Regional vehicle emissions are based on traffic activity data typically obtained from a conventional four-step travel demand model, with the usual limitation of link volumes predicted across multi-hour periods. From an emissions estimation perspective, aggregated (multi-hour) traffic estimates will mask temporal variations, thus biasing vehicle emissions estimates. This study explores and compares emission inventory outcomes using traffic data of differing resolutions. Specifically, using the El Paso roadway network, the paper addresses how regional emissions will change, in both direction and magnitude, after incorporating traffic data generated by dynamic simulation-assignment models. The analysis indicates that using finely resolved traffic data may significantly alter regional emissions estimation. In general, the magnitude of changes in emissions estimates varies with different pollutants and is correlated with facility type and area type. For congested roadway links with large speed variations, averaging time-dependent traffic data tends to result in underestimation of vehicle emissions relative to using detailed temporal information embodied in a bottom-up calculation method. The nitrogen oxide estimation on high-speed links, as well as the hydrocarbon and carbon monoxide estimation on low speed roadways seems to be more sensitive to traffic data aggregation.
Samenvatting