An investigation of road traffic noise on main roads in Greater London

Abstract

Measurements of road traffic noise were made at 46 locations on Single-carriageway hilly main roads, subjected to speed limits, in the Greater London area, where the flow of vehicles was relatively steady. Initially simultaneous recordings were made, as far as possible, at kerbside and 10 m from the kerb but, in the main part of the investigation, observations were made at 18 of the sites, with the microphone in the 10 m position. It was shown, for a given rate of flow of heavy and light traffic, that L10 and L50 do not appear to vary with gradients but, because of the relatively high standard deviations, averaging ± 3 dBA, for these quantities, one cannot rule out the possibility of small increases in L10 and L50 with gradient of up to a maximum of 2 dBA for the range of values of gradient from zero to about 5 per cent, followed by a decrease for higher values. No definite variation of L90 with gradient could be established. The simultaneous recordings at kerbside and at 10 m from the kerb indicated that the differences between the values of L10 for these positions and also between those of L50 remained fairly constant; deviations from these differences for some of the sites could in some cases be attributed to their environmental characteristics. A simple theory was developed for predicting 110 and it was found that the measured values of this quantity generally agreed with the theoretical predictions. Parameters expressing the rates of flow of both heavy and light vehicles as single quantities were established and the variations of the measured values of 150 and 190 could be well correlated with these parameters. A subsidiary investigation confirmed that, for a hilly road having a steep slope, the A-weighted indication of the sound level meter correlated better with subjective responses than the B.- and C.-weighted indications. Measurements with individual vehicles, specially provided for the purpose, did not show any variation of peak sound level with gradient, except for gradients of the order of 10 per cent. Recordings of peak levels from vehicles selected at random at a given site, when the density was low, yielded average values which could be used to predict 110 for that site if given the rate of flow of traffic.