Validazione e rappresentatività dei dati meteorologici

In recent decades, the increase in agrometeorological monitoring networks has made a vast amount of data available which is indispensable for providing indications regarding land management: agroclimatic classifications, agricultural and environmental pollution risk maps, irrigation programs, development and application of crop growth models, etc. In this context, analysis of the quality of data becomes fundamental in order to point out the presence of errors or possible measurement inconsistencies which could influence application results. Verification or validation of data is carried out on two levels: in the first, targeted actions are used to identify obvious errors and possible inconsistencies; in the second, statistical verifications are used to evaluate the spatial-temporal variability of observed phenomena and to define data normality. In the present chapter logical and empirical methods are reported which are used to establish if a "datum" is such or it is simply a "number".