Spring Predictability Barrier

Ask a model in March what the Pacific will do next winter and you get its least trustworthy answer of the year. That seasonal dip in forecast skill — for predictions made during, or reaching across, roughly February through May — is the spring predictability barrier, and it shapes how every El Niño forecast should be read.

The causes stack. Spring is ENSO's shift change: events peak in winter and decay or flip in spring, so the ocean is often near neutral with little signal to grab. The coupling between ocean and atmosphere runs weakest then, letting small disturbances — a stray westerly wind burst, a spell of trade-wind strength — steer the system disproportionately. Noise dominates signal, and models diverge.

The practical rules follow. Treat February–May forecasts of the coming winter with humility (the media's annual spring "El Niño is coming!" cycle earns its skepticism). Trust June-onward forecasts substantially more — and trust them most when the event is already observed rather than merely predicted.

That is precisely what distinguishes 2026's declaration: by June the ocean was already at strong-event levels with the atmosphere coupled, so the barrier had been cleared with the answer in hand. It's why the June outlook could responsibly attach 88% and 63% odds to strength tiers — numbers a March forecast could never honestly carry — and why each new monthly forecast update outranks the last. The verification runs weekly on the dashboard.