• DOI:10.3141/2247-01
• Corpus ID: 109916042

Pilot Models for Estimating Bicycle Intersection Volumes

@article{Griswold2011PilotMF,
  title={Pilot Models for Estimating Bicycle Intersection Volumes},
  author={Julia B. Griswold and Aditya Medury and Robert James Schneider},
  journal={Transportation Research Record},
  year={2011},
  volume={2247},
  pages={1 - 7}
}

• Julia B. Griswold, Aditya Medury, R. Schneider
• Published 27 December 2011
• Economics
• Transportation Research Record

Bicycle volume data are useful for practitioners and researchers to understand safety, travel behavior, and development impacts. Several simple models of bicycle intersection volumes were developed for Alameda County, California. The models were based on 2-h bicycle counts performed at a sample of 81 intersections in the spring of 2008 and 2009. Study sites represented areas with a wide range of population density, employment density, proximity to commercial property, neighborhood income, and street network characteristics. The explanatory variables considered for the models included intersection site, land use, transportation system, and socioeconomic characteristics of the areas surrounding each intersection. Four alternative models were developed with adjusted R2 values ranging from .39 to .60. The models showed that bicycle volumes tended to be higher at intersections surrounded by more commercial retail properties within1 .10mi, closer to a major university, with a marked bicycle facility on at least one leg of the intersection, surrounded by less hilly terrain within1 .2mi, or surrounded by a more connected roadway network. The models also showed several important differences between weekday and weekend intersection volumes. The positive association between bicycle volume and proximity to retail properties or a large university was greater on weekdays than on weekends, whereas bicycle facilities had a stronger positive association and hilly terrain had a weaker negative association with bicycle volume on weekends than on weekdays. The study found that further testing and refinement was necessary before accurate count predictions could be made in Alameda County or other communities.