← Back to the dashboard

How we track climate tipping points and permafrost

What the dashboard shows, where the numbers come from, and what they can’t tell you.

The short answer

Some parts of the climate system can change abruptly and stay changed once warming passes a threshold: ice sheets, ocean currents, forests, coral reefs, frozen ground. Scientists call these tipping elements. The dashboard compares today’s long-term global warming, about 1.4 °C above 1850–1900, with the temperature range at which each of 16 tipping elements could tip, and counts how many are already inside their range: six in November 2026.

An element counts as “inside its range” once warming passes the lowest temperature at which it might tip. That does not mean it has tipped. It means scientists can no longer rule it out. One element is further along: the Global Tipping Points Report 2025 judges that warm-water coral reefs have crossed the central estimate of their thermal tipping point, 1.2 °C in that report, after the worst bleaching event on record in 2023–25. The dashboard marks them as crossed on the report’s authority and says so.

How we know

The 16 elements and their thresholds come from Armstrong McKay and colleagues (2022) in Science, the most widely used assessment. For each element it gives a lowest, a central and a highest temperature at which tipping could happen. These are fixed reference values, not measurements, and they carry real uncertainty: AMOC collapse, for example, is placed anywhere from 1.4 to 8 °C.

Today’s warming is long-term warming, because the thresholds are stated for sustained warming and a single year can run a few tenths of a degree hotter or colder, as in an El Niño year. We take it from our Global Temperature Monitor, which applies the Copernicus method to the ERA5 record: the value of a 30-year linear trend at the latest month, about 1.44 °C in October 2026. Two other measures are shown beside it. The Indicators of Global Climate Change 2025, a yearly update of the UN climate panel’s figures, puts human-caused warming at 1.37 °C in 2025, rising 0.27 °C a decade. The last twelve months of the ERA5 record ran above 1.5 °C. NASA’s GISTEMP record is downloaded as a cross-check; NASA measures against 1951–1980, so we add 0.19 °C, NASA’s own conversion to 1850–1900.

The 2025 assessment. The Global Tipping Points Report 2025, led by the University of Exeter with 160 authors, is the broadest recent assessment. For each element where it gives a judgement, the gauge quotes it: ice sheets at risk since at least 1 °C and possibly already past tipping points; permafrost, Greenland, West Antarctica and the sub-polar gyre likely with a tipping point around 1.5 °C; mountain glaciers, boreal forests and the Atlantic overturning around 2 °C; the Amazon at risk in the range 1.5–2 °C once deforestation is counted; and an overturning tipping point that the report cannot rule out has already been passed. Its coral threshold, 1.2 °C, is lower than Armstrong McKay’s 1.5 °C, which is why the page shows both.

Measured signals. For some elements the gauge also shows a measurement from another dashboard on this site, with its date and a link: coral reef stations under bleaching alert, Greenland’s and West Antarctica’s ice loss, the overturning’s strength and trend at 26.5°N, the cold patch south of Greenland, the Arctic winter ice maximum, clearing of the Brazilian Amazon and permafrost ground temperature. These show what is happening, but they are not the threshold test itself. Where this site measures nothing about an element, the gauge says so.

Why the elements are connected

Tipping elements affect each other. Greenland’s meltwater freshens the North Atlantic and, in most models, weakens the Atlantic overturning; a weaker overturning moves the tropical rain belt south, which the West African monsoon depends on, and an overturning collapse would warm the Southern Ocean under West Antarctica. Some links are less clear: models disagree on what West Antarctic meltwater does to the overturning, and on whether the Amazon ends up drier or wetter. Wunderling and colleagues mapped these links, first in 2021 and in a broader review in 2024, and found that most are destabilising; the 2025 report counts the overturning in 45% of the interactions it assessed. The cascade on the dashboard is a simplified picture of those published findings, with the measured numbers from this site. It is not a forecast or a simulation.

How current each part is

What the dashboard can’t tell you

It cannot give a date for collapse. Passing a threshold can commit a system to change that then plays out over very different timescales: a coral reef can die within a few hot seasons, the Greenland ice sheet would take thousands of years to melt away.

