Travel Counts
If people feel it is safe, they are more likely to walk or cycle. Pedestrian and cyclist count data offers an objective alternative to the subjective evaluations of field surveys.
We analysed pedestrian and cyclist count data from automated counters in a US city. To isolate the effect of light from other parameters (purpose and destination) we compared pedestrian/cyclist numbers for a certain hour in the few days before and after daylight savings clock change. This hour tended to dark before clock change and to light after clock change (and vice versa) and was compared with control hours (permanently light or dark before and after clock change) using an odds ratio.
The results show significantly fewer people tend to walk or cyclwhen it is dark.
To validate the findings, the travel count data were analysed using an alternative method of analysis which considered pedestrian and cyclist counts across the whole year rather than just those days near to the clock change.
These results also show that significantly fewer people tend to walk or cycle when it is dark.
In this study the travel count method was extended to investigate the impact on light level on cycling numbers. This was done by plotting the ambient light level odds ratio for each location against a measure of light level at that location.
If the numbers of cyclists and pedestrians reduces after dark, what are they doing instead? There could be a change in travel mode (e.g. to drive rather walk) or there could be a decision to not travel at all. In this paper we investigated the change in numbers of motorised vehicles after dark alongside those of pedestrians and cyclists. The investigation again found a reduction in walking and cycling, but did not reveal any impact on the numbers of motorised vehicles.
Ongoing work
We are working with the police to acess crime data so that these analyses can be repeated for UK data, and extended to investigate the effect of different lighting conditions.