Abstract During the past 100 years, many large cities around the world prioritized individual transportation in cars over more sustainable and healthier modes of transportation. As a result, traffic jams, air pollution, and fatal accidents are a daily reality in most metropolis, in both developed and developing countries. On the other hand, walking and bicycling are effective means of transportation for short to medium distances that offer advantages to both the city environment and the health of its citizens. While there is a large body of research in modeling and analysis of urban mobility based on motorized vehicles, there is much less research focusing on non-motorized vehicles, and almost no research on comparing pedestrian and cyclist behavior. In this paper, we present a detailed quantitative analysis of two datasets, for the same period and location, covering pedestrian and bike sharing mobility. We contrast the mobility patterns in the two modes and discuss their implications. We show how pedestrian and bike mobility are affected by temperature, precipitation and time of day. We also analyze the spatial distribution of non-motorized trips in Greater Boston and characterize the associated network of mobility flows with respect to multiple metrics. This work contributes to a better understanding of the characteristics of non-motorized urban mobility with respect to distance, duration, time of day, spatial distribution, as well as sensitivity to the weather.

Highlights • A novel data-driven, spatial-temporal-weather comparison of pedestrian and bike-sharing mobility. • Cycling (more than pedestrian) flows present a bidirectional pattern, indicating commuting. • Pedestrian (more than cycling) flows present a significant hub-based structure. • Cycling is more affected by the weather in the Winter than walking. • Walking is more affected by the weather in the Summer than cycling.