Vehicle-induced air pollution is an important issue in the 21st century, posing detrimental effects on human health. Prediction of vehicle-emitted air pollutants and evaluation of the diverse factors that contribute to them are of the utmost importance. This study employed advanced tree-based machine learning models to predict vehicle-induced air pollutant levels, with a particular focus on fine particulate matter (PM2.5). In addition to a benchmark statistical model, the models employed were Gradient Boosting (GB), Light Gradient Boosting Machine (LGBM), Extreme Gradient Boosting (XGBoost), Extra Tree (ET), and Random Forest (RF). Regarding the evaluation of PM2.5 predictions, the ET model outperformed the others, as shown by MAE of 1.69, MSE of 5.91, RMSE of 2.43, and R2 of 0.71. Afterward, the optimal ET models were interpreted using SHAP analysis to overcome the ET model's lack of explainability. Based on the SHAP analysis, it was determined that temperature, humidity, and wind speed emerged as the primary determinants in forecasting PM2.5 levels.

This research employed machine learning approaches to classify acceptable or non-acceptable particulate matter (PM_2.5) concentrations using a dataset that was obtained from the Nairobi expressway road corridor. The dataset contained air quality data, traffic volume, and meteorological data. The Boruta Algorithm (BA) was utilized in conjunction with the Random Forests (RF) classifier to select the most appropriate features from the dataset. The findings of the BA analysis indicated that humidity was the most influential factor in determining air quality. This was closely followed by the variables of ‘day_of_week’ and the volume of traffic bound for Nairobi. The temperature of the site was determined to have a lower significance. The comparison among different machine learning classifiers for the classification of acceptable and unacceptable PM_2.5 concentrations revealed that the Extreme Gradient Boosting (XGBoost) classifier displayed superior performance in terms of Sensitivity (0.774), Specificity (0.943), F1-Score (0.833), and AU-ROC (0.874). The Binary Logistic Regression (BLR) model demonstrated comparatively poorer performance in terms of Sensitivity (0.244), Specificity (0.614), F1-Score (0.455), and AU-ROC (0.508) when compared to other ML models. The prediction of PM_2.5 has the potential to provide valuable insights to transport policymakers in their deliberations on urban transport policy formulation.

Background: Noise is one of the externalities of transportation systems in cities. To help understand which neighborhoods have noise levels exceeding the World Health Organization’s (WHO) thresholds, scholars have proposed the use of spatial noise maps. Most cities in developed nations have regularly produced these maps. Unfortunately, in developing nations, citizens still lack this important tool to help in the identification of quieter neighborhoods. One such, is Nairobi, the capital city of Kenya hence the reason for this study. In previous studies, varying methods have been used to produce Road Traffic Noise maps, such as Inverse Distance Weighting (IDW), Ordinary Kriging (OK), among others, each of which has its pros and cons. Aims: This study aimed to develop a GIS-based spatial model of Road Traffic Noise distribution in Nairobi city in Kenya. This was motivated by the fact that the ambient environmental conditions of urban neighborhoods are an indicator of how the welfare of their residents is taken care of. Study Design: This study deployed the GIS-based IDW method. Place and Duration of Study: The Department of Civil and Construction Engineering, University of Nairobi coordinated data collection across the city, between 6th July, 2025 and 12th July, 2025. Methodology: A seven-day Road Traffic Noise data for twelve hours from 6 AM to 6 PM using a handheld Class 1 Sound Level Meter was collected. The data was then bundled into two-hour intervals, resulting in six map presentations. GIS-based spatial maps were then subsequently produced using IDW method. Results: According to the maps, the noisiest time falls between 12 PM and 2 PM, while the low-noise periods occur from 8 AM to 10 AM and from 4 PM to 6 PM. Inferentially, noisy road corridors have high speed in the periods mentioned, and quieter corridors are heavily congested over the same time period. Notorious noisy road corridors in Nairobi include Ngong Road, Waiyaki Way, Eastleigh area, and Thika Road, and the quieter neighborhoods are in Runda and Karen areas. Conclusion: This study concludes that the GIS-based noise maps are helpful tools for residents to choose areas of habitation with a tolerable soundscape and to make noise management sustainable, by deploying GIS-based real-time noise maps accessible via the internet web display.