, Nicholas. The Economics of Climate Change: The Stern Review. Cambridge University Press, 2007.
- Hsiang, Solomon, Robert Kopp, Amir Jina, et al. "Estimating Economic Damage from Climate Change in the United States." Science 356(6345): 1362-1369, 2017. https://doi.org/10.1126/science.aal4369
- Hayhoe, Katharine. Saving Us: A Climate Scientist's Case for Hope and Healing in a Divided World. Atria/One Signal Publishers, 2021.
- Keeling, Charles David. "The Concentration and Isotopic Abundances of Atmospheric Carbon Dioxide in Rural Areas." Geochimica et Cosmochimica Acta 13(4): 322-334, 1958.