key: cord-0970546-lhpu9kz7 authors: Song, Jie; Zhang, Liye; Qin, Zheng; Ramli, Muhamad Azfar title: Spatiotemporal evolving patterns of bike-share mobility networks and their associations with land-use conditions before and after the COVID-19 outbreak date: 2021-12-31 journal: Physica A DOI: 10.1016/j.physa.2021.126819 sha: 3d11c65b1485aec500478553649b0690b93fb4dc doc_id: 970546 cord_uid: lhpu9kz7 Recent months have seen ever-increasing levels of confirmed COVID-19 cases despite the accelerated adoption of vaccines. In the wake of the pandemic, travel patterns of individuals change as well. Understanding the changes in biking behaviors during evolving COVID-19 situations is a primary goal of this paper. It investigated usage patterns of the bike-share system in Singapore before, during, and after local authorities imposed lockdown measures. It also correlated the centrality attributes of biking mobility networks of different timestamps with land-use conditions. The results show that total ridership surprisingly climbed by 150% during the lockdown, up from 63,370 trips before the outbreak. Biking mobility graphs became more locally clustered and polycentric as the epidemic develop. There existed a positive and sustained spatial autocorrelation between centrality measures and regions with high residential densities or levels of the land-use mixture. This study suggests that bike-share systems may serve as an alternative mode to fulfill mobility needs when public transit services are restricted due to lockdown policies. Shared-micromobility services have the potential to facilitate a disease-resilient transport system as societies may have to coexist with COVID in the future. Severe acute respiratory syndrome coronoavirus-2 (COVID-19) has inflicted enormous (Wang, 2014) . Accordingly, there is a need of alternative transport modes 85 to satisfy human mobility under such unusual circumstances. Cycling may be such an 86 alternative. There is evidence that the wide popularity of e-scooters in China was partly 87 ascribed to the 2003 SARS outbreak when public transit services were partially closed for 88 a long period (Weinert et al., 2007) . 89 Bike-share has become a recent subject of academic scrutiny. Promoting bike-share, as 90 well as other micro-mobility options such as shared e-scooters and e-bikes, may help lower 91 traffic congestion (Wang and Zhou, 2017) , mitigate gas emissions (Zhu and Diao, 2020) , 92 and in the face of pandemics, enhance the resilience of a urban transportation system 93 (Teixeira and Lopes, 2020). Bike-share is widely deemed as a first-and last-mile connector 94 to other public transport components (Gu et al., 2019). More importantly, it also helps 95 to withstand the disruptions in transportation infrastructure and sustain human mobility 96 needs. Saberi et al. (2018) found that the amount of bike-share trips nearly doubled after 97 the entire subway system was shutdown in London because of workers' strike. They further 98 discovered that the bike-share network was even more well-connected before the strike. In ing that the understanding of COVID-19 impact upon bike-share system is a global inter-127 est. First, an overwhelming theme is the understanding of spatiotemporal changing pat-128 terns of bike-share usages before and after the public crisis. COVID-19 related lockdown 129 policies have resulted in two major transformations regarding spatiotemporal patterns: In summary, multi-disciplinary efforts have sought to explore the spreading mecha-150 nism of contagious diseases among human communities and the role of public transport 151 in this process. Even through we are aware of the potential impacts of the COVID-19 152 outbreak upon the urban transportation system in general, and bike-share in particular, 153 we are far from understanding and quantifying such impacts completely as the epidemic 154 is still evolving. Current limited evidence seems to support the claim that COVID-19 is 155 negatively associated with bike-share trip demands, but more case studies are required to 156 justify this claim. Additionally, how land-use context may be spatiotemporally associated 157 with bike-share usage is still inadequately studied. Therefore, this study analyzed how the 158 dockless bike-share system in Singapore responded to COVID-19. It also attempted to 159 reveal the changing mobility patterns of bike-share users before and during the epidemic, Singapore is a country located in the southeastern Asia and has a total area of 721 163 km 2 and a population of over 500 million, one of the most populous region over the world. As a country advocating public transport, Singapore has one of the most efficient transit covering primary residential, commercial, and employment areas in the state (Fig 1) . The MRT and bus systems contribute to approximate 7.5 million rides per day, over 15 times 169 the daily number of trips made by taxi (Kang et al., 2013) . The dockless bike-share system 170 in Singapore was launched in early 2017, which is a great addition to the public transport 171 system. There were on average over 10,000 daily rides made by shared bikes after its 172 introduction. The bike-share market has undergone expansion and major transformation. sequence numbers. Next, we filtered out the trips with origins or destinations that were 181 out of the geographical boundary of Singapore (referred as location error