A Modeling Study Found That Congestion Pricing Implemented on a Typical Freeway in Los Angeles Can Reduce Nitric-oxide Emissions by Roughly 10 Percent During Morning Peak Hours.
Researchers Estimated Environmental Benefits for Implementing a Congestion Pricing Strategy in Los Angeles.
Made Public Date

Using Big Data to Estimate the Environmental Benefits of Congestion Pricing in the Los Angeles Metropolitan Area

Summary Information

In response to traffic congestion concerns, transportation agencies often consider congestion pricing strategy as a solution, through imposition of a fee for driving during congested hours, in hopes that drivers will be incentivized to alter their commuting habits or use other modes of transportation. Congestion pricing would also result in a direct co-benefit of pollution reduction, which was the objective of this study. Specifically, the magnitude of the pollution reduction co-benefit generated by pricing congestion was measured in this study using two statistical models. First model examined the effects of traffic congestion, measured by cars per miles, on Nitric oxide (NO) and Nitrogen dioxide (NO2) emissions of vehicles on freeways. Second model related speed with NO and NO2 emissions from vehicles on local roads. Real-time emissions data in the Los Angeles area were collected during a three-month period from August to October, 2019. In addition Freeway Performance Management System (PEMS) and weather data from National Weather Service were utilized for the analyses.


Emissions data were obtained through a mobile measurement technique in real-time using two equipped cars driving around Los Angeles, from Monday through Friday from 6AM to 6PM. The readings were then aggregated in five-minute intervals and three-digit latitude and longitude. A simple regression analysis was used where logged emissions readings by location at specific times was regressed on logged cars per mile, considering speed bins and a series of fixed effects. In order to capture how speed affects pollution through the performance of the vehicle, three speed bins were created: less than 20 mi/h, 20-50 mi/h, and greater than 50 mi/h. At speed less than 20 mi/h, vehicles were in heavy traffic and will stop-and-go frequently. Between 20-50 mi/h, vehicles were closer to their optimal performance, and at speeds greater than 50 mi/h, engine performance was also sub-optimal due to acceleration. Weather covariates, such as temperature, relative 

humidity, and wind speed were also taken into consideration. A measure of ten-minute lagged pollution was used to capture accumulation of pollution and fixed effects for hour of the day interacted with day of the week were used to capture temporal variations in traffic.


  • Results pointed to an elasticity of emissions with respect to traffic congestion in the order of 0.223 and 0.136 for NO and NO2, respectively, in the morning peak on freeways. That is, an increase in one percent of cars per mile in the morning peak (roughly 11 vehicles) would result in a 0.223 percent increase in NO emissions (roughly 0.2712 parts per billion).
  • Based on the elasticities found in this study, removing 500 cars per mile in the morning peak on a typical freeway with average speed between 20 and 50 mi/h would translate roughly into a 10 percent reduction of NO emissions on freeways, assuming these vehicles were not to appear in other locations or other time periods. A similar reduction was also estimated in the afternoon peak for the same speed range. This value should be interpreted as the upper bound effect on reduction in pollution.
  • Results obtained for the local roads suggested that, on local roads the effects of speeding on pollution, due to vehicles operating sub-optimal, appeared to be a bigger concern. For example, in the morning peak, moving from the middle to the highest speed bin, the elasticity of NO emissions moved from 0.077 to 0.098. This result also suggested that, for the middle speed bin for instance, the percentage change (increase) in NO emissions would be 0.077 percent resulting from a one percent change in cars per mile in the morning peak.
Results Type