Abstract:
Noise pollution due to heavy traffic and other sources is a common problem in the port city of Chattogram, the commercial capital and second-largest city of Bangladesh. High noise may lead to adverse auditory and non-auditory health effects as well as loss of efficiency and productivity. There is a lack of studies identifying the relationship between traffic and noise pollution in Chattogram. This study focuses on the assessment of noise pollution on major road corridors of Chattogram city and building empirical noise models to predict the noise level for future changed traffic flow as well as identify the most impactful portion of traffic on the noise level.
Forty one locations along four major road corridors of Chattogram city are selected for noise level and traffic volume data collection. Noise level readings are recorded for 5 minutes along with simultaneous video recording of traffic flow for vehicle counting in three different periods including weekday peak, weekday off-peak, and weekend. The instantaneous noise level recordings are analyzed later to get primary noise descriptors such as Leq, L10, L50, L90, and secondary noise descriptors including Noise Climate (NC) and Traffic Noise Index (TNI) for quantitative assessment of noise pollution. The traffic counts are made for several categories including NMV, light, medium, and heavy vehicles. Empirical models are generated to identify the association between noise pollution and traffic flow.
The quantitative noise assessment suggests that the equivalent noise level is above the recommended noise standard level by ECR’97 in all surveyed locations and the mean equivalent level is highest during weekday peak periods and lowest in off-peak periods. The highest Leq is 91.4 dBA on Station road while the lowest is 69.9 dBA in front of Shilpakala Academy. The commercial areas are prone to relatively higher noise levels and among various routes, the route Swadhinota Complex, Bahaddarhat to CEPZ is affected by the highest form of noise pollution. The community annoyance and dissatisfaction are highest in the off-peak period.
Empirical equations are built through the regression process for Leq, L10, L50, and L90. The regression model for Leq has an R2 value of 0.22 but the L10, L50, and L90 calculation models have slightly higher R2 values, respectively 0.32, 0.56, and 0.58. Modeling with filtered dataset by eliminating the rapid and short-duration noise event affected readings gives better R2 value (0.35) for Leq. Filtering the dataset again by eliminating the instantaneous noise level above 90 dBA yields even stronger model for Leq (R2 value 0.45). The highest co-efficient in different models suggest the light vehicle volume shows the most influence on the noise level. All the best-fitted equations are statistically significant with positive correlations between the logarithm of vehicle count and noise level. The positive association indicates that the increase in traffic count would also increase the noise level. The findings from this study would help the city authority to find critical locations for noise pollution along the surveyed corridor. Furthermore, the use of the empirical models may enable city planners to comprehend the effect on noise level due to altered traffic due to changes in land-use patterns in the city.
