Effects of improved spatial and temporal modeling of on-road vehicle emissions

Abstract

Numerous emission and air quality modeling studies have suggested the need to accurately characterize the spatial and temporal variations in on-road vehicle emissions. The purpose of this study was to quantify the impact that using detailed traffic activity data has on emission estimates used to model air quality impacts. The on-road vehicle emissions are estimated by multiplying the vehicle miles traveled (VMT) by the fleet-average emission factors determined by road link and hour of day. Changes in the fraction of VMT from heavy-duty diesel vehicles (HDDVs) can have a significant impact on estimated fleet-average emissions because the emission factors for HDDV nitrogen oxides (NO x ) and particulate matter (PM) are much higher than those for light-duty gas vehicles (LDGVs). Through detailed road link-level on-road vehicle emission modeling, this work investigated two scenarios for better characterizing mobile source emissions: (1) improved spatial and temporal variation of vehicle type fractions, and (2) use of Motor Vehicle Emission Simulator (MOVES2010) instead of MOBILE6 exhaust emission factors. Emissions were estimated for the Detroit and Atlanta metropolitan areas for summer and winter episodes. The VMT mix scenario demonstrated the importance of better characterizing HDDVactivity by time of day, day of week, and road type. More HDDVactivity occurs on restricted access road types on weekdays and at nonpeak times, compared to light-duty vehicles, resulting in 5-15% higher NO x and PM emission rates during the weekdays and 15-40% lower rates on weekend days. Use of MOVES2010 exhaust emission factors resulted in increases of more than 50% in NO x and PM for both HDDVs and LDGVs, relative to MOBILE6. Because LDGV PM emissions have been shown to increase with lower temperatures, the most dramatic increase from MOBILE6 to MOVES2010 emission rates occurred for PM 2.5 from LDGVs that increased 500% during colder wintertime conditions found in Detroit, the northernmost city modeled.Implications: Air quality model performance relies partly on on-road mobile source emission inventories accurately allocated, both spatially and temporally. This work demonstrates the importance of characterizing the mix of heavy-duty diesel versus lightduty gasoline vehicle activity on an hourly basis on weekdays and weekends by road type. Incorporating detailed activity data increases weekday average NO x and PM emissions 5-15%, with early morning hour emission increases approaching 100%, compared to using one average vehicle activity mix. Application of the methodologies described in this paper will improve the accuracy of on-road emission inventories in the understanding of ozone photochemistry and PM formation.

Measured and modeled CO and NO y in DISCOVER-AQ: An evaluation of emissions and chemistry over the eastern US

“…Incorrect emission factors are also another possible source of error. Lindhjem et al (2012) found that the use of MOVES2010 emissions factors increased NO x emissions by 50% in comparison to those found in the previous EPA model, MOBILE6. While further studies with the recently released 2011 NEI are needed, its CO/NO x ratio for Maryland is even lower than the values used in this study.…”

Temporal variation of traffic on highways and the development of accurate temporal allocation factors for air pollution analyses

“…Most of these methods do not account for temporal variation. In addition, the temporal profiles used to estimate hourly emissions may not reflect the travel patterns of heavy-duty and other vehicles (Lindhjem et al, 2012). …”

Spatiotemporal distributions of ambient oxides of nitrogen, with implications for exposure inequality and urban design

“…Finally, we used MOBILE6.2 rather than the MOVES estimator, which has replaced MOBILE6.2 since this work began. In a recent study of two metropolitan areas, Lindhjem et al (2012) found differences in NO x emissions estimated using MOVES versus MOBILE6. More intercomparison work is needed to fully evaluate the differences and their general implications.…”

“…Bai et al (2007) suggested a speed adjustment based on traffic volume, which was not applied here. Additionally, local deviations from default county distributions of vehicle type and age can impact emissions factors (Lindhjem et al, 2012). Finally, we used MOBILE6.2 rather than the MOVES estimator, which has replaced MOBILE6.2 since this work began.…”