The count is a warning score, not a probability. An element inside its range is not one that is about to collapse; it means warming has reached the range where scientists can no longer rule it out. Nor is “crossed” the end of the story for coral reefs: the 2025 report says returning warming below 1.2 °C, and eventually to 1 °C, could still retain functional reefs at meaningful scale.

And the thresholds are a snapshot of one assessment. Some are well constrained, others are wide guesses with few observations behind them, and new research keeps refining them.

Where the idea came from

Timothy Lenton and colleagues set out the idea of climate tipping elements in 2008. A year later, Johan Rockström and colleagues placed climate change among a wider set of “planetary boundaries”. Armstrong McKay and colleagues reassessed the evidence in 2022, adding some elements and dropping others, such as Arctic summer sea ice, where they found no clear threshold.

Further reading

Credits

The thresholds come from Armstrong McKay and colleagues, the 2025 assessment from the Global Tipping Points team at the University of Exeter, the temperature records from the Copernicus Climate Change Service (ECMWF) and NASA’s Goddard Institute for Space Studies, the human-caused warming estimate from Forster and colleagues, the interaction research from Wunderling and colleagues, and the measured signals from the agencies behind the other dashboards on this site. This dashboard collects and presents their work.

Permafrost and methane

The short answer

Permafrost is ground that has stayed frozen for at least two years in a row, often for thousands of years. Northern permafrost holds roughly 1,500 billion tonnes of carbon from plants and animals that never fully decomposed, about twice as much as the atmosphere holds today. As the Arctic warms, nearly four times faster than the planet as a whole, the ground thaws, microbes get to work, and some of that carbon escapes as carbon dioxide and, in wet ground, as methane.

This is a slow feedback, not a “methane bomb”. The permafrost region already emits methane, mostly from wetlands and lakes, and today that is a small share of global emissions. The concern is that it grows steadily for decades as more ground thaws.

How we know

Methane in the air. NOAA’s Global Monitoring Laboratory measures methane in air samples from stations around the world and publishes a global monthly average. Arctic stations, such as Utqiaġvik (Barrow) in Alaska, Alert in Canada and Zeppelin on Svalbard, show how much more methane Arctic air holds than the global average. That surplus points to northern sources, but it cannot tell permafrost apart from wetlands, oil and gas fields or air carried north.

The ground itself. Scientists measure permafrost temperature in boreholes drilled 10 metres or more into the frozen ground, coordinated by the Global Terrestrial Network for Permafrost. The first global comparison of those records (Biskaborn et al. 2019) found that over 2007–2016 continuous Arctic permafrost warmed by 0.39 °C, Antarctic permafrost by 0.37 °C, and permafrost worldwide by 0.29 °C. The page also shows the yearly records of 20 of the longest-running boreholes, some going back to 1978, from the network’s published data set. Every one of them with a long enough record got warmer.

How much methane the region releases. A 2024 assessment of the whole permafrost region (Hugelius et al.) estimates 20 to 51 million tonnes of methane a year for 2000–2020: the lower figure from measurements of methane in the air, the higher from field data and models scaled up. The world emits about 575 million tonnes a year from all sources, about two-thirds of it from human activity.

How current each part is

What the dashboard can’t tell you

It cannot say how much of today’s rise in global methane comes from permafrost. The recent acceleration is actively debated: tropical wetlands, agriculture, fossil-fuel leaks, fires and changes in the chemistry that breaks methane down all play a part. Separating them needs isotope measurements, aircraft campaigns and atmospheric models.

It cannot map individual thaw slumps or thaw lakes either. Abrupt thaw is patchy: one ice-rich hillside can collapse while the ground next to it stays stable. Boreholes measure single points; this page gives the broad picture, not a forecast for a particular place.

Further reading

Credits

The permafrost record exists because field scientists drilled into remote frozen ground and kept going back for decades, the Global Terrestrial Network for Permafrost brings their measurements together, and NOAA’s Global Monitoring Laboratory runs a methane network precise enough to see changes of a few parts per billion. This dashboard collects and presents their work.

← Back to the dashboard