Near-road fine particulate matter concentration estimation using artificial neural network approach

Evidence has shown a strong association between ambient particulate matter and adverse health problems. In urban areas, most of households are located near arterial roads, which are exposure to fine particulate matter directly. Hence, it is critical to understand the near-road fine particulate matter concentration and distribution for the purpose of health risk analysis. This paper applies artificial neural network to estimate the near-road fine particulate matter concentration. Factors influencing the detected concentration are classified into four categories: traffic-related, weather-related, detection location-related and background-related. The estimated values are compared with concentrations detected by monitoring campaigns in Gainesville, FL and Shanghai, China. Distinguished from previous research, this study illustrates the fine particulate matter dispersion and distribution within 50 m near road with portable fine particulate matter detectors and weather instruments. The results indicate that artificial neural network approach is capable of producing accurate estimation of pollutant dispersion near road. Besides, fine particulate matter concentration decayed about a half at 30 m distance from an arterial road in Gainesville, FL. Background contributes to more than 2/3 of the detected value at roadside in Shanghai, and the distance–decay pattern is not as obvious as that in Gainesville, which is different from previous studies reported in the literature. An artificial neural network model performs better after removing the background concentration and with higher concentration value of fine particulate matter.