The agenda today has three items; the first decides the other two. Item one is the number the World Meteorological Organization published on August 25: a 91 percent probability that at least one year between 2026 and 2030 will exceed 1.5 degrees Celsius above pre-industrial levels. Item two is the probability underneath it, that the five-year average itself crosses 1.5 degrees: 75 percent. Item three is the reasoning chain, and it is the item that decides how the board reads the other two.
Read the memo in order — the numbers and the reasoning follow the same line. The joint forecast from the WMO and the Met Office is not a single prediction; it is a probability distribution over the next five years. Eighty-six percent of that distribution contains at least one year hotter than 2024, which is already the hottest year on record. Every year from 2026 to 2030 is projected to land between 1.3 and 1.9 degrees above late-nineteenth-century levels. The numbers and the reasoning point the same way: the 1.5-degree guardrail, as a year-by-year measure, is likely to be breached before this board’s next strategic cycle ends.
Let me think about what a 91 percent probability actually obliges. I have read the WMO’s five-year outlooks across several cycles now, and I can tell you when a forecast changes register; this one did. A probability that high is not a scenario to be hedged against; it is a baseline to be managed. The board’s question is not whether the threshold will be crossed — at 91 percent, that question is already answered in the direction of the crossing. The question is what the organization does in the interval between now and the crossing, and then after it.
The first item decides the other two. If at least one year crosses 1.5 degrees within five years, then the assumptions embedded in every climate-linked budget line — insurance pricing, supply chain stress tests, asset useful lives, even the vocabulary of a company’s sustainability report — need to be re-dated. The reasoning is simple: you do not run a five-year plan on the assumption that a threshold your own forecasters put at 91 percent will hold.
The board’s question is not whether, but how. How does the organization treat the 1.5 degree target when the probability of breaching it in any single year reaches 91 percent, and the probability of breaching it on a five-year average reaches 75 percent? There is a defensible position available, and there is a hope. They are different things, and the difference matters for capital allocation.
The 1.5-degree line, read as an asset
Here is the defensible position. The Paris Agreement threshold is a long-term average target, measured over decades, not a single-year ceiling. A single warm year does not, by itself, invalidate the framework. But the forecast erodes the comfortable version of that argument, because 75 percent for the five-year average is not a footnote; it is the framework’s own metric turning against it. The distinction between a single-year breach and a five-year average breach is real, and it narrows by the month.
Let me be precise about the mechanism the forecast is describing. The strong El Nino pattern, which the WMO now expects to persist into 2028, sits on top of a warming baseline. El Nino redistributes heat in the Pacific; it does not create it. But its persistence raises the odds that 2027 in particular sets a new annual record, because the natural variability and the long-term trend push in the same direction in the same year. That is why the reasoning chain matters more than any single forecast figure: the warm baseline is the constant, and the El Nino is the multiplier.
I want to correct my own framing here, because it is easy to read this as a weather story. It is not. The regional detail in the forecast is the part that makes it an operational problem, not a headline. The WMO notes that the Arctic is warming faster than the global average, and that the Amazon basin is trending drier. Faster Arctic warming accelerates sea-ice loss; a drier Amazon raises the risk of drought and wildfire. Those are not abstract climate statistics; they are input assumptions for commodity prices, shipping routes, agricultural procurement, and physical-risk disclosures in every jurisdiction where this board operates.
There is a geopolitical layer here that a board memo should not pretend away. The forecast lands weeks before COP31, and it will be read in negotiating rooms as the scientific baseline for that session. A 91 percent probability published in the run-up to a climate conference changes the shape of the negotiation: it removes the luxury of debating whether the goal remains reachable and forces the debate onto what happens when the measured path and the agreed path diverge. For an organization with operations in multiple markets, that means the policy environment the forecast creates is part of the operating environment it must plan for. The number does its work in the atmosphere of policy before it does its work in the atmosphere of the planet.
The agenda item that follows from the first is a review of assumptions. If a five-year planning horizon now carries a 75 percent probability of exceeding the threshold that most transition plans use as their boundary condition, then the boundary has moved into the plan period. A transition plan built around a 1.5 degree pathway is not wrong; it is incomplete, because it does not yet price the interval where the pathway is breached and policy responds to the breach. The prudent response is not to abandon the pathway; it is to add a branch for the breach.