filter). We also 182 discarded those trips with duration less than 1 minute (referred as short trip filter). In our 183 data set, such trips only accounted for less than 1 percent of all samples. Additionally, we 184 deleted the trips with an average speed greater than 35 km/h (referred as speeding trip 185 filer). In Singapore, the speed limit for cycling is 25 km/h, and we believed that those where p i is the percent of land-use i and n is the number of categories. To compare the bike-share activities during different phases of the COVID-19 devel-200 opment, we first divided the original trip data according to four windows that marked 201 milestones of the pandemic evolution in Singapore. Table 3 where e ij is the edge between neighboring flows F i and F j . Each edge is assigned a weight, 234 the inverse distance between the two flows. In other words, a smaller distance denotes a 235 higher weight, indicating the two flows are more similar. Finally, a community detection algorithm was applied to group similar flows into same 237 clusters. A fastgreedy approach, Louvain algorithm, was employed in this study (Blondel 238 et al., 2008). The algorithm aims at maximizing the modularity of a network, which is a 239 score determining how the community partitioning performs. Modularity quantifies the where d ij is the edge weight (inverse average distance) between nodes (nodes) i and j, F i where in is the sum of the weights (inverse average distances in the network) of all edges 258 within C, tot is that of the weights of those edges incident to nodes in C, K i denotes 259 the total weights of the edges incident to node i, K i,in represents the total weights of the 260 edges originating from i to nodes in C, and M is the total weights of the network. The second phase of the algorithm is to create a new network whose nodes are the nodes. It is calculated by where n is the number of nodes that can reach i, N the total number nodes, and d(i, j) is 301 the shortest path distance between i and j. Node betweenness quantifies the role of a node as a nexus between other nodes. In value P R i = 1 N . Next within each iteration step, PageRank of a node i is updated by where W i is the number of incoming trips to node i, d is a damping factor with a default 314 value of 0.85, N E(i) is the set of nodes that link to i, and L j is the sum of trips of 315 out-going links from node j. The algorithm iterates until there is a convergence. zero, the spatial distribution is random. Similarly, local bivariate Moran's I index can be computed by where I i is the Moran's I index of cell i, and the definitions of the other parameters are Jun. Approximately 52% of trips are within the range of 1 to 25 minutes in duration 348 (Fig 5b) , but on average people spend more minutes on their rides on weekdays than on J o u r n a l P r e -p r o o f Journal Pre-proof weekends (Fig 5d) . Cycling trips are within a range of 200 to 2,400 meters in length (Fig 350 5c ). Bike usage has two surges a day (around 08:30 am and 07:30 pm), with considerably 351 more trips during the pm peak (Fig 5e) . 352 We also compared overall trip patterns during the four periods. Cycling behaviors re-353 main largely congruent in terms of duration (25 minutes on median) and distance (around 354 2,200 meters on median) (Fig 6a and b) . Overall, travelers' temporal patterns are similar 355 across the four windows, with twice usage peaks a day (Fig 6 c and d) . The usage spike 356 during morning rush hours is at around 08 or 09 am on weekdays, while the morning peak 357 occurs at 10 or 11 am on weekends. where the entire city is well-connected and long-distance flows are common. Addition-417 ally, the spatial structure transforms as the COVID-19 situation evolves. Before there is 418 a pandemic, there are two major clusters located in the northern periphery of the city 419 (Fig 9b) . Interestingly, when increasing COVID-19 cases are reported, a surge in bike 420 traffic surfaces in the Downtown Core (Figs 9a and 10a ). There are clearly two types of 421 flow clusters. One cluster is dominated by three red nodes that are closely connected to 422 the others with medium and low weighted in-degrees, a highly hierarchical structure (Fig 423 10b ). The other is over-represented by one central node that is surrounded by numerous 424 nodes with low degrees (Fig 10c) . The enhanced network visualizations help uncover the 425 patterns of how people move around neighborhoods via the bike-share system before and 426 during the COVID-19 crisis. Table 4 , average degree centrality in the fourth window is significantly higher than the 443 pre-pandemic level (Fig 11a) . Closeness centrality assumes important nodes are immediate 444 to other nodes in a graph. Therefore, nodes with large closeness values tend to form local facilities. In particular, we compared the temporal changes of global Moran's I indexes between 509 the residential variable and the centrality of biking mobility graphs ( Table 6 ). The cluster- trips. This omission may have mild impact upon our case study, because Singapore is 576 populous and well served by a dense bus network. Yet, the generalizability of this method 577 should be discussed on a case-by-case basis. The selection of a speed limit filter to remove 578 abnormal trips largely relies on the cutoff speed. The selected threshold may have con-579 tributed to biased filtering of