A practical model for predicting road traffic carbon dioxide emissions using Inductive Loop Detector data

Using Southampton, UK as a testbed, 514 GPS driving patterns (1 Hz speed-time profiles) were collected from 49 drivers of different vehicle types and used as inputs to an Instantaneous EM to calculate accurate vehicle emissions. In parallel, concurrent data were collected from ILDs crossed by vehicles during their journeys. Statistical analysis was used to examine relationships between traffic variables derived from the ILD data (predictor variables) and accurate emissions (outcome variable). Results showed that ILD data (when used in conjunction with categorisation of vehicle types) can form the basis for a practical road traffic CO2 EM that outperforms the next-best alternative EM available to LGAs, with mean predictions found to be 2% greater than proxy real-world values.