Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

The Extended Range (monthly) forecasts bridge the gap between medium-range and seasonal forecasting.  However, it is not an extension of the medium-range forecasts but is a completely separate system.  It runs daily from base time 00 UTC out to day 46 with 100 members.  It has horizontal resolution of 36 km, 137 model levels in the vertical and has and-atmosphere and ocean-atmosphere coupling.   A set of extended range re-forecasts provide a basis for ER-M-Climate.

Twice each week, on Mondays and Thursdays, Every day the ENS is extended to run to 46 days ahead at a reduced resolution (currently 36km) but still with land-atmosphere and ocean-atmosphere coupling.  The extended-range forecast provides an overview of the atmospheric evolution for the 32-day period (Day15 to Day46).  The standard ECMWF output products focus mainly on the week-to-week changes in the weather.  The chart products generally are presented in terms of anomalies relative to the extended-range model climate (ER-M-climate) and are mainly shown as 7-day means for calendar weeks Monday to Sunday.  From these one can infer whether conditions (e.g. temperature, rainfall, etc.) are likely to deviate from the average for a location and time of year, and whether such an anomaly could be extreme.

Extended Range forecasts are coupled to the Wave model (ECWAM) and the Dynamic Ocean model (NEMO) run with resolution of ~28km.

Since the ENS and Extended Range are run with different resolutions, the corresponding forecast fields are generated using effectively a different model having the same structure and physics etc, but which are based on a different reduced Gaussian grid in physical space, a different land-sea mask, and a different orography.  Users should be aware of this, especially when generating a time-series of products for a single location that crosses the ENS to Extended Range truncation forecast step (e.g. when generating a time-series of a variable or an meteogram spanning Day15), and develop interpolation procedures that take these changes into account.

...