Uncertainties and errors in predicting vehicle exhaust emissions using traffic flow models

Vehicle exhaust emissions predicted based on the outputs of traffic flow models are used directly to calculate traffic-related emissions, but also indirectly as input to 'air quality - human exposure' models. This research develops a data-driven methodological framework to probabilistic average speed-based emission predictions using two widely deployed macroscopic traffic flow models. These are the Cell Transmission Model (CTM), a discretised first-order LWR-type model, and METANET, a discretised second-order Payne-type model. Studying both allows quantitative comparison in their application to predicting emissions. While this research discusses all potential sources of uncertainty in this modelling chain, it focusses on those arising from the traffic flow modelling output. The methodology starts with an ensemble-based optimisation approach to estimate both calibration and validation prediction errors in the traffic flow model, and then proposes a Monte Carlo sampling approach to propagate these to emission predictions. This allows predicting emissions alongside their upper and lower bounds for any time period and road network, at different levels of detail. To ensure transferability of findings, this methodology has been tested on three motorway road networks, one of which operates under Variable Speed Limits (VSL). This permits the quantitative assessment of VSL-modified traffic flow models. In the results of this research, emissions of Oxides of Nitrogen (NOx) and uncertainty associated with their prediction are specifically reported for each road network under study. Finally, this research argues that the methodological framework developed can (and should) be applied to any other (relatively) simple or complex integrated 'traffic flow - emission' modelling chain used as part of policy and decision making process.

Media Info

  • Pagination: 1 file

Subject/Index Terms

Filing Info

  • Accession Number: 01660754
  • Record Type: Publication
  • Source Agency: ARRB
  • Files: ATRI
  • Created Date: Feb 20 2018 10:43AM