A data fusion approach for providing valid annual passenger transport statistics during the COVID-19 pandemic

Abstract

Abstract The global outbreak of the COVID19 pandemic dramatically changed people’s life’s and their travel behavior in 2020. Consequently, capturing these changes accurately and providing valid annual transport statistics constitutes a tremendous challenge all over the world. Against this background and the lack of a single comprehensive source of data revealing the ground truth, we present a data fusion approach to provide valid annual transportation statistics for Germany during the COVID19 pandemic. Therefore, we adapted our existing model approach at generating annual, national statistics for Germany on passenger transport. Unlike in the existing model, we do not model the whole year as one, but divide the year into pandemic stages in order to model passenger transport demand as adequately as possible within each stage. Three travel surveys capturing the altered travel behavior in the different stages of the pandemic were used in order to adapt our passenger kilometers travelled (PKT) model, which bases on a cross-sectional national household travel survey which in many countries serves as the data basis for providing annual transportation statistics. The main results show a decline in the overall number of trips in 2020 in Germany of around a quarter and around a third less kilometers travelled compared to 2019. These changes in travel behavior differ considerably between different modes of transport, trip purpose and the different stages of the pandemic in 2020. The results produced were validated on the basis of other studies and further sources of data such as floating car data and automated count stations for bicycle traffic and ensure reliable passenger transport statistics in the years of the COVID19 pandemic.

Keywords

• Data fusion
• National passenger transport statistics
• COVID19
• Travel behavior