Let me think about the second derivative, because boards are paid to see the second derivative. When a 91 percent probability becomes public knowledge, it changes the behavior of other actors before it changes the climate. Regulators read the same forecast. Insurers read it. Rating agencies read it. The forecast does not need to trigger a single storm to trigger repricing; it needs only to be widely read and believed. The window between the forecast and the first extreme event is where the costs move.
That is the part of this memo I expect the reader to push back on, so let me address it directly. The objection runs: forecasts revise; probabilities shift; we have seen predictions before. The counter is not that this forecast is infallible. The counter is that the pattern across forecasts is one-directional. The WMO has revised this outlook upward in successive cycles, and the underlying temperature record keeps setting new marks — 2024 was the hottest year on record, and the distribution puts an 86 percent probability on a hotter year within five years. An organization that treats a one-directional pattern as noise is making a choice, not a calculation.
The governance decision
The decisive item, then, is not the temperature number. It is the governance decision about how the forecast enters the organization. There are three ways a board can treat a 91 percent probability: ignore it and hope; cite it and defer; or act on it and re-date the assumptions. The first two look different but produce the same outcome — the cost shows up later, larger, and outside the control of the people who could have moved early. The third is the only one that keeps the decision inside the organization.
Let me correct a second tendency in my own draft. It would be easy to end with alarm, and alarm is not the assignment. The assignment is sequencing. The useful version of this memo is operational: which reviews should move to this quarter, which indicators should be tracked monthly, and which board committee should own the re-dating of the climate assumptions. Alarm ends the conversation; sequencing starts it. This board exists to make decisions, and a 91 percent probability is the most decision-ready input a forecast can offer.
One more consideration, and I include it because it will come up in the room even if nobody raises it first. The forecast is a global average, and a global average understates the distribution’s edges. The probability bands do not tell a board in one region whether its own supply chains are in the wet tail or the dry tail of the distribution. That is not a flaw in the forecast; it is the next layer of work. The WMO has given the organization the prior; the organization must now build the conditional scenario for its own geographies. A board that treats the global number as the full picture is making the same error as a board that treats a single weather event as proof of the trend, just in the other direction.
The sequence, in four steps
Concretely, the sequence looks like this. First, re-date the transition plan’s boundary assumptions to include a breach branch. Second, run the physical-risk stress test under a 1.5-plus scenario for 2027, which is the year the forecast flags as most likely to set a new record. Third, review insurance and supply chain clauses that reference climate thresholds, because counterparties will reprice the same probabilities at the same time. Fourth, treat the Arctic and Amazon lines in the forecast as commodity and logistics inputs, not as background color. That is the whole memo in one paragraph.
The agenda today has three items, and the first decided the other two. The temperature forecast is the item; the probability is the agenda; the re-dating of assumptions is the decision. Read in order, the numbers and the reasoning follow the same line, and the line leads to a governance choice, not a weather forecast.
What this means for the next cycle
There is a moment I keep returning to when I try to make this concrete. In an earlier planning cycle, I sat through a review where a climate scenario was presented as a fifty-year curve, and the room’s attention drifted because fifty years felt like a different administration’s problem. This forecast compresses that curve into the current planning window. The same slide, redated to 2027, would hold the room. That is the difference between a decision and a hope, and it is why the probability distribution matters more than the headline number.
So here is the board-level summary. The WMO forecast puts a 91 percent probability on a 1.5 degree year within five years, a 75 percent probability on a 1.5 degree five-year average, and an 86 percent probability on a year hotter than 2024, with 2027 flagged as the likely record-setter and the Arctic and the Amazon as the fastest-moving regional risks. Those numbers are the agenda. The decision is whether the organization treats them as a weather note or as an input to this quarter’s assumption review. The first item decides the other two, and this time the first item has a probability attached. That probability is the memo.
One closing note on tone, for the record. This memo is not an argument that the organization should abandon its climate commitments; it is the opposite. A commitment that is priced correctly is stronger than one that is priced at zero and quietly violated. The board that re-dates its assumptions this quarter is the board that keeps its own word plausible when the forecast materializes. That is the entire point of reading the memo in order: the numbers and the reasoning follow the same line, and the line ends in a decision that can be defended in the next cycle, whatever the temperature does.