Part 2 of a study of Capital Bikeshare GPS data
This series of posts discusses a study undertaken in the spring of 2015 where several Capital Bikeshare bikes were outfitted with GPS devices and tracked for several weeks. Read part one, which examines general attributes of trips, here.
As shown in the previous post, GPS data gives us the ability to look at precisely which roads and paths on which users like to bicycle. By separating miles ridden into different categories, I was able to calculate on what type of road member and casual users tend to ride, and visualize where infrastructure and new stations may be needed.
To calculate facility usage, I summed miles-ridden on three different surfaces: bicycle infrastructure (bikes lanes, paths, and protected bike lanes), roads that lack any cycling infrastructure, and, given the large amount of parkland in Washington, I decided to include miles ridden in land managed by the National Park Service (including the National Mall). Roads within NPS land are generally low-traffic, with many sidewalks and paths present. Parts of Beach Drive in Rock Creek Park are also closed to cars on weekends. Of the recorded trips on this road, all were on the weekend. Bike trails within NPS land were considered part of the park layer.
An analysis of D.C. GIS data reveals non-park bicycle facility mileage amounting to 10.58 percent of total bikeable roadway mileage – bike lanes accounting for 7.64 percent and non-park trails 2.94 percent. In this study, bicycle infrastructure was utilized for about 25 percent of miles ridden, indicating that users seek it out. This number jumps to 59 percent when National Park land is included, meaning that all Capital Bikeshare users ride on streets without any cycling infrastructure for about 41 percent of miles.
Further breakdown by membership type shows a large difference between the two user groups. Casual users ride the majority of their miles within National Park land (confirming the earlier heat map) and ride on bicycle infrastructure for less than half of their other miles. Perhaps better wayfinding could direct out-of-towners to the city’s bicycle infrastructure, improve their bicycling experience, and make it more likely that they will consider bicycling in the future.
Member riders more heavily use bicycle infrastructure, but stay clear of the busier parks. Half of their miles are ridden on roads without any bicycling infrastructure.
To determine most-traversed sections in this category, I further examined member segments outside of cycling infrastructure, as these could be prime candidates for future bicycling infrastructure improvements. The most-traversed segments lacking infrastructure included M Street and Pennsylvania Avenue NW in Georgetown, and Louisiana Avenue NW and Massachusetts Avenue NE near Union Station. 14th Street NW between Florida and Columbia Avenues NW, and K Street between 3rd Street NW and 1st Street NE were also very popular.
High ridership on these blocks are likely because these segments are gaps where a bike lane ends and picks back up further on. DDOT has recommendations in its Multimodal Long-Range Transportation Plan, moveDC [PDF], for additional infrastructure in all of these sections except for the K Street segment. DDOT has proposed protected bike lanes for the Louisiana, Massachusetts, and 14th Street sections identified, and a bike lane for M Street NW. A 2012 study [PDF] by Virginia Tech revealed that casual users were generally unhappy with the city’s cycling infrastructure. DDOT should start with the highly-trafficked segments identified in the above maps, as they will affect the most users.
Part three of this series will determine where casual users like to stop to take in the sights.
Photo: A man rides a Capital Bikeshare bike on the Pennsylvania Avenue NW protected bike lane (Elvert Barnes, Flickr, Creative Commons). Map and chart by the author.