Abstract
Vehicles are one of the most significant sources of air pollutant emissions in urban areas, and their real contribution always needs to be updated to predict impacts on air quality. Radar databases and traffic counts using statistical modeling is an alternative and low-cost approach to produce traffic activities data in each urban street to be used as input to predict vehicular emissions. In this work, we carried out a spatial statistical analysis of local radar data and calculated traffic flow using local radar data combined with different statistical models. Future scenarios about vehicle emission inventory to define public policies were also proposed and analyzed for Belo Horizonte (BH), a Brazilian State capital, with the third-largest metropolitan region in the country. The Normal-Neighborhood Model (i.e., the mixed effect model with random effect in the neighborhood, radar type, and in the regional area) was used to calculate traffic flow in each urban street. Results showed average reductions in CO (4.5%), NMHC (3.0%), NOx (3.0%) and PM2.5 (6.2%) emissions even with an increase in fleet composition (25% in average). The decrease is a result of the implementation of emission control programs by the government, improvements vehicles technologies, and the quality of fuels. Prediction of traffic data from radar databases has proven to be useful for avoiding the high costs of performing origin-destination surveys and traffic modeling using commercial software. Radar databases can provide many potential benefits for research and analysis in environmental and transportation planning. These findings can be incorporated in future investigations to implement public policies on vehicular emission reduction in urban areas and to advance environmental health effects research and human health risk assessment.
[Display omitted]
•The model Normal-Neighborhood was more suitable to perform a spatial distribution of vehicle flow.•Different fleet reduction combinations generate up to a 40% reduction in vehicle emissions.•Mobility and transportation solutions can be proposed using radar data.