Read the balance sheet first, then the headline. The headline: the World Meteorological Organization, with the UK Met Office, now gives a 91 percent probability that at least one year in 2026–2030 exceeds 1.5°C above pre-industrial levels, a 75 percent probability that the five-year average does the same, and an 86 percent probability that at least one year is hotter than 2024 — the hottest year on record.
How does the math work here? Let me do it the way I would with any P&L. A 91 percent probability, over a five-year window, is not a tail risk. It is the base case. If a company told you there was a 91 percent chance its largest revenue line would miss plan within five years, you would not call it a scenario — you would call it the forecast. The climate ledger has been accruing this entry for years, and this report is the moment the accrual becomes explicit.
The three numbers in the entry
Take the three probabilities in order, because they are not interchangeable. The first, 91 percent, is that at least one of the next five years breaches 1.5°C. The second, 75 percent, is that the five-year rolling average itself breaches 1.5°C. The third, 86 percent, is that at least one year beats 2024, the current record holder. Those are different lines on the same sheet: one is a spike, one is a sustained level, one is a record. And the math works out such that all three are more likely than not.
The implied annual temperature anomaly for 2026–2030 sits between 1.3°C and 1.9°C above late-19th-century levels. Notice that the range straddles the 1.5°C line. That is not a coincidence; it is the arithmetic of the window. The range was constructed around the number the Paris Agreement named, and the forecast is telling you where the distribution actually sits.
And where the money goes in this forecast is into timing, not direction. The direction — warming — was decided decades ago. The timing — which exact year breaches, whether 2027 does it given the possible El Niño lasting into 2028 — is what this forecast prices in.
Why the five-year average matters more
Let me be healthily sceptical about the 91 percent headline, because it is the number that will travel. A single hot year crossing 1.5°C is a milestone, but it is not the same as the climate system settling at 1.5°C. The Paris Agreement language refers to holding the increase to well below 2°C, pursuing 1.5°C — understood in practice as a long-term average, not a single calendar year buffeted by El Niño.
So the more honest line on the ledger is the second number: 75 percent that the five-year average itself exceeds 1.5°C. That is the number that removes the weather-vs-climate excuse. A single year can be dismissed as an anomaly, natural variability, a strong El Niño. A five-year average cannot be dismissed the same way; it is the climate, printed at quarterly intervals.
So, fair enough — but don’t let the distinction become a dodge. The 91 percent number is the one that gets the attention, the 75 percent is the one that should get the action, and the 86 percent is the one that tells you records are no longer safe. All three belong in the same entry.
The El Niño variable on the schedule
The forecast carries a scheduling note worth reading carefully: El Niño conditions may persist into 2028, which would make 2027 especially likely to set a new heat record. Think of El Niño as the calendar’s biggest scheduled event — a warm anomaly that lands on top of the long-term trend, briefly pushing the annual number toward the top of its range.
I started this piece intending to frame 2027 as a footnote to the main forecast. Let me correct that framing: it is not a footnote, it is the first scheduled test. If El Niño persists into 2028, 2027 becomes the year most likely to produce the first above-1.5°C annual print. That matters because the first breach will be treated — wrongly — as a cliff. It is not a cliff; it is a schedule. The forecast’s job is to tell you when the first payment on the accrued debt is due.
What the report adds beyond the temperature
The report does not stop at the global number, and neither should the reading. It flags that the Arctic is warming faster, and that the Amazon basin is trending drier — two lines that push up the risk of sea-ice loss, drought and wildfire. These are the expense lines attached to the temperature entry.
How does the math work for those lines? Arctic amplification is not a side detail; a faster-warming Arctic feeds the global circulation, affects mid-latitude weather, and accelerates ice melt that raises sea levels on a schedule the world has not fully priced. The Amazon drying is the same story with a different geography: drier conditions compound with existing fire risk, and the basin’s role as a carbon sink erodes precisely when the ledger needs it most. Both are consequential to the balance sheet, and both are the kinds of line items that appear quietly in a forecast annex and then dominate the actuals.
The quarterly rhythm of the climate ledger
The quarterly rhythm tells you more than the press release. The WMO publishes five-year forecasts on a regular cadence, and this one lands at a particular point in that rhythm: after the hottest year on record, before the next round of international climate negotiations. The report’s placement is not neutral. It is the last full forecast to sit on the table before governments meet again to set ambition levels.
