Versions Compared

Old Version 3

changes.mady.by.user Bob Owens

Saved on

compared with

New Version 4

changes.mady.by.user Bob Owens

Saved on

Key

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

Table of Contents

Some known problems

IFS output is constantly monitored for bias or inconsistency between forecast and observed values and a list of inconsistencies is maintained and dynamically updated with a view to rectification of the problems.  Users also can play a part by identifying and reporting when certain weather parameters are being poorly forecast.  Focus should be on issues that occur repeatedly, rather than on one-off failures.  Notifications can be made by email to servicedesk@ecmwf.int (please use subject "Feedback on forecast performance").  Once it has been corroborated, an issue may be added to ECMWF's "official list" for attention and ultimate removal of the problem in a future release (Cycle) of the IFS models.  Keep up-to-date with known IFS forecasting issues.  It should be noted that those issues that have been marked as resolved (indicated by grey text in the link) will still be relevant when inspecting historical data.

...

    • The Gulf Stream does not separate correctly in the western North Atlantic leading to incorrect warming/cooling of some sea areas.
    • In shallow waters, rapid changes due to upwelling can takeplace close to land leading to incorrect sea-surface temperatures.
    • Sea temperatures and salinity may not be correct in coastal waters (where there may be fresh water and lower salinity).
    • Locally incorrect sea ice cover (e.g. the gulf of Finland tends to be relatively uniformly covered with sea ice of a set concentration, or just open water).
    • Summer meltponds in sea ice sheets should be retained longer as sea ice.
    • Spurious sea ice can also occur also in deep water.
    • Spurious areas of ice act as if an island causing reduction of wave heights downstream.
    • Observations are not used in regions where the model sea depth is less than 500m in order to avoid assimilating data on the continental shelves where the ocean model has poor representativeness.

Additional Information

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

...