Modeling spatial patterns of traffic emissions across 5570 municipal districts in Brazil

Abstract

Traffic emissions pose a major environmental threat to human health in many countries. Most studies of traffic emissions have applied spatial models to achieve more accurate exposure assessments. Spatial analysis is a technique that takes into account various geographic phenomena, providing data that can be communicated to potentially affected people and government agencies considering policies to reduce exposure and adverse health effects. In Brazil, only a few studies have evaluated air pollution and no studies have conducted a spatial analysis of traffic emissions across the entire country. Brazil is a large continental region with 200 million inhabitants, where traffic emissions present a critical health challenge. In this study, we used three geostatistical approaches (Getis-Ord Gi*, K means, and spatial regression) to assess the spatial patterns of traffic emissions for all 5570 municipal districts in Brazil. We identified five groups of municipal districts (spatial clusters) with distinct patterns of traffic emissions based on six variables: population, gross domestic product, urbanization rate, length of highways, human development index, and distance from the state capital. One group represents municipal districts in the Northeast and Northwest regions, which have lower income, traffic, and emissions. In contrast, the others groups represent regions with high income, traffic, and emissions. Finally, we found a significant association between emissions inventories and the six variables used to evaluate spatial clusters. Our results can help inform the design of targeted, cost effective air pollution control strategies to reduce the adverse health effects of traffic emissions in developing countries.