predictive modeling of traffic-related air pollution in urban areas: insights from road mid-block sections

Roads and highways are vital to a nation's economic and social development. however, the surge in transportation demand has led to heightened vehicular emissions, particularly at urban mid-block sections. this study presents predictive models for estimating pollutant concentrations of co, hcho, tvocs, pm₂.₅, and pm10 a green house gas (ghg) co2 based on traffic flow and vehicle type data collected across 18 urban mid-blocks in warangal, tirupati, and vijayawada. three models such as mlr, svr, and ann were developed, with ann achieving the highest performance (r² > 0.90 for all pollutants). peak concentrations of co (1180 ppm) and pm₂.₅ (over 100 µg/m³) were observed during evening hours (7–8 pm), coinciding with traffic volumes exceeding 4000 pcu/hr. a strong correlation (r² > 0.7) between traffic volume and pollutant levels was confirmed across all models. these findings provide actionable insights for urban transport planners to forecast and mitigate traffic-related air pollution at mid-block sections in similar urban environments.