Sensitivity Analysis of the Calibration of Dataset for a Road Traffic Noise Multilinear Regressive Model

Quantitative evaluation of noise levels is of great importance for the overall evaluation of inhabited contexts. Road traffic noise is a pervasive kind of pollution, which causes health problems to people exposed to sound levels exceeding 55 dBA. Road Traffic Noise models are very useful in the assessment of noise when no actual data for the calibration are available but suffer from drawbacks when they are validated in a site that is different from the one where the calibration took place. To overcome this issue, a computed dataset of independent traffic variables has been generated, and then validated on a real site traffic dataset, obtaining good results. In this paper, a deeper investigation of the size of the dataset, together with a detailed statistical analysis of the dataset itself and the corresponding computing efforts are presented.