Analysis and Visualisation of Vehicle Emission Data
- 1. Source: Karl Ropkins, supervisor Contact: MOHD YASAR AHMAD Email: ts16mya@leeds.ac.uk Supervisor: Dr. Karl Ropkins; 2nd Marker : Dr Ann Jopson References Ropkins K., DeFries T. H., Pope F., Kemper J., Kishan S., Fuller G. W., Li H., Sidebottom J. (2017). SOME OBSERVATIONS BASED ON COMPLEMENTARY INTERNATIONAL EVALUATIONS OF EDAR VEHICLE EMISSIONS REMOTE SENSING TECHNOLOGY Riverside, Californina. Available from:http://www.cert.ucr.edu/events/pems/presentations/KRopkins_EDAR_PEMSPaper2017_v2.pdf Analysis and Visualization of Vehicle Emission Data Project by: MOHD YASARAHMAD, MSc. Transport Planning Figure-1 Figure-2 On-going Data Analysis The London Blackheath EDAR data will be analysed in detail within this study. One of the approaches which was also used in previous EDAR studies is Pareto analysis, which will be used to analyse the ‘high emitters’. The above example of a Pareto analysis, taken from the earlier study, illustrates one of the methods that is being used in on-going work. In this case the accumulative Pareto curve for NO emissions (calculated using EDAR data and fuel consumption estimates from the merged data sources) shows that 80% of all emissions were caused by 33% of the vehicles (Figure-1) and also that in this case buses and taxis were significant contributors to total passing fleet emissions (Figure-2). The NEXT STEPS of the project in line with the objectives will be the analysis of the fleet data subsets based on Euro class, age etc. In addition, analysis and visualization will focus on the vehicles that are emitting most in these sub-sets to investigate ‘hit emitters’ contribution. Background Road vehicle emissions are a major source of air pollutants. Historically dynamometer was used as a laboratory testing measure to provide emission rates for regulatory purposes. In-situ approaches such as tunnel, remote sensing methods like VERSS, probe vehicle and car chaser studies have been used to provide more “real-world” measures of vehicle emissions. Methodology Statistical programming language “R” will be used along with associated statistical packages. • The package boot will be used for bootstrapping to handle uncertainties when calculating vehicle fleet statistics • The packages lattice and ggplot2 are plotting packages and one or both will be used for visualization of statistical results. • Code recently written as part the EDAR evaluation study will be used to investigate the contributions of different emissions sources including potential ‘high- emitter’ contributions Source: Ropkins et al 2017 Introduction EDAR (Emissions Detecting and Reporting) is a new Vehicle Emissions Remote Sensing Systems (VERSSs) system which can accurately measure Measures CO, CO, NO, and PM etc. EDAR has been recently been evaluated and used to collect vehicle data at three sites in the UK. This study analyses data collected at one of those sites: London Blackheath. Source: Karl Ropkins, supervisor Source: Karl Ropkins, supervisor Source: google maps Vehicle Data Most of the vehicles seen at the Blackheath Hill site were cars and vans. Most of these were only seen once. Smaller numbers of buses and HGVs were also seen, as it is a residential area. Objectives A large number of factor effect vehicle emissions, e.g. the vehicle type, age and EURO category, maintenance level and driver behavior etc. This study will focus on the analysis, visualization and reporting of emissions trends of vehicle fleet and the relationship between emissions and these factors. Location Blackheath Hill is a major arterial route at a junction with a minor road in a residential area. VERSS data collected there was merged with vehicle registration data (e.g. DVLA records) for this analysis. . Previous Evaluation