A correction framework for improving the robustness of motor vehicle registration data

Abstract

An important key to reducing the environmental impact of motor vehicles is to identify those in-services vehicles which are likely to excessively contribute to air pollution. Such an assessment is dependent on quantifying vehicle scrappage which, in turn, relies upon the provision of temporally consistent motor vehicle registration data. There exist a number of issues that adversely impact on the temporal accuracy of motor vehicle registration data. This paper identifies these issues and proposes a cost effective correction framework for motor vehicle registration time series data. An application to Australian data demonstrated the efficacy of the framework, identifying the need to introduce an additional vehicle category into the data, adjusting annual vehicle counts and removing the erroneous incidence of the number of vehicles of a particular vintage increasing substantially beyond two years after the year of manufacture.