Page tree
Skip to end of metadata
Go to start of metadata

Long-Range (Seasonal) Forecast

Other names: System5 Forecast, SEAS5

This is a 51-member ENS that is run monthly, data time 00UTC, on the first day of each month for forecasts to 7 months ahead, and run quarterly for forecasts to 1 year ahead. The products are available on the 8th of each month.

The long-range forecast is a land and sea coupled, 51-member ensemble that is based on an older version of the IFS and that therefore uses slightly different model physics. It runs at 36km resolution, which is lower than the resolution typical of shorter range forecasts.  The long-range forecast provides a broad overview of the atmospheric evolution for the 7-month or 12-month period.  The chart products generally are presented in terms of anomalies relative to the seasonal model climate (S-M-climate) (e.g. to indicate periods which will potentially be warmer or colder than average for the time of year).   An important aspect of long-range forecasts is the identification of broad scale indicators or indices that represent important aspects of the global weather patterns (e.g. El Nino Southern Oscillation (ENSO) index).  Climagrams (a type of meteogram) give a month-by-month picture of the probability of appearance of these patterns, and of their possible amplitude.  The typical skill level of seasonal forecasts, as assessed using re-forecasts, is generally much lower than that of short range forecasts.  Skill levels are typically lower in the extratropics than in the tropics. Indeed in some regions, for some forecast variables, skill levels are vanishingly small. Users should familiarise themselves with verification figures from the outset to avoid over-interpretation of over-reliance on the seasonal forecast charts.

Long-range (Seasonal) Products
The long range output is available on the ECMWF web charts and ecCharts. Guidance on the  Interpretation of the seasonal products is discussed elsewhere in the guide.

Additional Sources of Information

(Note: In older material there may be references to issues that have subsequently been addressed)

  • No labels