Vital signs of a warming world

A snapshot, not a live feed — the numbers are embedded as of August 2026 and each series is credited to its source. The raw files are kept alongside the project so every line here is traceable back to the agency that measured it.

1 · The headline: global temperature

The single most-watched number in climate science is the global average surface temperature, expressed as an anomaly — how far each year sits above or below the 1951–1980 average. It wanders year to year with El Niño and volcanoes, but the trend since roughly 1970 is a near-straight climb of about 0.2 °C per decade. The last decade holds essentially every warmest year on record.

2 · The cause: atmospheric CO₂

Temperature is the symptom; the greenhouse gases are the mechanism. The Keeling Curve — carbon dioxide measured continuously at Mauna Loa since 1958 — is the cleanest line in all of Earth science. The little annual sawtooth is the planet breathing: CO₂ dips each Northern-Hemisphere summer as plants grow, then rebounds. Underneath it, the relentless rise from about 315 ppm to well past 420 ppm — a level the ice-core record says the atmosphere hasn’t seen in millions of years, and far above the pre-industrial ~280 ppm.

3 · Where the heat actually goes: the ocean

More than 90% of the extra energy trapped by greenhouse gases ends up in the ocean, not the air — which is why ocean heat content is arguably a steadier signal than surface temperature. It rises almost monotonically, with far less year-to-year wiggle, because water is a vast, slow thermal battery. The 0–700 m layer goes back to 1955; the deeper 0–2000 m record (dashed) starts with the Argo float era around 2005 and climbs even faster.

4 · A rising ocean

Warm water expands, and melting land ice adds volume — so the sea rises. Satellites have measured global mean sea level continuously since 1993 to millimetre precision, and the slope is about 3–3.5 mm per year and steepening. It looks small until you remember it’s a global average of the entire ocean surface, and that the curve is accelerating, not just rising.

5 · The vanishing ice: Arctic sea ice

Each September the Arctic Ocean’s floating ice shrinks to its yearly minimum. Satellites have tracked that minimum since 1979, and it has fallen by roughly 12–13% per decade — the clearest visible symptom of a warming north. (This is sea ice, which floats, so its melting doesn’t itself raise sea level; the land ice in the next chart does.)

6 · The land ice that does raise the sea

The Greenland and Antarctic ice sheets are the giants. A pair of satellites called GRACE weighs them from orbit by sensing tiny changes in Earth’s gravity, and both have been shedding mass since measurements began in 2002 — Greenland fastest, at thousands of gigatonnes lost. Every gigatonne that leaves the land ends up in the ocean.

7 · A fingerprint you feel: extreme rain

Warmer air holds more moisture — about 7% more per degree — so when it rains, it can pour. The honest caveat: total precipitation has no clean global trend, because a warming world redistributes rain rather than simply adding it. The signal lives in the extremes. This regional example — the share of the contiguous United States seeing a much-greater-than-normal amount of rain fall in single-day downpours — has drifted upward from its ~10% baseline, with the smoothed line clearly above where it sat a century ago.

Reading them together — and the honest caveats

Any one of these lines has a story you could quibble with. What’s striking is that seven independent measurement systems — ground thermometers, an infrared gas analyser, ocean floats, orbiting radar altimeters, passive-microwave sensors, gravity satellites, and a rain-gauge network — all point the same direction over the interval each one covers. That convergence is the argument.

• Anomalies, not absolutes. Temperature and ocean heat are shown as departures from a baseline period, which is standard practice but means the zero line is a choice, not a physical origin.
• Different windows. These records start when their instruments did — 1880 for thermometers, 1958 for Mauna Loa, 1979 for sea-ice satellites, 1993 for altimetry, 2002 for GRACE. Comparing slopes across different windows is apples-to-oranges; each line is only claimed over its own span.
• The ice-sheet series ends in 2020 in this packaging, and extreme precipitation is US-only — both noted at the charts. They’re the accessible, cleanly-sourced versions, not the last word.
• A snapshot. Everything here is embedded as of August 2026 and won’t update on its own; refreshing means re-pulling from the same agencies.
• Correlation vs. mechanism. These charts show the what. The why — that rising CO₂ is driving the warming — rests on the physics of the greenhouse effect, not on these curves alone.

