Traffic Noise Predictive Models Comparison with Experimental Data

The road traffic noise control is often performed by means of Traffic Noise predictive Models (TNMs). The most used and studied models are based on a statistical approach, in the sense that a functional relation is postulated, generally a logarithmic function, between equivalent level and some parameters (mainly vehicle flows, heavy vehicle percentage, distance source-receiver, etc.) and a fit on experimental data is performed, in order to tune the model on the data collected in the area under investigation. This procedure produce a kind of "site bias" in the prediction, that does not make the model suitable in any conditions and in any region. In this paper, after having summarized some of the most known models, a comparison between simulated and experimental data is performed, in order to highlight the behaviour of models in two different sites. The strong dependence on the site, and consequently on the vehicle flow volume and typology, will be sketched, together with the influence of climatic parameters.