摘要
轨迹时空模式的研究,对于突发公共安全事件监测及政策制定评估具有重要意义.目前大量的研究仅分析了物理轨迹时空模式变化,尚缺乏从群体虚拟轨迹的视角,分析其与物理轨迹的关联性和差异性.因此,本研究选取美国10 个大型样本城市作为研究对象,以OSM网络地图作为虚拟空间,利用地图瓦片访问频率构建群体虚拟轨迹,与物理轨迹对比分析居家令干预政策实施前后时空分布、时空衰减和空间要素偏好的变化.结果表明:①城市中群体虚拟轨迹访问频率具有一致的中心-边缘空间结构,但新冠肺炎疫情导致城市中心吸引力明显减弱;干预政策实施后,群体虚拟轨迹不同空间要素访问频率相比疫情前平均下降67.84%,且不同空间要素同步下降,仅在部分区域中对医疗设施的访问频率增加,谷歌移动指数代表的物理轨迹平均下降48.05%,仅居住场所增长22.15%.②物理轨迹访问频率变化与干预政策时间一致,然而虚拟空间中群体提前一周左右已经进入居家管制的活动模式.研究成果表明群体虚拟轨迹相比物理轨迹对于干预政策响应时间更早、时空衰减更大,为突发公共安全事件监测及政策制定评估提供了新途径.
Abstract
The evolution of information technology has facilitated a shift in human activities from physical to virtual spaces.Analyzing the spatiotemporal pattern changes in group virtual trajectories under the intervention policies of the COVID-19 pandemic provides a novel perspective for assessing the societal impact of emergent public events.Utilizing OpenStreetMap(OSM)as the virtual space and map tile access frequencies to construct group virtual trajectories,this study conducts a comparative analysis with physical trajectories in terms of spatiotemporal distribution,spatiotemporal attenuation,and spatial feature preferences.Ten large U.S.cities are selected as subjects.Spatiotemporal distribution is analyzed through heat maps to depict changes,employing spatial autocorrelation for clustering examination.Spatiotemporal attenuation is assessed by statistically analyzing the decay ratio of access frequencies for group virtual trajectories pre-and post-implementation of stay-at-home policies.Layered analysis investigates spatial decay trends,contrasting with nighttime lights radiance values representing physical space activity.Spatial feature preferences are explored by statistically analyzing temporal changes in group virtual trajectories and the Google Community Mobility Index,representing physical trajectories,across various place types.Using New York as an exemplar,group virtual trajectories are linked with Points of Interest,and the Latent Dirichlet Allocation model extracts thematic preferences of spatial features.The results indicate that,concerning spatiotemporal distribution,the access frequency of group virtual trajectories manifests a spatial structure reminiscent of center-to-edge,aligning with the spatial distribution characteristics of the physical space population as reflected by nighttime lights.However,the COVID-19 pandemic has markedly diminished the allure of the city center,leading to a substantial decrease in the access frequency of group virtual trajectories.In terms of spatiotemporal decay,the access frequency of group virtual trajectories declines by 38%and 63%in March and April 2020,respectively,compared to February.In contrast,the radiance values of nighttime lights in urban areas decrease by 8%and 5%for the corresponding periods.Moving from the city center to the suburbs,both the access frequency of group virtual trajectories and nighttime lights radiance values exhibit exponential decay,with the decline in virtual space activity being more pronounced.Regarding spatial feature preferences,the implementation of stay-at-home policies results in a sharp decline in both the access frequency of group virtual trajectories and physical trajectories.However,both trajectories maintain activity periodicity and consistency.Group virtual trajectories,compared to physical trajectories,undergo a lifestyle transition about a week earlier post-pandemic,indicating that physical trajectories respond in real-time to stay-at-home policies,while group virtual trajectories often demonstrate foresight.After the implementation of stay-at-home policies,the access frequency of group virtual trajectories decreases by 19.79%compared to the Google Community Mobility Index,representing physical trajectories.This indicates a more pronounced spatiotemporal decay for group virtual trajectories,accompanied by synchronized reductions.In contrast,physical trajectories exhibit a noticeable increase only in residential places.This study underscores that group virtual trajectories,when compared to physical trajectories,exhibit swifter responses to intervention policies,undergo more pronounced spatiotemporal decay,and furnish a novel avenue for monitoring emergent public safety events and evaluating policy formulation.
NETL
NSTL
万方数据