Modeling Bicycle Ridership: Challenges, Opportunities, and Scalable Solutions for Northwest Arkansas

 This study, which aims to be spatiotemporally comprehensive of the study region for more rigor and reliability by producing a model, involves a thorough comparative analysis of data from various sources, including the primary datasets of Strava and counter data, then adding covariate and location-specific data to fine-tune and represent the model according to the region's homogeneity. We will tune the values of the model parameters to obtain the best fit for our data from these sources and then assess accuracy by comparing our model predictions for independent ground truth counter measurements not used to fit the model but validated and tested to report accurate estimates using a different model. The primary goal of this study is to find if that counter data can calibrate with the user-fed Global Positioning System (GPS)-derived data from Strava to reduce ridership bias and count bicyclists frequenting the trails that are representative of all riders, both Strava users and non-Strava users. This poster was presented at the 2024 July ESIP Meeting in Asheville, NC (July 22-26, 2024).