The use of statistical analysis techniques in the study of urban vehicular traffic noise

Noise pollution phenomena, when they are produced by vehicular traffic, are usually described through the L-eq, to which is added, at most, the study of the time history of the measured sound pressure level during the monitoring. The present paper illustrates the application of statistical analysis methods to a large quantity of sound pressure level data, carried out during a measurements campaign performed in many urban centres, characterized by different vehicular traffic typologies. In particular, it is shown that the individuation and the employment of suitable statistic describers can allow the user to identify the most meaningful noise sources, contributing, in a decisive way, to their acoustical characterization, and to the development of predictive models usable in the noise impact evaluation procedures, as well as in the reduction of the city acoustic classification and of the urban traffic plans. Such techniques furthermore represent a necessary tool for the evaluation of the statistic validity degree of the measurements results especially when the emissive sources features are casual or pseudo-casual as in the case of the urban vehicular traffic noise.