The traffic noise model embedded in ISO 717-1 was originally developed with the 1980s Nordic traffic data, raising concerns regarding its applicability to the contemporary traffic situation. This study conducted field measurement of current traffic noise on representative urban roadways to evaluate the suitability of these legacy rating criteria. Results showed that modern traffic noise spectra exhibited significant divergence from the ISO reference, with discrepancies reaching up to 10 dB at 100 Hz. When applying both spectral datasets to assess the sound insulation performance of windows and composite wall systems, the standard exhibited a consistent tendency to overestimate low-frequency energy components while simultaneously underestimating the actual performance of building elements. Finally, psychoacoustic analysis revealed that the proposed spectral correction term failed to yield a stronger correlation between the sound reduction index and subjective perception compared to the existing reference curve. Collectively, these findings suggest that more accurate traffic noise spectrum corrections should not be rely exclusively on the source characteristics, but must additionally account for human auditory perception mechanisms to maintain their validity.
Cai et al. (Mon,) studied this question.