Creation of a runnability index with temporal influences

Abstract

This research focuses on the effect of precipitation on running behaviour of people in the city of Utrecht. From scientific literature, different features of the built environment are identified that influence running behaviour. These different features are used to create a benchmark runnability index, which is then extended with a variable depicting influences on running behaviour depending on the precipitation level. By analysing running routes scraped from Endomodo, the differences between dry, moderate precipitation and heavy precipitation runs were found, which were used to quantify the influences of the environment given the precipitation level. With these new models, the runnability scores at road level and neighbourhood level were determined, which showed that there exist some hotspots in Utrecht for running. Furthermore, with a correlation analysis between the different runnability indices and running data, it was found that the extended runnability indices described the running behaviour for different precipitation levels better, which shows that including weather in the runnability index is useful. Further research should be focused on getting more running data to analyse the differences and similarities between runs with different precipitation levels. With this knowledge, the runnability index could then be improved even further.