Evaluation and modeling of traffic noise in an urban area of Chhattisgarh, India
Abstract
Introduction: Traffic noise modeling is a rapidly growing field. Researchers are continually improving existing models and creating new ones that take into consideration complex aspects such as traffic flow patterns and the influence of geography. This study aims to test few models that may be suitable for the Indian scenario along with development of new model.
Materials and methods: In the present study, evaluation and modeling of traffic noise have been carried out. The study was carried out in 20 locations in Raipur city. Half of the locations were selected for validation of results, and half were selected for studying the best-suited model for our selected area. Six models best suited to our location were selected after performing the literature review in brief. Traffic data was collected, and models were tested.
Results: On comparing the data, it was found that out of six models, the Burgess model was found to be the most accurate, as its predicted noise levels are consistently closest to the measured noise levels across all ten locations. But the coefficient of correlation (R) for this model was found to be in the range of 0.31 to 0.64. Burgess model uses the framework of concentric zones to analyse how noise varies based on location within a city, taking into account factors such as land use, population density, and the types of activities prevalent in each zone. Further, we developed our own model by using the multiple regression method and validated our results. On performing the statistical analysis, highest value of R2 (0.83 and 0.82) were found for locations PL1 and PL8 respectively. Mean Absolute Deviation (MAD) values ranged from 0.859 to 2.175, and Root Mean Squared Error (RMSE) values ranged from 0.884 to 2.203 for all locations.
Conclusion: The high R² values, close to 1, and the low RMSE values indicate that our model fits the data well. Therefore, we can conclude that the developed model is highly suitable for predicting noise levels at our location.
