Tipping points
What a threshold beyond which a system does not return actually is, which ones are documented, and why uncertainty about their position is not a reassuring argument.
Updated on 15 November 2026 · Natural hazards, Modelling and projections, Cryosphere, Ocean and sea level
Most consequences of warming are gradual: a little more heat, a little more sea, a little less ice. A few are not. A tipping point is a threshold beyond which a system moves of its own accord towards another state, and does not come back when the cause eases.
What separates a tipping from a worsening
Three properties define the phenomenon, and all three are needed.
There is self-sustainment: past the threshold, the transformation continues without the initial perturbation growing. A feedback that was damping becomes amplifying.
There is irreversibility on human timescales: returning below the threshold does not return the system to its former state. The outward and return paths do not coincide — physicists speak of hysteresis. A vanished ice sheet does not reform at the temperature at which it melted, because once its altitude is lost, its surface sits in warmer air.
And there is a change of state, not a mere intensification: a forest becoming savanna, an ocean circulation changing regime.
The inventory, and what it is worth
The reference survey identifies sixteen elements liable to tip, nine of them planetary in reach. For each it proposes a triggering temperature range, and it is on those ranges rather than on the central values that attention belongs.
McKay et al., 2022Four elements are given as possibly crossable within the current warming range: the Greenland and West Antarctic ice sheets, warm-water coral reefs, and some permafrost regions. The ranges remain wide and confidence uneven between elements.
The Amazon case is distinctive because the threshold there is not only climatic. The forest recycles its own rain: part of the moisture falling in the western basin has already been evaporated several times by the canopy upstream. Deforestation and drought therefore attack the same mechanism from two sides.
Flores et al., 2024The signs that can be observed
A system approaching a tipping does not announce it clearly, but it leaves statistical traces. It takes longer to return to equilibrium after a perturbation — this is critical slowing down — and its variability rises. Such traces are sought in long observational records, and their interpretation remains debated: slowing can have other causes.
The Atlantic meridional overturning circulation is the most studied case. Work finds a weakening relative to the past millennium, reconstructed from indirect indicators for want of a direct measurement long enough.
Caesar et al., 2018Permafrost illustrates a neighbouring difficulty: the Arctic carbon balance depends largely on what happens in winter, a season long under-instrumented. Winter measurements show CO₂ losses above what models assumed.
Natali et al., 2019Why uncertainty is not reassuring
The common objection is that these thresholds are poorly known, therefore speculative. The reasoning reverses.
A wide range means the threshold may lie lower as readily as higher. Facing irreversible damage, the asymmetry is total: erring by overestimating the threshold costs mitigation spending; erring by underestimating it costs a state of the system that no later spending buys back.
Added to this is the possibility of cascades — one tipping triggering another. The survey identifies plausible couplings, but the literature quantifies them poorly, and that gap is itself a matter of the unknown rather than the reassuring.
Kemp et al., 2022That is the underlying argument: the best-studied scenarios are the central trajectories, not the tails of the distribution. Yet it is the tails that carry the risk one cannot insure against.
Sources
- Global Tipping Points Report — University of Exeter
- IPCC AR6 WGI, Chapter 4, Box 4.1: Abrupt Change and Tipping Points — IPCC
Studies cited in this brief
David I. Armstrong McKay et al. · Science · 2022
- Natural hazards
- Modelling and projections
Bernardo M. Flores et al. · Nature · 2024
- Biodiversity and ecosystems
- Natural hazards
Levke Caesar et al. · Nature · 2018
- Ocean and sea level
- Climate observation
Susan M. Natali et al. · Nature Climate Change · 2019
- Cryosphere
- Carbon cycle and greenhouse gases
Luke Kemp et al. · Proceedings of the National Academy of Sciences · 2022
- Natural hazards
- Policy and governance