EFI and SOT are powerful tools for identifying potentially extreme weather compared to climatology for a given location and time of year.  A high EFI value indicates that an extreme event is more likely than usual but the values do not represent probabilities of that event.   Users should note:

  • Extreme weather indicated by EFI and SOT:
    • has been assessed against IFS model climatology as derived for M-climate, ER-M-climate or S-M-climate and these can differ from observed climatology.

    • does not necessarily indicate high impact as:

      • Rainfall has varying significance according to location (e.g. 2mm rainfall in the desert might be very unusual but have little physical impact).

      • Windstorm impact can depend on whether trees are in leaf, whether ground is saturated, stability of buildings, etc.

      • Past history is important but is not directly accounted for (e.g. the impact of a heavy rain event on saturated ground is greater than if the ground can absorb the water).
  • Products are only as good as the IFS model output:
        • Tropical cyclone representation is limited by resolution.
        • The threat from intense, very localised convection is unlikely to be fully captured in time or space.


Although a high EFI value indicates that an extreme event is more likely than usual, the values do not represent probabilities.  Any forecasts or warnings must be based on a careful study of probabilistic information derived from ensemble forecasts.  Examples are meteograms and plumes in addition to the EFI.   Users should note:

  • Potentially extreme situations (e.g. wind storms) are often characterized by high dynamical instability in the atmosphere and consequently large spread among ENS members.  An extreme event might be preceded by wide-ranging shallow slope CDFs yielding EFI values that are not particularly high. 
  • The EFI may indicate that anomalous wind speeds or rainfall rates are more likely than normal, but inspection of the CDF diagram is necessary to show what this means for a specific threshold (e.g. 5mm/12hr).   If the climatological risk is 5% and the predicted probability is 20%, the risk is four times larger than normal, but any action will depend on whether this 20% probability is high enough for a specific end-user to undertake protective action.


In a few cases there could be an apparent discrepancy between EFI values as indicated by the ecChart probe tool and the CDF curve for the same selected location.  This is because the EFI and CDF may use different but nearby grid points.  The discrepancy, however, is rarely significant.

The M-climate on the CDF plot should only be considered as a reference so is not suitable for accurate calculations of EFI and SOT.  The M-climate is lead-time dependent and there might be differences for different lead-times, especially in the tails of the CDF plot.