I modelli di previsione del rumore da traffico: variabili caratteristiche e modalità di rilevamento dei dati

The traffic noise forecasting can be carried out using methods with different degree of detail and reliability; they can be classified in three main groups: empirical models (or manual methods) based on regression equations and graphs; scale physical models (simulation reproducing with a high degree of detail very complex territorial conditions); automatic calculation codes (allow to overcome the limits of the empirical models and are less expensive than the physical ones). Obviously the goodness of the results depends on the complexity of the model and on the precision of input data. The paper propose a formulation of empirical models showing how several models can be set up id reliable data are available after in situ measurements. This makes several formulation possible, but it emphasizes the lack of reliability in experting them to differemt contexts, maiking the current empirical models very limited, even if well performing in the contexts in which they have been calibrated.