Introduction

An atmospheric numerical prediction model is the main component in a much larger whole termed a numerical prediction suite, which comprises all of the processes implemented for making operational weather forecasts. Among these processes we distinguish the acquisition of meteorological data, the objective analysis (and more generally the observation data assimilation), the forecasting process using one or more atmospheric models, the determination of weather parameters at local scale, and lastly the display of all the available information.

In order to grasp the behaviour of a prediction model, we need to deploy a set of verification methods that allow us to compare predictions with the real situation. There are a good number of quality indices that provide statistical information on the model’s behaviour. It is essential to verify models so as to monitor their imperfections and try to rectify them.