An emission model to predict hourly street level traffic emission for air quality management in megacity Delhi

Road traffic emission is considered to be the major source of pollution exposure in megacities around the globe. Traffic emission makes urban air pollution to be highly spatially heterogeneous with sharp concentration gradients that can vary substantially within a few meters near the road. The spatially heterogeneous and temporally varying emissions are required to account for  concentration gradients that have a direct impact on the population exposure to outdoor air pollution. However, estimating such a detailed emission is very complex as it requires multi-category emission factors and a huge amount of georeferenced detailed traffic activity data such as traffic volume and speed, distance traveled, vehicle category share, fuel share, engine share, technology share etc. In the absence of detailed data, emission estimations have been limited to coarser resolution which may not be suitable for high resolution air quality modeling, exposure assessment and management..

Here we present an emission model to estimate multi-pollutant hourly gridded on-road traffic emission over Delhi. The model uses the globally adopted COPERT (Computer Programme to Calculate Emissions from Road Transport) emission functions to calculate the emission as a function of speed for 127 vehicle categories. For traffic activity, the emission model uses advanced traffic volume and speed data for Delhi obtained from TRIPP (Transportation Research and Injury Prevention Programme, IIT Delhi). Further the model considers the congestion (travel time delay based on TOMTOM) and speed-volume relation for different road categories to estimate hourly traffic volume and speed for each road link in Delhi that is used to calculate the hourly emissions using the COPERT emission functions Further, the emissions are gridded at 100 m × 100 m resolution to generate high-resolution spatio-temporal emission maps for Delhi and shown in Fig. 1 for four different hours of the day.

We analyzed the modeled emissions to identify peak emission hours, pollution hotspots and most polluting vehicles. The hourly variation of emissions show distinct bimodal distribution with morning and dominant evening peaks for almost all pollutants linked with congestion and peak traffic.

Figure 1. Estimated gridded NO_x emission at 100m × 100m spatial resolution at different time of the day; the time is displayed in the upper-right corner of each subplot.

The emissions are high near the busy roads and traffic junctions. The emission flux in the central areas of Delhi (Fig. 1) is 40-50% higher than mean emission flux due to the higher road and traffic density and lower average speed. Diesel vehicles have been found to be the dominant contributor to PM, BC and NO_x emission. Our results suggest that the top 5 polluting vehicle categories account for more than half (55% - 91%) of the emissions. This study provides very detailed spatio-temporal emission maps for megacity Delhi, which can be used in air quality models for developing suitable strategies to reduce the traffic related pollution. The developed model can be applied for developing emission inventory and real-time emission with the growing availability of real-time traffic data.