The Bicycle Cycle: A Sinusoidal Modal for Predicting Bicycle Demand

The lack of bicycle demand estimation methods for locations with severe seasonal change stands as a substantial barrier to transportation agencies to plan and design bicycle facilities. As bicycling continues to increase in the United States, there is a growing need for a simple method that can accurately predict the seasonal bicycle demand. Although there are many methods to extrapolate demand from a count they are often complex, limited by location, or require many calibration factors. This paper develops and validates a simply calibrated mathematical model for seasonal bicycle demand using a sinusoidal function that generically fits locations with seasonal change. This function has the ability to estimate average daily bicycle counts (ADB) and average annual daily bicycle counts (AADB) at any location in a community using a community calibration factor. This calibration factor is established ideally using one full year of continuous count data to check its validity. However, the minimum data necessary to approximate a calibration factor is just two short monthly counts one in the winter and one in the summer. Two locations with of annual bicycle count data was set aside and used in a test scenario to compare and validate the model. Ultimately, this model expands the estimation ability of count data by assuming a sinusoidal function for seasonal demand, thus providing a powerful and cost effective aid to transportation agencies.