CMCC SPS4 seasonal forecast maps
CMCC Seasonal Prediction System 4 forecast maps from Copernicus C3S. Compare CMCC SPS4 with ECMWF, UK Met Office, Météo-France and DWD on synchronized maps.
The Euro-Mediterranean Center on Climate Change’s Seasonal Prediction System 4 (SPS4) uses the CMCC-CM3 coupled model. Its components include CAM6 atmosphere, CLM 5.1 land surface, NEMO 4.2 ocean and CICE6 sea ice. SPS4 became operational in August 2025 and is C3S system 4.
This site uses 50 forecast members and 30 hindcast members per reference year for probabilities. The hindcast reference is restricted to 1993–2016 to match the other products shown here.
Configuration source: CMCC C3S contribution documentation. Identifiers and ensemble schedules: C3S multi-system description.
Coverage and processing on this site
| Centre | Euro-Mediterranean Center on Climate Change, Italy |
|---|---|
| System | SPS4 (C3S system 4) |
| Forecast members used for probabilities | 50 |
| Hindcast members used for probabilities | 30 per hindcast year |
| Reference period | 1993–2016 hindcast climatology |
| Published grid | C3S 1° global output; ±85° latitude shown. Native model grids differ. |
| Imported lead times | Lead months 1–6; four three-month windows. Some systems forecast further ahead. |
| Variables | 2 m temperature, precipitation, mean sea-level pressure, sea-surface temperature, 500 hPa geopotential height, 10 m wind speed and 850 hPa temperature |
| Products | Ensemble mean anomaly, tercile summary, median exceedance, tercile and outer-quintile probabilities |
| Source datasets | C3S monthly anomalies and monthly member statistics. Z500 and T850 use the corresponding pressure-level anomalies and pressure-level member statistics. |
These details describe the imported August and September 2026 runs. Each import resolves the system identifier for that initialization and checks it against the source metadata. Imports are run manually; a new monthly forecast may be available from the centre before it appears here. Check the initialization and system on the map panel.
Monthly ensemble-mean anomalies are supplied by C3S. Probabilities and three-month products are derived locally from operational C3S data; they are not synthetic sample forecasts. Member counts above describe the local probability calculation, not necessarily the full ensemble offered by a centre. See the C3S product definitions and the probability guide for reference and processing conventions.
CMCC against the other models
Each link opens CMCC next to one other centre for the same October 2026 initialization, with a shared valid period and scale. Differences can reflect model formulation, initialization, ensemble sampling and each system’s hindcast climate; they do not establish which model is more skilful.
- CMCC vs ECMWF SEAS5: January 2027 temperature
- CMCC vs UK Met Office GloSea6-GC5.1: January 2027 temperature
- CMCC vs Météo-France System 9: January 2027 temperature
- CMCC vs DWD GCFS 2.2: January 2027 temperature
- ECMWF vs UK Met Office seasonal forecasts
- Europe seasonal temperature outlook: five models compared
- How the seasonal outlook changed between monthly runs