So why doesn’t everyone believe it? A view from six Ontario cities

Every chart above is a global or continental average, and every one is smooth and unambiguous. But nobody lives in “the globe.” You live in one place, and remember the last few years — and at that scale the climate signal is buried in weather noise. That gap, more than anything, is why the science and the sidewalk disagree.

Here are six Ontario cities, from Thunder Bay in the north to Windsor in the south, each an independent homogenized temperature record from Environment Canada, plotted on the same baseline as the global line above. The thin lines are the cities; the bright line is their six-city average; the red dashed line is the global GISTEMP curve from the top of this page. Click any legend entry to show or hide it (or use Hide all and add lines back one at a time) — isolating the global line against a single city makes the difference in steadiness obvious. The scale stays fixed so the comparison is honest.

Look at any single city and you can see why someone would doubt the trend. The lines thrash up and down by two or three degrees from one year to the next — and against that, a warming rate of ~0.2 °C per decade is almost invisible over the span a person actually remembers. Sudbury’s most recent year in this record even lands below the 1951–1980 average. Cherry-pick a cold recent winter against a warm year a decade ago and you can “show” cooling in a way that feels honest to the person doing it.

Now watch what averaging does. The six-city average (bright line) is already far calmer than any single city, and the global line (red) — an average of thousands of stations — is smooth enough to see the trend plainly. That’s not a trick of the adjustment; it’s arithmetic. Independent noise cancels when you average (roughly by the square root of the number of stations), leaving the shared signal. Over the full 128-year overlap, the Ontario average and the globe actually agree on the warming to within a few hundredths of a degree — the cities are warming. You just can’t see it from your own back yard.

Three honest local wrinkles that give skeptics real-feeling ammunition, none of which overturn the trend:

• The warming hides where people look less. In this region it shows up most in winter and in overnight lows, not in blistering summer afternoons — so “summers don’t feel hotter” can be locally true.
• The Great Lakes moderate southern Ontario, and this corner of the continent had a genuinely weak mid-century warming spell — so a flat-looking stretch on a local line is real, not fabricated.
• Canada’s headline “warming twice as fast” is an Arctic story. Southern Ontario warms closer to the global pace, so an Ontarian doesn’t personally experience the dramatic number they hear on the news.

And the rain? Local precipitation is even noisier than temperature, with no clean trend at a single station — which is exactly why the extreme-rain chart earlier had to be a continent-wide aggregate. Here are the same six cities’ annual totals, each as a percentage of its own mid-century normal:

The precipitation lines mostly wander around 100% with no obvious direction — which is the honest local picture, and another reason a resident concludes “nothing’s really changed.” The takeaway isn’t that doubters are irrational; it’s that lived experience and climate science are measuring different things. One is a single noisy place over a few remembered years; the other is the planet, averaged, over more than a century. When you match the ruler to the claim — average enough places over enough time — the Ontario back yard and the global record tell the same story.

The variability trap: pick a city and a season

The noise that hides the trend isn’t spread evenly across the calendar — and that’s where the misconception really takes root. Winter is both the fastest-warming season and by far the noisiest, so it has the worst signal-to-noise of any season: the warming is genuinely there, but it’s buried under swings several times its size. Summer, quieter, shows its trend more plainly.

Use the tool below to see it for yourself. Pick a city and a season. The left panel is every year’s seasonal average with its warming trend line; the right panel bins those same years into decades as a violin — the shape of each decade’s spread, with every year as a dot — and you can watch the shapes climb while they keep overlapping, or flip it to Year violins or Year boxes, where each year becomes its own scrollable violin (or box-and-whisker) of that season’s daily temperatures — so you can see how an individual winter actually played out, day-spread and all. That overlap is the whole problem: a cold recent winter really can land below a mild winter from decades ago, which feels like a refutation even as the trend marches up underneath. The y-axis is fixed across every city and season, so you can see at a glance how far apart the seasons sit and how differently they spread.

Two things worth noticing as you click around. First, almost every city and season is warming — the trend lines tilt up, some steeply — yet you have to squint past the scatter to trust it, especially in winter. Second, a lot of the warming shows up not as balmy winters but as fewer brutal cold extremes: the bottom whiskers lift faster than the tops. People notice a hot summer day far more than a cold snap that simply didn’t happen — so the most human-perceptible evidence is the evidence we’re least wired to register. The doubt is understandable. The trend is still there.