The Science of Climate Change: What the Data Actually Says

In January 2026, six of the world’s major independent climate monitoring agencies — NASA, NOAA, Copernicus (ECMWF), Berkeley Earth, the UK Met Office’s HadCRUT, and the World Meteorological Organization — simultaneously released their annual global temperature analyses. They all reached the same conclusion, independently, using different datasets and different methodologies. The year 2025 was the third warmest in recorded history, at 1.47°C above the pre-industrial baseline — only marginally cooler than 2023’s 1.48°C and significantly cooler than 2024’s 1.60°C, which stands as the hottest year ever directly measured. The last 11 years have been, in order, the 11 warmest years ever recorded. The three-year average temperature for 2023–2025 has exceeded 1.5°C above pre-industrial levels for the first time in human history.

That 1.5°C threshold matters because it is the more ambitious of the two temperature limits specified in the 2015 Paris Agreement — the level of warming beyond which climate scientists assess that risks to human societies, ecosystems, and physical systems escalate sharply. The Paris Agreement’s 1.5°C target refers to sustained long-term warming, not single-year exceedance — but a three-year average above 1.5°C, during a period that included a La Niña cooling event that should have suppressed temperatures, is precisely the kind of milestone that signals that long-term warming is approaching the threshold. Copernicus Director Carlo Buontempo stated plainly: “We are bound to pass it; the choice we now have is how to best manage the inevitable overshoot and its consequences on societies and natural systems.”

This article presents the scientific evidence for climate change as it stands in 2026 — the data from temperature records, ocean heat measurements, carbon dioxide concentrations, sea ice, sea level, and extreme weather — with the clarity and specificity that the evidence warrants. This is not a political document. It is a factual account of what the world’s scientific monitoring infrastructure is measuring, why scientists interpret those measurements the way they do, and what the physical mechanisms connect human activity to observed climate changes. The data is publicly available from every agency cited here. The conclusions are the consensus product of thousands of independent researchers working on every continent.

The Temperature Record: What It Shows and How We Know

Global average surface temperature is measured from a network of more than 25,000 meteorological stations on land, combined with sea surface temperature measurements from ships and ocean buoys, and supplemented by satellite data and atmospheric reanalyses that fill gaps in the station network. The major agencies each process this data independently, using different statistical methods to handle missing data, account for the changing distribution of stations over time, and correct for urban heat island effects that could bias land measurements upward. The fact that all of these independent analyses — using different raw data sources and different analytical approaches — produce consistent temperature records that agree to within measurement uncertainty is one of the most important pieces of evidence that the global temperature trend is a real physical signal and not an artefact of any single measurement system or analytical methodology.

The temperature record extends back to approximately 1850, when reliable global-scale instrumental measurements begin. Before that, palaeoclimate evidence — ice cores, tree rings, coral records, sediment samples — provides evidence of past temperatures over thousands to millions of years. That proxy evidence consistently indicates that current temperatures are the highest they have been in at least several thousand years, and possibly since the last interglacial period approximately 120,000 years ago, when sea levels were several metres higher than today.

Since 1850, global average surface temperature has increased by approximately 1.2°C on the long-term trend, with most of the increase occurring after 1970 and the rate of warming accelerating in recent decades. The 2023–2025 period represents a step change beyond the previous temperature trajectory — 2023 in particular broke records by an unprecedented margin that prompted extensive scientific analysis to identify its causes. The primary factors identified were record-high greenhouse gas concentrations, an unusually strong El Niño event that peaked in late 2023 and 2024, and a reduction in marine aerosol pollution from shipping (following new low-sulphur fuel regulations) that inadvertently removed a temporary cooling effect. The El Niño has since weakened and transitional La Niña conditions emerged in late 2025 — yet 2025’s temperature was still the third highest on record, only marginally cooler than 2023. This indicates that the underlying long-term warming trend is the dominant signal, with natural variability providing temporary fluctuations around it.

The Cause: The Greenhouse Effect and Human Emissions

The physical mechanism connecting human activity to global temperature increase is well understood and has been for over a century. The basic greenhouse effect was described by scientists Eunice Newton Foote (1856) and John Tyndall (1859), who established that carbon dioxide and water vapour absorb infrared radiation — the heat that Earth’s surface radiates into the atmosphere. Without any greenhouse effect, Earth’s average temperature would be approximately negative 18°C; the natural greenhouse effect maintains it at approximately 15°C. Adding more greenhouse gases to the atmosphere strengthens this effect, trapping more heat.

