The world’s climate authority has effectively declared that 2024’s record heat was not a peak but a stepping stone. In its Global Annual to Decadal Climate Update, released in June 2025, the World Meteorological Organization projected a 91 percent probability that at least one year between 2025 and 2029 will surpass 2024 as the hottest on record. It also assigned a 75 percent chance that the five-year global average will exceed 1.5 degrees Celsius above pre-industrial levels, the threshold that the 2015 Paris Agreement treated as a guardrail against the worst climate damages.
Those numbers land on a planet still absorbing the consequences of 2024, which the WMO’s State of the Global Climate report pegged at 1.55 degrees Celsius above the 1850-to-1900 baseline. Every year since 2015 has ranked among the ten warmest in the modern instrumental record. The question is no longer whether the planet is warming at a dangerous pace, but how quickly the remaining margin to internationally agreed limits is disappearing.
Where the 1.55°C number comes from
The 2024 anomaly is not the output of a single thermometer network. The WMO draws on six independent global temperature datasets maintained by institutions including NOAA’s National Centers for Environmental Information, NASA’s Goddard Institute for Space Studies, the UK Met Office’s HadCRUT series, and the European Union’s Copernicus Climate Change Service. Each uses different statistical methods to blend land-station readings with ocean-surface measurements, and each fills data-sparse regions (particularly the Southern Ocean and the Arctic) in slightly different ways.
The datasets do not perfectly agree. Differences on the order of a few hundredths of a degree reflect choices about how to interpolate gaps and weight polar regions. But all six converge on the same conclusion: 2024 was exceptionally warm by any measure. When the WMO consolidates them into a single estimate, the resulting 1.55°C figure carries a narrow uncertainty band, giving scientists high confidence in the direction and scale of the trend.
Copernicus, which operates the ERA5 reanalysis system through the European Centre for Medium-Range Weather Forecasts, adds a further layer. Its Climate Data Store provides gridded fields that assimilate observations into a physically consistent atmospheric model, allowing researchers to trace not just global means but regional extremes, seasonal shifts, and vertical temperature structure. That granularity matters because global averages can mask the unevenness of warming: the Arctic, for instance, is heating roughly three to four times faster than the global mean.
What drives the 91 percent forecast
The WMO’s five-year outlook is produced by the UK Met Office on behalf of the organization, using an ensemble of climate models initialized with current ocean and atmospheric conditions. The approach has a strong track record: similar forecasts issued in previous years correctly anticipated the record-breaking warmth of 2023 and 2024.
Two forces dominate the projection. The first is the relentless rise in greenhouse gas concentrations. Atmospheric carbon dioxide surpassed 420 parts per million in 2024, its highest level in at least 800,000 years, and global emissions have yet to peak. That accumulating blanket of heat-trapping gases raises the baseline temperature against which natural variability plays out.
The second force is the El Niño-Southern Oscillation cycle. The powerful 2023-to-2024 El Niño redistributed heat from the tropical Pacific into the atmosphere, amplifying global temperatures. As of early 2026, conditions have shifted toward ENSO-neutral territory. A La Niña phase, which tends to cool global averages temporarily, could delay a new single-year record. But the WMO’s high probabilities suggest its models judge that background warming has already locked in a baseline high enough to make a new record likely regardless of short-term ocean cycling.
The WMO has not published the full ensemble model code or individual member runs behind the 91 and 75 percent figures, so independent replication from public links alone is not currently possible. That is not unusual for operational forecasts of this kind, but it means the probabilities should be understood as expert-system outputs rather than figures any outside analyst can reproduce from scratch.
What 1.5°C actually means, and what it does not
A critical distinction often lost in headlines: the Paris Agreement’s 1.5°C target refers to long-term warming averaged over decades, not a single year or even a five-year window. A five-year average above 1.5°C would not, by itself, mean the Paris threshold has been permanently breached. But it would signal that the world is flirting with that boundary far sooner than most national climate plans anticipated, and that the window for avoiding a sustained overshoot is closing fast.
The practical consequences of warming at and above 1.5°C are already visible. The WMO’s 2024 climate report documented accelerating glacier retreat, record-low Antarctic sea-ice extent for the third year running, and marine heatwaves that triggered mass coral bleaching across every tropical ocean basin. Global sea levels rose at roughly 4.77 millimeters per year over the past decade, more than double the rate of the 1990s. Extreme heat events that were statistically rare in a pre-industrial climate are now occurring with a frequency that strains public health systems, agricultural planning, and energy grids.
None of those impacts switch on at a precise temperature threshold. They intensify along a continuum, and every fraction of a degree matters. But 1.5°C has become a shorthand for a level of warming beyond which certain damages, particularly to coral reef systems, low-lying island nations, and Arctic ecosystems, become very difficult to reverse.
What the numbers do not tell us
Temperature projections, however robust, are boundary conditions for risk, not forecasts of specific outcomes. The WMO’s probabilities do not translate directly into estimates of economic damage, crop failure, displacement, or mortality. Those cascading impacts depend on local geography, infrastructure quality, governance, and adaptation investment, variables that climate models do not capture.
There is also residual uncertainty in the baseline itself. The 1850-to-1900 reference period relies on instrumental records that were sparse, particularly over the Southern Hemisphere oceans. Different agencies fill those gaps with different statistical techniques, and small shifts in baseline construction can move the anomaly by a few hundredths of a degree. When the policy threshold sits at exactly 1.5°C, those hundredths matter for political framing, even if they do not change the physical reality of a rapidly warming planet.
Finally, natural variability has not been abolished. A major volcanic eruption injecting sulfate aerosols into the stratosphere could temporarily cool global temperatures, as Mount Pinatubo did in 1991. An extended La Niña sequence could suppress surface warming for a year or two. Neither scenario would reverse the long-term trend driven by greenhouse gas accumulation, but either could delay the crossing of symbolic milestones and shift public perception of urgency.
Where this leaves the rest of the decade
The WMO’s forecast frames the second half of the 2020s as a period in which the gap between political commitments and physical reality will become harder to ignore. Under the Paris Agreement, nations are expected to submit updated emissions-reduction plans, known as nationally determined contributions, that collectively bend the global temperature curve below 1.5°C. The current round of pledges, even if fully implemented, falls well short of that goal. The United Nations Environment Programme’s most recent Emissions Gap Report estimated that existing policies put the world on track for roughly 2.6 to 2.8°C of warming by 2100.
Against that backdrop, a 91 percent chance of a new record year and a 75 percent chance of breaching 1.5°C on a five-year average are not abstract statistics. They are a measure of how quickly the atmosphere is outrunning the policy response. The data infrastructure behind those numbers, built over decades by agencies like NOAA, Copernicus, and the WMO, is among the most rigorously cross-checked in all of science. What it is telling us, as of mid-2026, is that the trajectory has not changed, and the next few years will test whether anything else will.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
