Optimized Adjustment of Speed Resolution and Time Alignment Data for Improving Emissions Estimations

Abstract

Emissions models involving the use of vehicle-specific power (VSP) have been developed and studied on the basis of instantaneous emissions data and vehicle activity data for more than a decade. However, various issues associated with the data quality have never been fully resolved or even understood. This research sought to identify the best speed resolution and adjustment of time alignment to improve emissions estimations. First, VSP distributions at different speed resolutions were calculated and compared on the basis of the second-by-second speed data. The results indicated that VSP distributions cannot be accurately characterized when the speed resolution is lower than 0.1 km/h or 0.1 mph, and the effect of different speed units (km/h and mph) on VSP distributions can be ignored when the speed can be expressed in at least one decimal digit. Then, a correction for the time alignment that relied on the correlation coefficient of the emissions and the positive VSP values was applied to the emissions data. It was shown that the coefficient of variation at the 95% confidence level of emission rates tended to decrease after the correction was implemented. It was further found that the time alignment differed for different pollutants. Finally, emission factor on the basis of the speed distributions at different speed resolutions and emission rates for before and after the correction of the time alignment were developed.