Carbon dioxide (CO₂) is the primary driver of human-caused warming. Atmospheric CO₂ concentration was approximately 280 parts per million (ppm) in 1750, before the Industrial Revolution. By 1960, when systematic global monitoring began at Mauna Loa, it had reached approximately 315 ppm. In 2024, atmospheric CO₂ reached 424 ppm — the highest concentration in at least 3 million years. The 2024 increase alone was 3.75 ppm — the largest one-year increase on record. CO₂ concentrations have been rising consistently for 65 consecutive years of continuous Mauna Loa monitoring, without a single year of decline. The source of this increase is identifiable from isotopic analysis of the CO₂ itself: fossil fuel combustion and land-use change produce CO₂ with specific carbon isotope ratios that are measurably different from natural CO₂ sources, and the isotopic composition of atmospheric CO₂ has shifted precisely in the direction expected from an increasing fossil fuel contribution.

Methane (CH₄) and nitrous oxide (N₂O) are the other major human-emitted greenhouse gases — methane from fossil fuel operations, livestock agriculture, and landfills; nitrous oxide primarily from fertiliser use. Atmospheric methane concentration has increased by approximately 160 percent since pre-industrial times and continues to rise at an accelerating rate. All three major greenhouse gases reached record atmospheric concentrations in 2025, according to Copernicus Atmosphere Monitoring Service data.

Attribution science — determining what fraction of observed warming is caused by human activity versus natural factors — uses climate models and statistical techniques to separate human and natural contributions to observed temperature change. The results are unambiguous: without human greenhouse gas emissions, natural factors alone (solar variability, volcanic activity, natural climate cycles) cannot explain the observed temperature increase over the past century. Including human greenhouse gas emissions accurately reproduces the observed warming trend. The IPCC’s Sixth Assessment Report (2021-2022) stated that human influence has warmed the climate at an “unprecedented rate” and that it is “unequivocal” that human influence has warmed the atmosphere, ocean, and land.

Ocean Heat: The Most Important Climate Indicator

Global average surface temperature is the most widely reported climate metric, but ocean heat content is arguably the more fundamentally important one. Approximately 93 percent of the excess energy trapped by increased greenhouse gases goes into the oceans — the oceans’ vast heat capacity makes them the planet’s dominant heat reservoir. Surface air temperature fluctuates significantly with natural variability (El Niño, La Niña) from year to year; ocean heat content provides a more stable, cumulative measure of the energy imbalance driving long-term warming.

The ocean heat content record is unambiguous. 2025 was the warmest year on record for ocean heat content — the cumulative heat stored in the world’s oceans has increased by approximately 500 zettajoules since the 1940s (one zettajoule is a billion trillion joules). The increase in ocean heat in 2025 alone, compared to 2024, was approximately 23 zettajoules — equivalent to approximately 200 times the world’s total electricity generation in 2024, all absorbed by the ocean in a single year. The rate of ocean warming has accelerated over the past decade, and approximately 33 percent of the global ocean surface in 2025 ranked among its top three warmest historical conditions.

Warmer oceans have cascading consequences for the climate system and for humanity. They fuel more intense hurricanes and typhoons, which draw their energy from warm sea surface temperatures. They cause thermal expansion of seawater, contributing to sea level rise independently of ice melt. They bleach and kill coral reefs — the Great Barrier Reef experienced its most severe mass bleaching event on record in 2024. They disrupt marine ecosystems and the fisheries that feed hundreds of millions of people. And they absorb CO₂ more slowly as they warm — reducing the ocean’s capacity to act as a carbon sink and allowing more CO₂ to remain in the atmosphere per unit of emission.

The Weakening Carbon Sinks: A Critical Feedback

One of the most alarming findings in recent climate science — documented in the Global Carbon Project’s 2025 report — is evidence that natural carbon sinks are weakening. Approximately half of annual human CO₂ emissions are absorbed by natural processes: the land biosphere (forests, soils, vegetation) absorbs roughly 30 percent, and the oceans absorb roughly 22 percent. The remaining 48 percent accumulates in the atmosphere. This natural buffering has significantly slowed the rate at which atmospheric CO₂ increases relative to what human emissions alone would produce.

