NWP and Reality

Owing to the non-linear nature of the forecast system, output from NWP models does not behave in a simple or regular way.   Errors can be due to:

  • The inherent structure of the NWP model.
  • Errors in the initial handling of information and data.
  • Propagation and non-linear growth of analysis errors moving downstream.


NWP model programs and systems produce forecasts that are true to the physical representation and parameterisation of the atmosphere and its processes.   But minor inconsistencies in initial data can and do amplify.  This is the whole point of the ensemble forecast so that uncertainty may be investigated.  Users should be aware of how best to deal with error growth and understand how forecasts can be improved by later modification.

Model forecasts rely heavily on the data available, particularly new up-to-date data that can be added to the model background fields.  Analysis increments show where there are significant differences between observed data and background fields.  These point out particular areas where the model expectations may be incorrect.  However, there may be others that have not been identified in this way. 

Users must decide whether the data or the background is anomalous or incorrect and then later assess whether the model has adapted, at least partially, to the latest data.   Adjustment to incorrect data, or non-fitting of correct data, both have the potential to amplify/cause forecast errors.

Finally, meteorological disturbances and the associated energy propagate downstream (downstream spread of influence).  This occurs both within NWP models and in reality.  Anomalies in the analysis affect forecasts well downstream during the forecast period.  Some differences will perhaps weaken or amplify to yield some very incorrect results.  Users need assess the development and progress of disturbances arising from observed data departing from background fields.