Developing a modelling system to supplement and distribute road traffic volumes from odometer readings

Statistics Norway has recently developed a database of road traffic volumes based on odometer readings. The database provides annual figures for the total vehiclekilometres (VKms) performed by Norwegian passenger cars, buses and goods road vehicles on national and foreign territory.

In order to provide the information requested in Eurostat’s annual collection of road traffic volumes, Statistics Norway will have to estimate total road traffic VKms for Norwegian motorcycles, mopeds, other motorized vehicles and bicycles. Additional information will also have to be obtained in order to distribute the total road traffic VKms by territory, nationality of vehicle, type of road and gender and age of driver. 

The objective of the project described in this report has been to establish a modelling system that will enable Statistics Norway to provide information on the various levels of detail requested by Eurostat (as described in Appendix 1). The chosen method is a top-down model that combines annual figures for the total VKms from the database of road traffic volumes with information from other available data sources, such as transport surveys, travel surveys, traffic counts and the Norwegian Register of Driving Licenses.

The quality of the estimates of road traffic VKms derived from the top-down model will be dependent on the quality of the estimates of total road traffic VKms and the quality of the estimated distribution factors. The quality of the estimated total road traffic VKms is generally considered to be good, although changes in traffic volumes caused by economic or social trends may be slow to register.

The quality of the estimated distribution factors is dependent on the quality of the underlying data sources (as described in chapter 3 and chapter 4). Some of the main surveys that provide the necessary data for estimating the distribution factors, such as the Norwegian travel survey, are typically carried out every four years. By relying on estimated distribution factors that remain constant over several years,the top-down model for estimating road traffic VKms will not be sensitive to short term changes in driving patterns. T

The main challenge when constructing the Norwegian road traffic VKm model has been to identify existing data sources for calculating distribution factors of good or fairly good quality. There is, for instance, a definite lack of good data sources for the road traffic volumes performed by Norwegian vehicles abroad and by foreign vehicles on Norwegian territory. Efforts should be made to replace distribution factors of poor quality with distribution factors based on more accurate and relevant data.

On the whole, the Norwegian road traffic VKm model is likely to produce road traffic volume estimates of adequate quality, although the quality assessments of individual estimates may range from good to uncertain. It is a strength that the model unites all the relevant data which are available about Norwegian road traffic volumes, and that the model is flexible enough to allow input data and distribution factors to be updated whenever more accurate or relevant data are available.

In 2010, both Eurostat and the United Nations Economic and Social Council (UNECE) adopted a modified structure of the tables in the voluntary data collection of road traffic volumes. The consequences of these modifications to the Norwegian road traffic volume model are discussed in chapter 6. Some measures planned by Statistics Norway in order to improve the availability and quality of road traffic volume data are outlined in chapter 7.