cycling trips. Finally, the modal transfer between bike-share 580 and public transit was not verified because of an absence of public transport data. There 581 are arguments that bike-share may complement or compete with public transport. Simi-582 larly, we infer a competing effect arose during the outbreak, which is subject to validation. Our follow-up tasks will focus on these aspects. The structure of spatial net-591 works and communities in bicycle sharing systems Influencing factors for potential bike-593 sharing users: an empirical analysis during the covid-19 pandemic Fast unfolding of com-596 munities in large networks Shared 601 bicycles in a city: A signal processing and data analysis perspective system on an existing bike-share system in China. Transportation Research Part A: 633 Policy and Practice Examining spatiotemporal changing pat-635 terns of bike-sharing usage during covid-19 pandemic Should bike-sharing continue operating 638 during the covid-19 pandemic? empirical findings from nanjing, china. JOURNAL OF 639 TRANSPORT AND HEALTH 23 A static bike reposi-641 tioning model in a hub-and-spoke network framework A Dynamic Model Averaging for the 644 Discovery of Time-Varying Weather-Cycling Patterns Exploring human movements in Sin-647 gapore: A comparative analysis based on mobile phone and taxicab usages Data Mining org/citation.cfm?doid=2505821.2505826 Impact of covid-19 on usage patterns of a bike-sharing system: Case study 652 of seoul The risk of airborne influenza transmission 655 in passenger cars The bike-sharing system in slovakia and the 658 impact of covid-19 on this shared mobility service in a selected city A robust analysis of the impacts of the stay-at-home policy 661 on taxi and citi bike usage: A case study of manhattan Impacts of covid-19 on the usage of public 664 bicycle share in london Effectiveness of surgical 667 masks against influenza bioaerosols Transmis-670 sion risk of infectious droplets in physical spreading process at different times: A review COVID-19 pandemic and the stethoscope: Do not forget to sanitize 673 49 How does our natural and 676 built environment affect the use of bicycle sharing? Transportation Research Part A: 677 Policy and Practice Vulnerability and resilience of transport systems -A 679 discussion of recent research How can airborne transmission of 687 covid-19 indoors be minimised? Environment International 142 How can airborne trans-691 mission of covid-19 indoors be minimised? Environment international 142 Cycling 693 in the era of covid-19: Lessons learnt and best practice policy recommendations for a 694 more bike-centric future cal Activity During the COVID-19 Outbreak in South Korea: Implications 697 for Physical Distancing Strategies Integrating Bicycling and Public Transport in North Amer-701 ica Assessment of Models to Estimate Bus-Stop Level 704 Transit Ridership using Spatial Modeling Methods. Journal of Public Transporta-705 tion From non-cyclists to frequent cyclists: Factors 708 associated with frequent bike share use in New York City Understanding 711 the impacts of a public transit disruption on bicycle sharing mobility patterns: A case 712 of Tube strike in London The impacts of covid-19 on older adults' active trans-715 portation mode usage in isfahan, iran Examining the associations between 718 perceived built environment and active travel, before and after the covid-19 outbreak 719 in shiraz city, iran Impacts of covid-19 721 pandemic on user behaviors and environmental benefits of bike sharing: A big-data 722 analysis Understanding the usage of dockless bike sharing 724 in Singapore Updates on COVID-19 (Coronavirus Disease Local Situation The link between bike sharing and subway use during the 731 COVID-19 pandemic: The case-study of New York's Citi Bike The motivations for using bike sharing during 734 the covid-19 pandemic: Insights from lisbon Recognition of aerosol transmission 738 of infectious agents: a commentary How Change of Public Transportation Usage Reveals Fear of the 740 SARS Virus in a City Bike-sharing systems and congestion: Evidence from us cities Bus stop, property price and land 747 value tax: A multilevel hedonic analysis with quantile calibration The transition to electric bikes in China: 750 History and key reasons for rapid growth Coronavirus Death Toll and Trends Unravel the 756 landscape and pulses of cycling activities from a dockless bike-sharing system A spatiotemporal and graph-760 based analysis of dockless bike sharing patterns to understand urban flows over the 761 last mile. Computers, Environment and Urban Systems 77 The deadly coronaviruses: The 2003 SARS pandemic and the 2020 novel coronavirus 765 epidemic in China Cough aerosol in healthy participants: Fundamental knowledge 768 to optimize droplet-spread infectious respiratory disease management Detecting the dynam-772 ics of urban structure through spatial network analysis Understanding spatiotemporal patterns of biking behavior by analyzing 776 massive bike sharing data in Chicago A reliable traffic prediction approach for bike-sharing system by exploit-780 ing rich information with temporal link prediction strategy Unravel the impact of covid-19 on the spatio-784 temporal mobility patterns of microtransit Mapping Large Spatial Flow Data with Hierarchical Clustering Understanding the spatiotemporal patterns of public bicycle us-790 age: A case study of Hangzhou