Sfyridis, Alexandros; (2022) Road transport and emissions modelling in England and Wales: A machine learning modelling approach using spatial data. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text Sfyridis_A_2022_Doctoral_Thesis_Final.pdf - Accepted Version Download (7MB) | Preview

Abstract

An expanding street network coupled with an increasing number of vehicles testifies to the significance and reliance on road transportation of modern economies. Unfortunately, the use of road transport comes with drawbacks such as its contribution to greenhouse gases (GHG) and air pollutant emissions, therefore becoming an obstacle to countries’ objectives to improve air quality and a barrier to the ambitious targets to reduce Greenhouse Gas emissions. Unsurprisingly, traffic forecasting, its environmental impacts and potential future configurations of road transport are some of the topics which have received a great deal of attention in the literature. However, traffic forecasting and the assessment of its determinants have been commonly restricted to specific, normally urban, areas while road transport emission studies do not take into account a large part of the road network, as they usually focus on major roads. This research aimed to contribute to the field of road transportation, by firstly developing a model to accurately estimate traffic across England and Wales at a granular (i.e., street segment) level, secondly by identifying the role of factors associated with road transportation and finally, by estimating CO2 and air pollutant emissions, known to be responsible for climate change as well as negative impacts on human health and ecosystems. The thesis identifies potential emissions abatement from the adoption of novel road vehicles technologies and policy measures. This is achieved by analysing transport scenarios to assess future impacts on air quality and CO2 emissions. The thesis concludes with a comparison of my estimates for road emissions with those from DfT modelling to assess the methodological robustness of machine learning algorithms applied in this research. The traffic modelling outputs reveal traffic patterns across urban and rural areas, while traffic estimation is achieved with high accuracy for all road classes. In addition, specific socioeconomic and roadway characteristics associated with traffic across all vehicle types and road classes are identified. Finally, CO2 and air pollution hot spots as well as the impact of open spaces on pollutants emissions and air quality are explored. Potential emission reduction with the employment of new vehicle technologies and policy implementation is also assessed, so as the results can support urban planning and inform policies related to transport congestion and environmental impacts mitigation. Considering the disaggregated approach, the methodology can be used to facilitate policy making for both local and national aggregated levels.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as JSONXMLCSV

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item