The 2025 analysis found evidence that climate change has caused a long-term decline in land and ocean carbon sink efficiency, with sinks being approximately 15 percent weaker over the past decade than they would have been without the impacts of climate change itself. The mechanism is partly self-reinforcing: warmer temperatures stress forests, making them more susceptible to wildfire, drought, and pest outbreaks that reduce their carbon uptake and in some cases convert them from carbon sinks to carbon sources. Record wildfire emissions — Europe recorded its highest-ever annual wildfire emissions in 2025 — represent the land carbon sink being partially reversed, with stored carbon being returned to the atmosphere. This feedback — warming causes sink weakening, which causes more CO₂ to accumulate, which causes more warming — is one of the mechanisms that makes climate projections highly dependent on the trajectory of emissions in the coming decades.

The 3.75 ppm increase in atmospheric CO₂ in 2024 — the largest single-year increase on record — reflects the combination of persistently high fossil fuel emissions and the weakening of natural carbon sinks. It is a direct measurement of the accelerating accumulation of CO₂ that drives warming.

Sea Ice, Glaciers, and Sea Level: The Physical Fingerprints

Climate change leaves physical fingerprints across the planet’s cryosphere — the ice-covered regions — that are directly measurable and broadly consistent with what physical models predict would happen as temperatures rise.

Arctic sea ice extent has been declining consistently since satellite observations began in 1979, with the minimum summer ice extent (September) showing the most dramatic decline. Monthly sea ice extent in 2025 set new record lows for January, February, March, and December — the beginning and end of the year saw the lowest Arctic sea ice coverage ever recorded for those months. Antarctica, which had shown more complex patterns in previous decades, also recorded its warmest annual temperature on record in 2025, and Antarctic sea ice extent reached its lowest combined polar value since satellite records began in February 2025. The loss of sea ice creates additional warming through the ice-albedo feedback: sea ice reflects approximately 80–90 percent of incoming solar radiation back to space, while open ocean absorbs approximately 94 percent of it. As sea ice retreats, more solar energy is absorbed by the dark ocean surface, amplifying Arctic warming — the Arctic is currently warming approximately three to four times faster than the global average.

Glacier retreat is documented on every continent where glaciers exist. The World Glacier Monitoring Service tracks thousands of reference glaciers globally and consistently reports net mass loss — glaciers are losing more ice to melting in summer than they gain from snowfall in winter, on an accelerating trend. Glaciers in the Alps, Himalayas, Andes, and other mountain regions serve as freshwater reservoirs for hundreds of millions of people whose water supplies depend on glacial meltwater during dry seasons. The long-term decline of these glaciers is creating what scientists call “peak water” — a period of initially increased meltwater followed by declining flows as the ice volume diminishes — with serious consequences for water security in glacier-dependent regions.

Global mean sea level has risen by approximately 20–23 cm since 1900, with the rate accelerating significantly in recent decades. Current sea level is rising at approximately 3.7 mm per year — more than twice the average rate of the 20th century — driven by the combination of thermal expansion of warming seawater (approximately 40–50 percent of recent rise) and melting of the Greenland Ice Sheet, Antarctic Ice Sheet, and mountain glaciers (approximately 50–60 percent combined). The Greenland Ice Sheet, if completely melted, would raise global sea levels by approximately 7 metres; the Antarctic Ice Sheet contains enough ice to raise sea levels by approximately 60 metres. Neither of these outcomes is projected within any plausible near-term scenario, but their ongoing mass loss contributes to sea level rise that threatens low-lying coastal cities and small island nations on timescales relevant to current infrastructure planning.

Extreme Weather: Attribution Science Connects Events to Climate

One of the most significant advances in climate science over the past decade has been the development of event attribution — the ability to quantify how much climate change altered the probability and intensity of a specific extreme weather event. Rather than simply stating that climate change makes extreme events “more likely,” attribution science now provides probability statements: for the 2023 European summer heatwaves, for example, attribution studies found that climate change made such temperatures essentially impossible without human-induced warming — the heatwaves would not have occurred in the pre-industrial climate.

