Was the forecast right for Paris on Sun Oct 4?
Paris, Île-de-France Region · daily high temperature in °C · forecasts from each model's archived runs, verified against the LFPB observations
14 days out, 4 models put Paris's Sun Oct 4 high anywhere from 17° to 22°;by the final call, the day before, the field had settled to a median of 24°. The day verified at 25° against the LFPB observations — cloudy.
GEM landed closest, off by 0.2°. AIGFS ran 3° cool.
What every model said, at every lead
Daily-high calls for Sun Oct 4 as each model's runs approached it — the last archived run per day. NBM and The Spread's own blend are consensus products and stay out of these ranges and rankings; the Paris scoreboard shows them, labelled.
| Lead | AIFS | AIGFS | ECMWF | GEM | GFS | ICON | UKMO | Range |
|---|---|---|---|---|---|---|---|---|
| 14d · 2026-09-20 | 17° | 17° | 22° | — | 21° | — | — | 17–22° |
| 13d · 2026-09-21 | 18° | 21° | 28° | — | 29° | — | — | 18–29° |
| 12d · 2026-09-22 | 16° | 18° | 18° | — | 29° | — | — | 16–29° |
| 11d · 2026-09-23 | 16° | 18° | 21° | — | 15° | — | — | 15–21° |
| 10d · 2026-09-24 | 21° | 21° | 19° | — | 20° | — | — | 19–21° |
| 9d · 2026-09-25 | 19° | 22° | 15° | 23° | 23° | — | — | 15–23° |
| 8d · 2026-09-26 | 23° | 21° | 22° | 24° | 23° | — | — | 21–24° |
| 7d · 2026-09-27 | 23° | 21° | 19° | 25° | 20° | — | — | 19–25° |
| 6d · 2026-09-28 | 24° | 21° | 20° | 26° | 22° | 22° | 21° | 20–26° |
| 5d · 2026-09-29 | 22° | 20° | 20° | 23° | 24° | 20° | 21° | 20–24° |
| 4d · 2026-09-30 | 23° | 21° | 21° | 23° | 23° | 19° | 22° | 19–23° |
| 3d · 2026-10-01 | 22° | 21° | 22° | 22° | 23° | 22° | 22° | 21–23° |
| 2d · 2026-10-02 | 23° | 21° | 23° | 24° | 19° | 22° | 22° | 19–24° |
| 1d · 2026-10-03 | 24° | 22° | 24° | 25° | 25° | 24° | 24° | 22–25° |
| actual | 25° |
Read it honestly
This is one date, not a league table. A model can win a day it deserved to lose; rankings only mean something over a window. For that, see which model has been most accurate in Paris — or every city The Spread scores.
Forecasts are what each model's archived runs said at the time, captured at The Spread's sweep times — never reconstructed after the fact. Truth is the LFPB observations. The actual values are the ones in hand when this receipt was settled, usually 2 days after the date; a correction published after that is not carried into it. Archive depth varies by city; receipts exist only for dates the archive watched arrive.