There is a difference between the rhythm and the event. The event is the release. The rhythm is the accumulation — each successive five-year forecast has moved the probability of a 1.5°C breach upward, and this is the edition where the probability crossed into the range that finance professionals would call the base case rather than a scenario. That crossing is worth naming. It changes how a rational actor treats the number: you do not hedge against a 91 percent event; you plan for it.
Read the numbers the way a portfolio manager would read a covenant. A covenant that is likely to be breached within five years gets negotiated today, not at the breach date. The same logic applies to the climate ledger: the 91 percent number is the covenant that has moved from probable to effectively certain, and the negotiation window is the next few years.
What gets booked against it
This is where the memo turns from the forecast to the response, and here I want to be commercially blunt. A 91 percent forecast of a 1.5°C breach within five years is an accrual. The professional question is what the world books against it — and the honest answer, right now, is not enough on the mitigation side and a growing amount on the adaptation side.
Let me think about what actually gets built in response to a number like this. Sea defenses get upgraded on a timetable, not a hope — because coastal asset values are on the balance sheet. Drought insurance gets repriced in agricultural regions. Wildfire risk gets folded into property underwriting. These are all accrual responses: the market has already started booking the cost of a warmer world, line by line, because the forecast keeps making the probability hard to ignore.
The mitigation side is the part that is still being argued over as if it were discretionary spending. It is not discretionary; it is the reduction of a known liability. The 91 percent number makes that point with arithmetic that no amount of debate can discount.
The concrete moment in the ledger
There is a moment I keep coming back to whenever a forecast like this lands. It is the release day, the quiet part of the morning before the coverage starts, when the analysts read the numbers without commentary. That is the only time the probability is read at face value — 91 percent, 75 percent, 86 percent — before anyone rounds them into a headline. The gap between the face value and the headline is where the real story gets lost.
Here is the plain reading. Within five years, it is more likely than not that the world sees its first calendar year above 1.5°C, more likely than not that a five-year average crosses the same line, and more likely than not that the current hottest-year record is broken — possibly as soon as 2027 if El Niño persists. The direction was never in doubt; the report prices the timing.
What the report is not saying
Let me be careful to state what this forecast does not claim, because the ledger has to be read both ways. It does not claim the Paris Agreement’s 1.5°C target is formally dead. The target is defined in the agreement as a long-term holding of the increase; a single breached year, or even a five-year average, is the subject of ongoing scientific and diplomatic interpretation. Nor does it claim that warming becomes uncontrollable in 2031 if the number is hit — the word “breach” describes a threshold crossing, not an abrupt cliff in the climate system.
What it does claim, with arithmetic, is that the buffer is gone. The margin between the current trajectory and the 1.5°C line has been spent, and this forecast is the acknowledgment that the payments on the accrued warming are now due within the forecast window. That distinction — between a threshold that is being crossed and a system that is ending — is the difference between reading this as a footnote and reading it as a summons.
One more correction to my own framing: I have been using the word “accrual” as if the climate were a company keeping books. The metaphor holds up to a point, and the point is exactly the probability. A company books an accrual when a liability is probable and its timing is estimable. That is precisely the state of the climate ledger: the liability is probable — 91 percent — and the timing is now estimable — within five years. The metaphor fails, as metaphors do, because the climate cannot restate its books or raise new equity. It can only be paid.
The verdict
Read the balance sheet first: 91 percent, 75 percent, 86 percent — three lines, all above 50, all pointing the same way. The first 1.5°C breach is no longer a tail event; it is the base case. The question the world is answering with its policies is not whether to book the liability, but what to book against it. Fair enough — but don’t call a 91 percent forecast a warning. Call it what it is: an entry on the ledger, with the invoice date now roughly known.
And to close the loop the way a sharp reader would: the distinction between the 91 and the 75 is where the honest policy debate lives. The 91 tells you a breach year is coming; the 75 tells you the average will follow. Nobody who reads ledgers would book the first number and ignore the second. The schedule is set, the first test is 2027, and the world has spent the buffer it used to think it had. The only question left is what gets built, repriced and rewritten between now and the first invoice.