In 2025, Copernicus data documented that 50 percent of global land experienced more days than average with at least “strong heat stress” — defined as a feels-like temperature of 32°C or above. Half the planet’s land surface was more often dangerously hot in 2025 than was historically typical. In 2025, 23 weather and climate events in the United States alone exceeded $1 billion in damage, causing $115 billion in total economic losses — making 2025 the third most expensive year for major weather disasters on record. Europe recorded its highest-ever annual wildfire emissions. Record-breaking heatwaves affected South and Southeast Asia, Europe, and parts of Africa. Severe flooding events struck across multiple regions. The WHO designates heat stress as the leading cause of weather-related fatalities worldwide.

Attribution science does not claim that climate change causes all extreme weather — these events have always occurred. It quantifies the change in probability: for heat extremes, climate change has made them significantly more frequent and more intense. For some types of precipitation extremes (heavy rainfall events), a warmer atmosphere holds more water vapour, intensifying rainfall events when they occur. For drought, higher temperatures increase evaporation and moisture stress, intensifying droughts in many already-dry regions. For hurricane intensity (though not necessarily frequency), warmer sea surface temperatures provide more energy, making peak wind speeds and rainfall rates higher in the most intense events.

The 1.5°C Threshold: What Exceeding It Means

The Paris Agreement’s 1.5°C target was established based on scientific assessments of how climate risks — to food systems, water availability, human health, biodiversity, and physical infrastructure — scale with temperature. The IPCC’s Special Report on 1.5°C (2018) synthesised the evidence: at 1.5°C, climate impacts are severe but manageable for most regions and populations; at 2°C, they are significantly worse across essentially every category; beyond 2°C, some risks become catastrophic and potentially irreversible on human timescales.

Specific differences between 1.5°C and 2°C warming that IPCC assessments have identified include: coral reef systems losing 70–90 percent of their coverage at 1.5°C versus effectively all reefs at 2°C; the Arctic experiencing a sea-ice-free summer once per century at 1.5°C versus once per decade at 2°C; approximately 420 million fewer people exposed to extreme heat waves and approximately 65 million fewer people exposed to exceptional heat waves at 1.5°C versus 2°C; approximately 50 percent of the global species range loss difference between 1.5°C and 2°C for insects and plants. The scientific evidence for why the 0.5°C difference matters is specific and substantial — it is not an arbitrary number chosen for political convenience.

The current trajectory of global greenhouse gas emissions — which reached a new record in 2024, according to the Global Carbon Project — is not consistent with limiting warming to 1.5°C. The remaining “carbon budget” — the total amount of CO₂ that can be emitted globally while maintaining a 50 percent probability of staying below 1.5°C — is approximately 250–350 gigatonnes of CO₂ at current emission rates. At 2024’s emission rate of approximately 37 gigatonnes per year, this budget would be exhausted within approximately 7–9 years without significant emissions reductions. The 2°C budget is approximately four times larger, but similarly requires rapid and deep emissions reductions to preserve.

What the Science Does and Does Not Say

The scientific evidence for human-caused climate change is robust, multi-evidenced, and cross-validated by independent monitoring systems in ways that leave no reasonable scientific doubt about the basic conclusions: the Earth is warming, the primary cause is human greenhouse gas emissions, and the physical consequences — sea level rise, ice loss, extreme weather intensification, ocean warming — are already measurable and accelerating.

What the science does not determine is the policy response. How rapidly to reduce emissions, which technologies to prioritise, how to distribute the costs and benefits of climate action globally, how to balance present and future welfare — these are questions of values, economics, politics, and ethics that go beyond the empirical domain of climate science. Scientists can characterise the physical risks associated with different temperature trajectories; they cannot determine which risks are acceptable or how to allocate the burden of avoiding them. That determination belongs to democratic institutions, international negotiations, and individual human choices.

What the science also does not guarantee is that any particular outcome is predetermined. The physical system responds to the cumulative total of CO₂ emissions over time, not to any single year’s emission level. Every tonne of CO₂ not emitted meaningfully affects the trajectory of warming, and every fraction of a degree of warming avoided meaningfully reduces climate impacts. The urgency of the scientific picture — the unprecedented three-year temperature average, the record ocean heat, the weakening carbon sinks, the accelerating sea level rise — reflects the cumulative consequence of past emissions and the challenge of current trajectories. It also defines, precisely, what the scientific evidence says about what is at stake in choices being made right now.