Designing a bicycle and pedestrian traffic monitoring program to estimate annual average daily traffic in a small rural college town. (June 2017)

• Record Type: Journal Article
• Authors: Lu, Tianjun; Buehler, Ralph; Mondschein, Andrew; Hankey, Steve
• Is Part Of: Transportation research. Volume 53(2017)
• Page Start: 193
• Page End: 204
• Publication Date: 2017-06
• Subjects: Non-motorized transportation -- Active travel -- Performance measure -- Traffic patterns; Transportation -- Research -- Periodicals; Transportation -- Environmental aspects -- Periodicals; 354.76
• DOI: 10.1016/j.trd.2017.04.017
• Languages: English
• ISSNs: 1361-9209

Abstract

Highlights: Network metrics (functional class, centrality) can be used for count site selection. All automated counters were highly correlated with the manual counts. Active travel was correlated with road type, facilities, and proximity to campus. Comprehensive monitoring programs can be conducted for active travel. Abstract: Cycling and walking are commonly recognized as energy-efficient alternatives to motorized transport. Research and practice lack a comprehensive set of methods to assess spatiotemporal patterns of traffic volumes across an entire transportation network. Current non-motorized traffic monitoring programs are primarily implemented in urban areas and for singular components of the network (e.g., off-street trails, specific corridors). Our approach synthesizes ongoing efforts in non-motorized traffic monitoring to estimate Annual Average Daily Traffic (AADT), across an entire network in Blacksburg, VA – a small, rural college town. We selected count sites across the network, stratified by street functional class (e.g., major roads, local roads), centrality of the link relative to origins and destinations, and planned bicycle facilities. We collected 45, 456 h of pedestrian and cyclist counts using three types of automated counters: pneumatic tube (n = 12), passive infrared (n = 10), and radio beam (n = 3) at both reference locations (n = 4; 1-year) and short-duration locations (n = 97; 1-week) during 2015. We found a strong correlation between manual validation counts and automated counts. We used day-of-year scaling factors to estimate AADT for bicycles and pedestrians and found that temporal and spatial patterns differed between modes. Pedestrian volumes were higher and more variable than bicycle volumes (median [interquartile range] AADT for pedestrians: 135 [89–292]; bicycles: 23 [11–43]); both modes were positively correlated with street functional class, presence of facilities, and proximity to campus. Our approach provides insight for planners or policymakers interested in comprehensive monitoring programs to track performance measures or for use in environmental and health impact studies.