Spatio-temporal analysis of car distance, greenhouse gases and the effect of built environment: A latent class regression analysis. (July 2015)

• Record Type: Journal Article
• Authors: Zahabi, Seyed Amir H.; Miranda-Moreno, Luis; Patterson, Zachary; Barla, Philippe
• Is Part Of: Transportation research. Volume 77(2015)
• Page Start: 1
• Page End: 13
• Publication Date: 2015-07
• Subjects: Greenhouse gas emissions -- Neighborhood typologies -- Latent class regression -- Travel behavior -- Built environment characteristics; Transportation -- Research -- Periodicals; 388.011
• DOI: 10.1016/j.tra.2015.04.002
• Languages: English
• ISSNs: 0965-8564

Abstract

Highlights: Household-level GHGs are determined for 3 OD surveys in Montreal, Canada. Temporal–spatial variations of car usage and GHGs are investigated. BE attributes have a statistically significant but small effect on GHG and distance. It is found that the impact of BE and socio-demographics varies across subgroups. A significant reduction in trip GHGs was observed over the three waves. Employment, income and household structure have an important effect on trip GHG. Abstract: This work examines the temporal–spatial variations of daily automobile distance traveled and greenhouse gas emissions (GHGs) and their association with built environment attributes and household socio-demographics. A GHGs household inventory is determined using link-level average speeds for a large and representative sample of households in three origin–destination surveys (1998, 2003 and 2008) in Montreal, Canada. For the emission inventories, different sources of data are combined including link-level average speeds in the network, vehicle occupancy levels and fuel consumption characteristics of the vehicle fleet. Urban form indicators over time such as population density, land use mix and transit accessibility are generated for each household in each of the three waves. A latent class (LC) regression modeling framework is then implemented to investigate the association of built environment and socio-demographics with GHGs and automobile distance traveled. Among other results, it is found that population density, transit accessibility and land-use mix have small but statistically significant negative impact on GHGs and car usage. Despite that this is in accordance with past studies, the estimated elasticities are greater than those reported in the literature for North American cities. Moreover, different household subpopulations are identified in which the effect of built environment varies significantly. Also, a reduction of the average GHGs at the household level is observed over time. According to our estimates, households produced 15% and 10% more GHGs in 1998 and 2003 respectively, compared to 2008. This reduction can be associated to the improvement of the fuel economy of vehicle fleet and the decrease of motor-vehicle usage – e.g., a decrease of 4% is observed for fuel efficiency rates and 12% for distance according to the raw average estimates from 1998 with respect to 2008. A strong link is also observed between socio-demographics and the two travel outcomes. While number of workers is positively associated with car distance and GHGs, low and medium income households pollute less than high-income households.