Analysis of Transport Shares based on Combined Mobile Network and Public Transport Data

Sammanfattning

Public transport is an integral part of society. It allows people’s mobility demands to be met while having the potential to reduce congestion and decrease the amount greenhouse gases released. Sweden is dealing with increased population growth in urban centers and to reach the Swedish climate targets, a functioning public transport system is more important than ever. It is therefore of interest if the amount of public transport travel has increased or decreased during recent years. The increase in availability of mobile network data makes it possible to capture the overall travel demand in areas. By combining mobile network data with public transport data such as sensor or ticket data it is possible to calculate the public transport share of the overall travel in time and space. The data sets were combined through a series of database joins to calculate the public transport share of overall travel. The combined data is then evaluated through an inter quartile range filter to detect outliers and the detected outliers are adjusted through imputation to retain the data for analysis. The public transport share is then evaluated through a descriptive analysis and k-means clustering to try and discern data-driven patterns. Both analyses are conducted in different temporal and spatial resolutions. The results are that the cluster analysis enables categorization of weekdays and weekends based on the public transport share each hour. Cluster analysis can also used in order to classify zones based on their public transport share over time and find zones that have similar trend in their public transport share over time. The analyzes of the public transport share in time shows that work and school commuting stands for a large portion of public transport travel. This causes seasonality in the data during summers when the level of work and school commuting is low. It also shows that public transport share of overall travel has increased slightly especially in 2025. When the county is analyzed on a municipal and district level the larger urban centers of Linköping and Norrköping saw an increase in public transport share, both in total numbers and as a relative change. This effect was far less pronounced in rural areas where the numbers were more stable.