Measurement and prediction of road traffic noise at different floor levels of buildings in a mid-sized Indian city

Abstract

It is widely accepted that residents of high-rise buildings located beside roads are inevitably exposed to significant levels of noise pollution due to traffic operation. This study aims to unveil crucial insights into the dynamic fluctuations of noise levels within the vertical dimensions of high-rise buildings nestled amidst the bustling streets of mid-sized Indian cities where heterogeneous traffic is under operation. Two cycles of data were collected on all floors for hourly duration on the façade line for two high-rise buildings. Traffic volume, speed, distance of the building from centre of the carriageway and honking were used to develop models for Leq and L10 using Artificial Neural Network (ANN). The R2 values obtained for Leq and L10 were 0.93 and 0.90, respectively indicating the efficacy of the models in predicting noise level at all floors. Results show that the sound level rises to a certain height of the building and then begins to decrease. Floor level 3, 4, 5 and 6 (height approx. 8.4–21 m) had the maximum noise level compared to the other floors, while the 1st and floor level above 10 were quieter floors. Frequency analysis reveals that buildings closer to roadways experience dominant honking noise, particularly at 3.15 kHz, whereas traffic noise prevails at frequencies up to 200 Hz. This work has the novelty of addressing the impact of heterogeneous traffic noise at various floor levels of Varanasi mid-sized city which has narrow carriageway. Such a traffic has prominence of honking which is modelled perhaps for the first time. The outcome of this research has importance in urban planning and formulating legislations.