seasonforecasts

How the seasonal outlook changed between monthly runs

Compare the previous and latest initializations of each seasonal model on the same valid month to see where the outlook held and where it shifted.

Written and maintained by seasonforecasts · Published 11 September 2026 · Reviewed 12 September 2026 · Forecast data imported 12 September 2026 · Sources: C3S seasonal forecast catalogue and product descriptions

Why compare successive runs

Every seasonal centre issues a new run each month. A month later the model has a month of new ocean, land and atmosphere observations and is one lead closer to the target. If the outlook for a given month survives that update it has earned some trust; if it flips, the earlier signal was weak. Run mode on this site places the same model’s initializations side by side, aligned on the valid month, so the only thing that differs between panels is the start date.

Europe temperature: August 2026 run against September 2026 run

Each table shows the months both runs cover. The change column is the September 2026 value minus the August 2026 value; a positive change means the newer run is warmer for that month.

ECMWF

ECMWF: Europe-average 2 m temperature anomaly (°C)
Valid monthAugust 2026 runSeptember 2026 runChange
September 2026+0.83 lead 2+1.22 lead 1+0.39
October 2026+0.96 lead 3+0.66 lead 2−0.30
November 2026+0.93 lead 4+0.74 lead 3−0.19
December 2026+1.12 lead 5+0.99 lead 4−0.13
January 2027+1.10 lead 6+0.93 lead 5−0.17

Largest change: September 2026, +0.39 °C.

UK Met Office

UK Met Office: Europe-average 2 m temperature anomaly (°C)
Valid monthAugust 2026 runSeptember 2026 runChange
September 2026+1.15 lead 2+0.99 lead 1−0.16
October 2026+1.05 lead 3+1.06 lead 2+0.01
November 2026+1.19 lead 4+1.24 lead 3+0.05
December 2026+1.21 lead 5+1.21 lead 40.00
January 2027+1.61 lead 6+1.30 lead 5−0.31

Largest change: January 2027, −0.31 °C.

Météo-France

Météo-France: Europe-average 2 m temperature anomaly (°C)
Valid monthAugust 2026 runSeptember 2026 runChange
September 2026+0.91 lead 2+0.91 lead 10.00
October 2026+0.65 lead 3+0.80 lead 2+0.15
November 2026+0.76 lead 4+0.71 lead 3−0.05
December 2026+0.79 lead 5+0.78 lead 4−0.01
January 2027+0.72 lead 6+0.86 lead 5+0.14

Largest change: October 2026, +0.15 °C.

CMCC

CMCC: Europe-average 2 m temperature anomaly (°C)
Valid monthAugust 2026 runSeptember 2026 runChange
September 2026+0.78 lead 2+1.64 lead 1+0.86
October 2026+0.84 lead 3+1.14 lead 2+0.30
November 2026+0.62 lead 4+0.96 lead 3+0.34
December 2026+0.48 lead 5+0.58 lead 4+0.10
January 2027+0.48 lead 6+0.48 lead 50.00

Largest change: September 2026, +0.86 °C.

DWD

DWD: Europe-average 2 m temperature anomaly (°C)
Valid monthAugust 2026 runSeptember 2026 runChange
September 2026+1.38 lead 2+2.07 lead 1+0.69
October 2026+1.20 lead 3+1.37 lead 2+0.17
November 2026+1.20 lead 4+1.05 lead 3−0.15
December 2026+1.39 lead 5+1.38 lead 4−0.01
January 2027+1.41 lead 6+1.01 lead 5−0.40

Largest change: September 2026, +0.69 °C.

Europe average: 35°N–71°N, 11°W–40°E, land and sea cells, area-weighted. Values are ensemble-mean anomalies relative to each system’s own 1993–2016 hindcast climatology, computed from the same 1° grids the maps draw. Data imported from the C3S catalogue on 2026-09-12.

How to read a change

  • A change smaller than 0.25 °C in the Europe average stays within one colour band on the map. Treat it as no change.
  • The newer run is always at a shorter lead, so it should be sharper. A newer run that is weaker or has changed sign is the interesting case.
  • An area average can hide compensating shifts: warming in the east and cooling in the west cancel. Open the run comparison for the month and look at the pattern.
  • Two runs is a short history. The site accumulates initializations over time, and a signal that persists across three or more runs is a better guide than any single change.
  • Consistency between runs is not skill. A model can be consistently wrong. See comparing seasonal models and monthly runs for how to weigh agreement across models and across runs together.

Other comparisons