Fig 1.
(A) Location and Divisions of study area; (B) Population density distributed in the study area. (Fig 1A and 1B were created by the author (Y. Zhang) based on the data, this is the original copyright.)
Fig 2.
The spatial distribution of bike stations and capacities in the study area.
(A) spatial distribution of stations that were built before and after the system expansion. (B) The number of parking slots at each of bike stations.
Fig 3.
A diagram of computing the weight of each cell.
Table 1.
The aggregate use of the system by all users in March 2012, March2013, and March 2014.
Table 2.
The aggregate use of the system by steady users, former users, and new users on weekdays.
Table 3.
The aggregate use of the system by steady users, former users, and new users on weekends.
Fig 4.
Comparisons of hourly demand during morning peak and evening peak hours at stations.
(1) the points represent each of bike stations. (2)Y axis represents the average number of hourly demand during morning peak hours, and X axis represents the average number of hourly demand during evening peak hours.
Fig 5.
The spatial distribution of demand by all users on weekdays.
Demand represents the number of trips per day. The upper panel represents the result of March 2014. The lower panel represents the results of March 2012 (left) and March 2013 (right).
Fig 6.
Comparisons of daily demand by one group and another group at stations (weekdays).
(1) The points represent each of bike stations. (2) Blue, red, and yellow symbols denote “stations 12”, “stations 13”, and “stations 14” respectively. (3) Figures A describe the user groups of Mar 2012 versus Mar 2013, and figures B describe user groups of Mar 2013 versus Mar 2014.
Fig 7.
Getis-Ord Gi* statistic of the spatial distribution of demand by each user group (weekdays).
Figures A describe the user groups of Mar 2012 versus Mar 2013, and figures B describe user groups of Mar 2013 versus Mar 2014.
Fig 8.
The spatial distribution of D/S by all users on weekdays.
The D/S represents the ratio of trips/day to the number of parking slots. The upper panel represents the result of March 2014. The lower panel represents the results of March 2012 (left) and March 2013 (right).
Fig 9.
Differences in spatial demand by all users between weekdays and weekends.
The “NDemand” represents the difference in normalized demand between weekdays and weekends. The upper panel represents the result of March 2014. The lower panel represents the results of March 2012 (left) and March 2013 (right).
Fig 10.
Differences in spatial demand between user groups of before and after the system expansion (weekdays).
The “NDemand” represents the difference in normalized demand between user groups. The left panel represents the comparison between March 2012 and March 2013, and right panel represents the comparison between March 2013 and March 2014.
Fig 11.
Differences in spatial demand between new users and steady users after system expansion (weekdays).
The “NDemand” represents the difference in normalized demand between new users and steady users. The left panel represents the comparison between new users and steady users in March 2013. The right panel represents the comparison between new users and steady users in March 2014.