A threshold can create a misleading kind of certainty. Before it, the story sounds like prevention. After it, the story can sound as though the ending has already been written.
Climate change does not work that way. Crossing a specified warming level does not make all further outcomes identical, and returning below it would not necessarily undo every loss incurred along the way.
That is the point behind the UN’s renewed emphasis on limiting the magnitude and duration of overshoot ahead of COP31. In remarks in Fiji on 7 October, climate chief Simon Stiell said the response must minimise both. The UN’s subsequent account rejected the idea that overshoot is a reason to surrender. These are calls for action, not a new finding that every future degree of warming is fixed. [1][2]
The distinction is urgent because the term is often used without explaining what is being measured—or what can still change.
One warm year is not the whole climate goal
A temperature figure needs a baseline, a geographic scope and a period. A monthly or annual reading above 1.5°C relative to pre-industrial conditions is not the same measurement as sustained long-term global warming at that level.
The World Meteorological Organization has repeatedly made this distinction in its annual-to-decadal outlooks. Its 2026 update anticipates continued exceptional warmth and a high likelihood of temporary annual exceedances. Those are forecasts for specified periods, not a declaration that a single year determines the long-term Paris outcome. [3]
The qualification should not become reassurance that short-term exceedances do not matter. They are observations or projections of exposure in the world people inhabit. It simply prevents one measurement from being substituted for another.
It also matters when comparing reports. An ocean-surface anomaly relative to a twentieth-century average cannot be read as though it were the same number as global land-and-ocean warming relative to 1850–1900. A dramatic figure becomes more informative, not less, when its denominator and period are visible.
Overshoot is a pathway, not a pardon
The IPCC uses temperature overshoot to describe exceeding a specified warming level and later returning to it or below it within a stated period. Its assessment finds that larger and longer overshoot brings greater risks than staying below the level, and that some impacts can be irreversible. [4]
That means two pathways ending near the same temperature can still involve different harm on the journey. A loss that occurs during the hotter period cannot be assumed to disappear because a later average improves.
The IPCC’s assessment identifies risks to less resilient ecosystems and longer-lived physical consequences. It also describes the additional difficulty of returning temperatures downwards if warming weakens natural carbon stores or triggers other feedbacks. These are reasons to minimise the peak and its duration, not to treat a later return as a harmless detour. [4][5]
The phrase “temporary overshoot” can therefore be deceptively gentle. Temporary describes the temperature pathway. It does not guarantee that every ecological, economic or human effect is temporary too.
Removal is not an unlimited undo button
The IPCC explains that sustained net-negative carbon dioxide emissions could reduce a temperature exceedance. It also identifies feasibility and sustainability constraints, with greater overshoot requiring more removal to return to a given level. [4]
That is different from assuming that an unspecified future technology can compensate for any amount of delay. A plan needs to explain the scale, timing and consequences of the removal it relies on. A possibility within a model is not automatically an available service at the required scale.
The practical implication is straightforward: avoiding additional warming reduces the amount of future correction required and the harm experienced in the meantime. It does not eliminate the need to consider removal, but it prevents removal from being used as an excuse to stop examining emissions now.
Adaptation has a similarly important but distinct role. It can reduce exposure or vulnerability in particular settings. It cannot be assumed to make every level of warming equally manageable, nor can a global mitigation target substitute for protecting people facing present conditions.
Fatalism is a bad reading of the evidence
The strongest climate story is neither that a single threshold divides safety from catastrophe nor that missing a goal makes every subsequent effort irrelevant. Both erase differences that matter.
The evidence instead points to a continuing relationship between choices and risk. How far temperatures rise, how long they remain elevated and what systems are exposed shape the consequences. Uncertainty about the exact outcome is not a reason to assume that the direction of action no longer matters.
That is the question to carry into the negotiations ahead. Do proposed measures plausibly limit additional warming and protect people already facing harm? Are their assumptions visible? Is a promise of later correction being used to avoid an earlier decision?
The world does not receive a fresh start after overshoot. Nor does it lose the ability to influence what happens next. The task becomes harder, not meaningless. Every argument about the threshold should leave that distinction intact.
Sources & notes
Explore the sources cited in this article.

