Estimating the hourly variability of bicycle trip patterns and characteristics from automatic bicycle counters: Case study in Prague, Czech Republic

Abstract

Nowadays it is common for many cities around the world to use automatic bicycle counters laid at main road segments to monitor the cyclists’ traffic flow at any given time in the given location. In most cases, this data is then used to identify the peak period of the bicycle traffic. This paper introduces a method to classify the characteristics of bicycle traffic flows and inferring different proportion trip purposes based on the variations of hourly counts. Whilst of many previous studies investigated the key determinants of bicycle travel demand detected by these counters, none of these studies tried to distinguish different portions of different traffic characteristics in each counting section. In most cases the trip purposes of cycling were grouped into either prevailing utilitarian or prevailing recreational. In reality, however, these recorded bicycle traffic contains both trip purposes. The method which is introduced in this paper was developed and tested on the longitudinal datasets obtained from the bicycle counting system in the city of Prague over five years’ period, i.e. 2010 to 2015. The system consists of 26 counting facilities in various city quarters covering wide variety of road segments. In achieving the study objective, first, a theoretical assumption was done to differentiate three different basic traffic characteristics and their respective hourly variations, i.e.: utilitarian, recreational affected by workhours, and recreational not affected by workhours. These hypothetical hourly variations were tested and confirmed based on the observed Prague sample. Obtained results were discussed with respect to geographical location, built environment and knowledge of local situation. The hourly variations of the recorded bicycle flow were influenced by the build environment factors such as specific land use in the proximity of counting section. The results also indicate some specific routes (and its direction) preferences.