The planet's warming stripes, 1850–2025
Each stripe is one year's global average temperature vs the 1961–1990 baseline (HadCRUT5). Same color scale as every location on this site — which is why the planet's stripes look paler than your city's: two thirds of Earth is ocean, and oceans warm more slowly than land. Your location has probably warmed more than this.
CO₂ in the air — measured, not modeled
Monthly average CO₂ at Mauna Loa Observatory, Hawaiʻi, since 1958 (the “Keeling curve”, NOAA). The zig-zag is the northern hemisphere’s vegetation breathing — drawing carbon down each summer, releasing it each winter. The relentless climb underneath is us.
Why a single mountain in Hawaiʻi?
CO₂ mixes through the whole atmosphere within about a year, so a clean, remote station measures the planetary background, not local traffic. Mauna Loa is simply the longest continuous record — hundreds of stations worldwide, run by different countries, show the same curve.
800,000 years of CO₂ — and then us
Air bubbles trapped in Antarctic ice preserve ancient atmosphere. Through eight ice ages, CO₂ moved between ~180 and ~300 ppm. The vertical line at the right edge is the industrial era.
What am I looking at?
Each point is CO₂ from ice cores (Bereiter et al. 2015 composite), joined at the right by modern direct measurements. The slow waves are ice ages coming and going over tens of thousands of years — driven by wobbles in Earth’s orbit, with CO₂ amplifying the change. What is different now is not just the level — the highest in this entire record — but the speed: a rise that used to take 10,000 years now happens in about 30.
Global temperature, 1850–2025
Yearly global average vs the 1961–1990 baseline (HadCRUT5) — the same chart every location page shows for your place, drawn for the whole planet. The dark line is the 10-year average: that is climate; the single bars are weather.
How do we know it's us and not the sun?
Three independent lines of evidence. Timing: warming tracks CO₂ precisely, while the sun’s output has been flat to slightly declining since the 1980s — the decades of fastest warming. Fingerprint: the lower atmosphere warms while the upper atmosphere cools — exactly what greenhouse physics predicts, and the opposite of what a hotter sun would do. Nights and winters warm fastest — again the greenhouse signature. This is why the IPCC calls human influence “unequivocal”. And these charts need no climate models at all — they are thermometer and air measurements.
The futures still open — scenarios, not forecasts
Nobody can “predict” 2100, because the biggest unknown is a human choice: how much more we emit. Science therefore works with scenarios — consistent “what if” storylines. Solid line: measured history. Dashed lines: IPCC assessed best estimates per scenario (vs the 1850–1900 pre-industrial baseline).
Which scenario does this site use for your city — and why?
The “Outlook to 2050” on every location page uses three global high-resolution climate models (CMIP6 HighResMIP) on the high-emission pathway, bias-corrected against the same ERA5 data as our historical charts.
How that relates to the chart above — both come from the same scientific project (CMIP6), but from two different experiments that trade off breadth against detail:
| The global chart above | Your city’s outlook | |
|---|---|---|
| Experiment | ScenarioMIP (IPCC assessed) | HighResMIP |
| Scenarios | all four SSPs | one — borrows SSP5-8.5 as its forcing |
| Resolution | ~100–250 km grid | ~25 km — city scale |
| Ends | 2100 | 2050 (by experiment design) |
| Good for | comparing the futures we could choose | seeing one future at your place |
Why the high-emission one? Two honest reasons. First, it is the only pathway the high-resolution experiment ran — city-scale model runs are so expensive that the scientific community computed just one, ending in 2050. Second, before mid-century it barely matters: the pathways only diverge strongly after 2050, because near-term warming is largely already committed. Up to 2050 the scenarios differ by only ~0.2 °C — less than the model spread we show.
Worth knowing: as a story about 2100, SSP5-8.5 is now widely considered unlikely — it assumes a return to massive coal growth, and current policies track closer to the middle pathways. That criticism applies to the era we deliberately don’t show. For the 2050 window on our city pages, the high-emission and middle pathways are nearly indistinguishable, so the outlook is a fair approximation of any future short of rapid decarbonization — not a worst-case exaggeration.
What the models cannot do: their grid cells are ~25 km, so they smooth over neighborhoods, and they underestimate local extremes — your city’s worst future heatwaves will likely be worse than the model average suggests. The band on your city’s outlook chart shows where the three models disagree, not a guarantee.
Common questions
“The climate has always changed — why is this different?”
True — and the ice-core chart above shows it. What is different is the cause and the speed. Past changes followed Earth’s orbital wobbles over tens of thousands of years; today’s CO₂ spike is measurably from fossil carbon (its isotopic fingerprint is distinct) and is happening ~100× faster. Ecosystems and societies adapted to the slow version; the fast version is the problem.
“It was cold last winter — so much for warming.”
Weather is a single roll of the dice; climate is the loaded dice. Cold snaps still happen on a warming planet — they are just becoming rarer, while record-warm days outnumber record-cold ones by a growing margin. That is why every chart on this site emphasizes 10-year averages over single years.
“Can we trust these datasets?”
The global temperature series is produced independently by rival teams (UK Met Office, NASA, NOAA, Berkeley Earth — the last one founded by a former skeptic to check the others). They use different methods and agree within hundredths of a degree. CO₂ is simpler still: direct air measurement, replicated at hundreds of stations. None of this depends on climate models.
“Is it too late?”
No — and “too late” is the wrong frame. Warming stops roughly when emissions reach net zero, and every tenth of a degree avoided is measurably fewer heatwaves, droughts, and floods. The gap between the scenario lines above is not fate; it is the decision space. The difference between 1.8 °C and 4.4 °C is a choice being made now.
Global averages hide your story
The planet as a whole: about +1.3 °C. But nobody lives in the global average — land warms faster than oceans, and your place has its own stripes.
Sources
Global temperature: HadCRUT5 (Met Office Hadley Centre / CRU). CO₂: NOAA Global Monitoring Laboratory, Mauna Loa. Ice cores: Bereiter et al. 2015 composite via NOAA NCEI. Scenario estimates: IPCC AR6 WG1, Summary for Policymakers. Local projections: CMIP6 HighResMIP via Open-Meteo. All datasets are bundled with this site (no third-party requests) and refreshed about yearly.